Path: chonsp.franklin.ch!not-for-mail From: Neil Franklin Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 03 Jan 2003 22:00:21 +0100 Organization: My own Private Self Lines: 155 Message-ID: <6uvg16uexm.fsf_-_@chonsp.franklin.ch> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> NNTP-Posting-Host: chonsp.franklin.ch X-Trace: chonsp.franklin.ch 1041627624 849 10.0.3.2 (3 Jan 2003 21:00:24 GMT) X-Complaints-To: news@chonsp.franklin.ch NNTP-Posting-Date: 3 Jan 2003 21:00:24 GMT X-Newsreader: Gnus v5.7/Emacs 20.4 Xref: chonsp.franklin.ch alt.folklore.computers:125674 alt.sys.pdp10:12474 dropped the prime stuff, added pdp10 jmfbahciv@aol.com writes: > In article <9dn1va.bu7.ln@via.reistad.priv.no>, > Morten Reistad wrote: > >According to : > >>Yup. We do have different definitions. I had forgotten that my > >>idea of PDP-10 architecture having RISC-flavor was not politically > >>correct. :-) > > > >The PDP1/6/10 instruction set style had lots of things that > >anticipated RISC; Actually it simply predates the CISC "detour". The extended PDP10 (KL-10) made in midst of the CISC days had a few strongly CISCy additions (all the string stuff). > >but it fell through the RISC tests on three > >important counts: Effective address calculation ; sheer number > >of instructions and hardware interfaces. > > All right. Now I'm feeling uneducated :-). How does effective > address calculation affect the category? Gadzooks! RISC ist mainly defined as any architecture where instructions are included or excluded on whether they are "reduced complex", i.e. they can be implemented without getting in the way of the rest of the instructions. So stuff that speeds one specific (sub-)job 10 times, but adds 10% time to all other (sub-)jobs get throun out, if it is not used often enough to speed up by 10% to offset that. Main "show stoppers" are things that stall the pipeline by needing data that has not been prefetched (usually because it is only known to be needed too late) or lengthen the basic clock cycle (usually by requiring complex decoders). The PDP10 EA calculation hits on both these places. The culprint is the indirect bit (bit 13, whose number which is very appropriate :-)). a) it forces an memory read of the indirect address, whose address of that address it is only known after the memory access direct before it. So the pipeline stalls for the full time until memory answers. A "slow" 100MHz (=10ns per cycle) running from 100ns memory (that is full memory access cycle) falls to 10% speed if that happens constantly. A 200MHZ processor is already down to 5% speed with same memory. b) even after reading the word coming back, it may have I bits in it (for double/multiple indirect), so the data can not be immediately used but needs first looking at. So this is an case of waiting for decoder. This only applies to using I. The X indexed stuff is only fast memory indirect (so no access time trouble), and only max one indirection (so can be rolled out into hardware (an X=0 driven multiplexer that takes Y or Y+F(X)). So it is even a case of: AND 2,3(0) is fast and good, while AND 2,@3 is slow and bad! And yes, trying to minimise the impact of such indirect addressing is one of the headaches any fast PDP10 maker (such as FPGS cloners) has got to address, if maximal speed is wanted. This means trying to detect the difference of AND 2,@3 (which can be speeded up) and AND 2,@333 (which can not) and speeding up the just the later. But such detection (a 14bit comparison) slows down basic instruction decoding, thus we have again the "how often" trouble. Does speeding AND 2,@3 up by 3-10% (is even dependant on memory speed!) save more time than slowing all non-@ instructions by about 1%? > I've written > this question three times and it's still very clumsy. It is something from deep within your "guy thing" category. > With hardware interfaces...are you talking about peripherals > or memory? I don't know either what he meant by this. Peripherals usually are not relevant for CPU instruction speed. Nor for CPU classification. > I'll leave the number of instructions to somebody else. I sure > didn't think we had many; quite a few of them were just there > to complete the symmetry. They were a lot easier to document. It is not about reduced count, but reduced complexity. This missunderstanding is actually quite widely spread. > >Interestingly enough, this is where the emulator writers struggle > >with performance. Or rather the host processor running the emulator program. It goes into pipeline stalls, even if it has an pipeline friendly architecture. Software can kill any pipeline. Processors can only offer software (usually the compiler) the chance of avoiding this trouble. An emulator for an non-pipelinable processor will allways stall its host processor. As seen in above AND examples. PDP10s can be programmed pipeline friendly or pipeline stalling. But lots of software prefers I over X usage, because it is simpler (and surely faster on the simple microcoded KS-10, possibly the complex microcoded KL-10 had some "dirty trick" to make it same speed) and more universal (anywhere in memory, not just addresses 01-17). So fast PDP10 makers have to try to optimise at least some common cases, at cost of chip space, and clock speed. I can only hope that the gcc compiler port uses X over I as often as possible (or can be changed to in future optimise its code in this way). But then, X is slower than I on old some traditional microcoded processors, but X is faster than I on FPGAs, and I don't know what KI-10 did here (could be both ways). So this needs to be an target processor model dependant selectable option to the compiler! That would be -mtune=... option to gcc/pdp10, for those that want to know. Thanks to Lars for putting that option in, it will be needed one day in the future. > After I learn more about the three, perhaps we can get to this. > First, what is the emulator emulating? What ever the emulated target architecture does. > >Otherwise the large number of regular registers, orthogonal > >instruction set and flat memory space were RISCy features. > Yes. PDP10 is quite good on many aspects. But the central RISC stuff is not one of them. That "I" bit is the worst anti-RISC thing one can do. -- Neil Franklin, neil@franklin.ch.remove http://neil.franklin.ch/ Hacker, Unix Guru, El Eng HTL/BSc, Programmer, Archer, Blacksmith - hardware runs the world, software controls the hardware code generates the software, have you coded today? ###### Message-ID: <3E1618F4.AB20363B@yahoo.com> From: Peter Flass X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 9 Date: Fri, 03 Jan 2003 23:01:45 GMT NNTP-Posting-Host: 24.194.50.82 X-Complaints-To: abuse@rr.com X-Trace: twister.nyroc.rr.com 1041634905 24.194.50.82 (Fri, 03 Jan 2003 18:01:45 EST) NNTP-Posting-Date: Fri, 03 Jan 2003 18:01:45 EST Organization: Road Runner Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.stealth.net!news.stealth.net!newsfeed1.cidera.com!Cidera!cyclone.nyroc.rr.com!cyclone-out.nyroc.rr.com!twister.nyroc.rr.com.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125698 alt.sys.pdp10:12476 Neil Franklin wrote: > The PDP10 EA calculation hits on both these places. The culprint is > the indirect bit (bit 13, whose number which is very appropriate :-)). > So this is why indirect addressing has gone out of fashion. Pity. I never thought much about it before. I was going to ask when the pendulum was going to swing back towards ease-of-programming over sheer effeciency, but since nearly everything is now programmed in HLLs, ease-of-programming is probably no longer a consideration. ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 4 Jan 2003 02:00:04 -0000 Lines: 29 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> <3E1618F4.AB20363B@yahoo.com> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1041645601 21279 80.177.7.220 (4 Jan 2003 02:00:01 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Sat, 4 Jan 2003 02:00:01 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news-FFM2.ecrc.net!newsfeed.icl.net!newsfeed.fjserv.net!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125741 alt.sys.pdp10:12482 "Peter Flass" wrote in message news:3E1618F4.AB20363B@yahoo.com... > Neil Franklin wrote: > > The PDP10 EA calculation hits on both these places. The culprint is > > the indirect bit (bit 13, whose number which is very appropriate :-)). > > > So this is why indirect addressing has gone out of fashion. Pity. I > never thought much about it before. I was going to ask when the > pendulum was going to swing back towards ease-of-programming over sheer > effeciency, but since nearly everything is now programmed in HLLs, > ease-of-programming is probably no longer a consideration. Hmmm... I find RISC processors *easier* to write code for than their addressing mode laden CISC counterparts... After all if I have enough registers to burn a few on constructing EAs using plain old arithmetic instructions I can construct whatever addressing mode *I* want... The very simplicity of the system model utilised by RISCs make them so much easier to code for. Compare the IA-32 system programming ref to the Alpha equivelent... If that isn't enough to convert you to the RISC cause I don't know what is. :) IA-64 takes CISC and makes it even mode complicated. :/ Cheers, Rupert ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 4 Jan 2003 03:03:06 GMT Organization: TSS Inc. Lines: 26 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> <3E1618F4.AB20363B@yahoo.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1041649386 35589 10.0.0.43 (4 Jan 2003 03:03:06 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 4 Jan 2003 03:03:06 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125743 alt.sys.pdp10:12483 In article , Rupert Pigott wrote: >The very simplicity of the system model utilised by RISCs make >them so much easier to code for. Compare the IA-32 system >programming ref to the Alpha equivelent... If that isn't enough to >convert you to the RISC cause I don't know what is. :) IA-32 isn't CISC. It's just layers of cruft laid down on an originally fairly simple and straightforward 8-bit processor instruction set. In fact IA-32 has relatively few CISC-style bottlenecks compared to systems like the 68000... but for most of the '80s micros didn't tend to use many advanced implementation mechanisms like pipelining and the 68000 family handily beat the pants off the 80[123]86 at equivalent clock speeds. It wasn't until the 486 that the x86 really took off. >IA-64 takes CISC and makes it even mode complicated. :/ IA64 is neither RISC nor CISC, it's VLIW in drag. Overly complex VLIW, admittedly, but that doesn't make it CISC. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> From: Morten Reistad X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: mrr@acer.reistad.priv.no (Morten Reistad) Message-ID: Lines: 219 Date: Sat, 4 Jan 2003 13:09:22 +0100 NNTP-Posting-Host: 212.186.246.50 X-Complaints-To: abuse@chello.no X-Trace: amstwist00 1041693502 212.186.246.50 (Sat, 04 Jan 2003 16:18:22 MET) NNTP-Posting-Date: Sat, 04 Jan 2003 16:18:22 MET Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-lei1.dfn.de!newsfeed.freenet.de!amsnews01.chello.com!amstwist00.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125737 alt.sys.pdp10:12479 According to Neil Franklin : >dropped the prime stuff, added pdp10 > >jmfbahciv@aol.com writes: > >> In article <9dn1va.bu7.ln@via.reistad.priv.no>, >> Morten Reistad wrote: >> >According to : >> >>Yup. We do have different definitions. I had forgotten that my >> >>idea of PDP-10 architecture having RISC-flavor was not politically >> >>correct. :-) >> > >> >The PDP1/6/10 instruction set style had lots of things that >> >anticipated RISC; > >Actually it simply predates the CISC "detour". The extended PDP10 >(KL-10) made in midst of the CISC days had a few strongly CISCy >additions (all the string stuff). There is still an element of orthogonality that seems to be missing from other pre-1970 instruction sets I have seen. >> >but it fell through the RISC tests on three >> >important counts: Effective address calculation ; sheer number >> >of instructions and hardware interfaces. >> >> All right. Now I'm feeling uneducated :-). How does effective >> address calculation affect the category? Gadzooks! [excellent explanation included:] >RISC ist mainly defined as any architecture where instructions are >included or excluded on whether they are "reduced complex", i.e. they >can be implemented without getting in the way of the rest of the >instructions. > >So stuff that speeds one specific (sub-)job 10 times, but adds 10% >time to all other (sub-)jobs get throun out, if it is not used often >enough to speed up by 10% to offset that. > > >Main "show stoppers" are things that stall the pipeline by needing >data that has not been prefetched (usually because it is only known >to be needed too late) or lengthen the basic clock cycle (usually by >requiring complex decoders). > >The PDP10 EA calculation hits on both these places. The culprint is >the indirect bit (bit 13, whose number which is very appropriate :-)). > >a) it forces an memory read of the indirect address, whose address of >that address it is only known after the memory access direct before it. >So the pipeline stalls for the full time until memory answers. A "slow" >100MHz (=10ns per cycle) running from 100ns memory (that is full memory >access cycle) falls to 10% speed if that happens constantly. A 200MHZ >processor is already down to 5% speed with same memory. This is where it helped a lot with a fast memory cache. The effect of this was dramatic, even for the rather tiny cache in the 2065. But consider that the cache is a fix for a problem introduced with the instruction set. >b) even after reading the word coming back, it may have I bits in it >(for double/multiple indirect), so the data can not be immediately >used but needs first looking at. So this is an case of waiting for >decoder. > >This only applies to using I. The X indexed stuff is only fast memory >indirect (so no access time trouble), and only max one indirection (so >can be rolled out into hardware (an X=0 driven multiplexer that takes >Y or Y+F(X)). > >So it is even a case of: AND 2,3(0) is fast and good, while >AND 2,@3 is slow and bad! > > >And yes, trying to minimise the impact of such indirect addressing is >one of the headaches any fast PDP10 maker (such as FPGS cloners) has >got to address, if maximal speed is wanted. > >This means trying to detect the difference of AND 2,@3 (which can >be speeded up) and AND 2,@333 (which can not) and speeding up the >just the later. > >But such detection (a 14bit comparison) slows down basic instruction >decoding, thus we have again the "how often" trouble. Does speeding >AND 2,@3 up by 3-10% (is even dependant on memory speed!) save more >time than slowing all non-@ instructions by about 1%? > >> I've written >> this question three times and it's still very clumsy. > >It is something from deep within your "guy thing" category. Does it help to show it with pseudo-code ? Quoting from Ralph E Gorin, Introduction to DECSYSTEM-20 Assembly Languale programming, page 43 he has a code example in something resembling PASCAL to show the EA calculation : (comments excluded) RUN := TRUE; PC := STARTPC; WHILE RUN DO BEGIN MA := PC; IR := MEM[MA][0:12]; REPEAT <<----- dependency here Y := MEM[MA][18:35]; X := MEM[MA][14:17]; I := MEM[MA][13:13]; IF X = 0 THEN MA := Y <<--- and here ELSE MA := Y + MEM[X][18:35]; UNTIL I:= 0; PC := PC + 1; ExecuteIntruction; END; As you can see from the pseudo-code there are two dependencies in this code that will break a pipeline. >> With hardware interfaces...are you talking about peripherals >> or memory? > >I don't know either what he meant by this. Peripherals usually are not >relevant for CPU instruction speed. Nor for CPU classification. I was thinking of the instruction set and associated complexness of the i/o instructions on the PDP10. The processor has a programmer visible bus structure with the "E" bus where there are channels and a hardware interrupt system. This is all integrated into the CPU design in the PDP10. According to a RISC philosphy it does not belong in the CPU. >> I'll leave the number of instructions to somebody else. I sure >> didn't think we had many; quite a few of them were just there >> to complete the symmetry. They were a lot easier to document. > >It is not about reduced count, but reduced complexity. This >missunderstanding is actually quite widely spread. > If you take a look into one of the emulators you will be struck with how complex this intruction set really is. And it is not the odd string or JFFO that contributes the bulk of it. Having the halfword instructions contributes. The full set of MOVE,JUMP,SKIP,AOS,SOS,AOJ, SOJ,CAM and CAI all as full classes of modifyable instructions contribute. The Test instructions contribute even more. It is the sum of this that disqualifies the PDP10 as RISK. > >Interestingly enough, this is where the emulator writers struggle >> >with performance. > >Or rather the host processor running the emulator program. It goes >into pipeline stalls, even if it has an pipeline friendly architecture. Coding the instruction dispatch and the effective address calculation is a real challenge to get efficient, and both have lots of impact on performance; as they are used at least once every instruction. I have seen the emulator writes attack these from a number of different angles as work progressed. >Software can kill any pipeline. Processors can only offer software >(usually the compiler) the chance of avoiding this trouble. An >emulator for an non-pipelinable processor will allways stall its host >processor. > >As seen in above AND examples. PDP10s can be programmed pipeline >friendly or pipeline stalling. But lots of software prefers I over >X usage, because it is simpler (and surely faster on the simple >microcoded KS-10, possibly the complex microcoded KL-10 had some >"dirty trick" to make it same speed) and more universal (anywhere >in memory, not just addresses 01-17). > >So fast PDP10 makers have to try to optimise at least some common >cases, at cost of chip space, and clock speed. > > >I can only hope that the gcc compiler port uses X over I as often as >possible (or can be changed to in future optimise its code in this >way). > >But then, X is slower than I on old some traditional microcoded >processors, but X is faster than I on FPGAs, and I don't know what >KI-10 did here (could be both ways). > >So this needs to be an target processor model dependant selectable >option to the compiler! That would be -mtune=... option to gcc/pdp10, >for those that want to know. Thanks to Lars for putting that option >in, it will be needed one day in the future. > > >> After I learn more about the three, perhaps we can get to this. >> First, what is the emulator emulating? > >What ever the emulated target architecture does. These emulators emulate a machine not too different from a stock 2060 or 1090. CPUs are emulated in great detail; even to the point of extended addressing. They are geared towards supporting software, and the choices are TOPS20 v4 (or non-extended early v5), full-blown Tops20 v7 and TOPS10 of the latest and fanciest (7.04?). Disk and tape I/O is emulated in great correctness, even if some hardware quirks are left out. Communications via Ethernet or front end processors is done with more economy, and consentrates on mapping onto host machine hardware. >> >Otherwise the large number of regular registers, orthogonal >> >instruction set and flat memory space were RISCy features. >> > >Yes. PDP10 is quite good on many aspects. But the central RISC stuff >is not one of them. That "I" bit is the worst anti-RISC thing one can do. But, again, as you said: it predates CISC. This instruction set must be designed in the fifties (the first PDP1 came out in 1960 as far as I recall.). And the topic drift brings us right back on target for this news group. -- mrr ###### From: Lars Brinkhoff Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 04 Jan 2003 13:35:28 +0100 Organization: nocrew Lines: 9 Sender: lars@junk.nocrew.org Message-ID: <85adihm6sv.fsf@junk.nocrew.org> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> NNTP-Posting-Host: junk.nocrew.org (213.242.147.30) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: fu-berlin.de 1041684011 13237231 213.242.147.30 (16 [140306]) X-Orig-Path: junk.nocrew.org!not-for-mail User-Agent: Gnus/5.0808 (Gnus v5.8.8) Emacs/20.7 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!uni-erlangen.de!fu-berlin.de!uni-berlin.de!junk.nocrew.ORG!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125731 alt.sys.pdp10:12477 Neil Franklin writes: > I can only hope that the gcc compiler port uses X over I as often as > possible Of course. -- Lars Brinkhoff http://lars.nocrew.org/ Linux, GCC, PDP-10, Brinkhoff Consulting http://www.brinkhoff.se/ HTTP programming ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 4 Jan 2003 15:09:58 -0000 Lines: 33 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> <3E1618F4.AB20363B@yahoo.com> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1041692996 28855 80.177.7.220 (4 Jan 2003 15:09:56 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Sat, 4 Jan 2003 15:09:56 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fr.clara.net!heighliner.fr.clara.net!proxad.net!proxad.net!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125738 alt.sys.pdp10:12480 "Peter da Silva" wrote in message news:av5ita$12o5$1@citadel.in.taronga.com... > In article , > Rupert Pigott wrote: > >The very simplicity of the system model utilised by RISCs make > >them so much easier to code for. Compare the IA-32 system > >programming ref to the Alpha equivelent... If that isn't enough to > >convert you to the RISC cause I don't know what is. :) > > IA-32 isn't CISC. It's just layers of cruft laid down on an originally > fairly simple and straightforward 8-bit processor instruction set. In > fact IA-32 has relatively few CISC-style bottlenecks compared to systems Accumulator style architecture... Not really your classic RISC or is it ? :) > like the 68000... but for most of the '80s micros didn't tend to use many > advanced implementation mechanisms like pipelining and the 68000 family > handily beat the pants off the 80[123]86 at equivalent clock speeds. > > It wasn't until the 486 that the x86 really took off. > > >IA-64 takes CISC and makes it even mode complicated. :/ > > IA64 is neither RISC nor CISC, it's VLIW in drag. Overly complex VLIW, > admittedly, but that doesn't make it CISC. True. But it is jolly Complicated compared to a classic RISC. Cheers, Rupert ###### From: Paul Repacholi Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 04 Jan 2003 23:12:11 +0800 Organization: none Lines: 44 Sender: prep@k9 Message-ID: <87smw9osok.fsf@prep.synonet.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> NNTP-Posting-Host: news-01.core.usertools.net Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: nnrp.waia.asn.au 1041694569 18832 202.154.80.9 (4 Jan 2003 15:36:09 GMT) X-Complaints-To: usenet@nnrp.waia.asn.au NNTP-Posting-Date: Sat, 4 Jan 2003 15:36:09 +0000 (UTC) User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 Cache-Post-Path: heidi.pe!unknown@p148.qv1-01.dial.usertools.net X-Cache: nntpcache 2.3.3 (see http://www.nntpcache.org/) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!news1.optus.net.au!optus!news.uwa.edu.au!nntp.waia.asn.au!nnrp.waia.asn.au!127.0.0.1!nobody Xref: chonsp.franklin.ch alt.folklore.computers:125735 alt.sys.pdp10:12478 Neil Franklin writes: > The PDP10 EA calculation hits on both these places. The culprint is > the indirect bit (bit 13, whose number which is very appropriate > :-)). > a) it forces an memory read of the indirect address, whose address > of that address it is only known after the memory access direct > before it. So the pipeline stalls for the full time until memory > answers. A "slow" 100MHz (=10ns per cycle) running from 100ns memory > (that is full memory access cycle) falls to 10% speed if that > happens constantly. A 200MHZ processor is already down to 5% speed > with same memory. > b) even after reading the word coming back, it may have I bits in it > (for double/multiple indirect), so the data can not be immediately > used but needs first looking at. So this is an case of waiting for > decoder. > This only applies to using I. The X indexed stuff is only fast > memory indirect (so no access time trouble), and only max one > indirection (so can be rolled out into hardware (an X=0 driven > multiplexer that takes Y or Y+F(X)). > So it is even a case of: AND 2,3(0) is fast and good, while AND 2,@3 > is slow and bad! There is an added complication to this. The AND 2,@3 instruction can require access to the registers OR memory for the result. Until the EA is calculated, you don't know! Worse, if that is a register store, it can be a source operand, or a EA operand of a following insn... The memory/register aliasing is a severe problem for compilers and I don't think anyone has developed good methods of handeling it, other that banishing it from existance. -- Paul Repacholi 1 Crescent Rd., +61 (08) 9257-1001 Kalamunda. West Australia 6076 comp.os.vms,- The Older, Grumpier Slashdot Raw, Cooked or Well-done, it's all half baked. EPIC, The Architecture of the future, always has been, always will be. ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 4 Jan 2003 19:29:10 GMT Organization: TSS Inc. Lines: 47 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1041708550 68620 10.0.0.43 (4 Jan 2003 19:29:10 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 4 Jan 2003 19:29:10 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.stealth.net!news.stealth.net!logbridge.uoregon.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125740 alt.sys.pdp10:12481 In article , Rupert Pigott wrote: >"Peter da Silva" wrote in message >news:av5ita$12o5$1@citadel.in.taronga.com... >> In article , >> Rupert Pigott wrote: >> >The very simplicity of the system model utilised by RISCs make >> >them so much easier to code for. Compare the IA-32 system >> >programming ref to the Alpha equivelent... If that isn't enough to >> >convert you to the RISC cause I don't know what is. :) >> IA-32 isn't CISC. It's just layers of cruft laid down on an originally >> fairly simple and straightforward 8-bit processor instruction set. In >> fact IA-32 has relatively few CISC-style bottlenecks compared to systems >Accumulator style architecture... Not really your classic RISC or >is it ? :) For the 8080, which is where that architecture originated, they didn't have a huge transistor budget. For systems like that unnecessary register access can slow you down. An accumulator style architecture is probably the RISCiest thing you could do in that case: having the output of the ALU go to any register would make it more complex. To find something more like a classical RISC, you would need to look at something like the 1802. It's still an accumulator architecture, but it has 16 16-bit registers. Now the 1802 has a couple of bottlenecks... it needs to indirect through the P and X registers, and it uses the top of the stack as an accumulator. Probably because they were running out of opcodes, and it was that or cut down on the number of registers. >> IA64 is neither RISC nor CISC, it's VLIW in drag. Overly complex VLIW, >> admittedly, but that doesn't make it CISC. >True. But it is jolly Complicated compared to a classic RISC. Oh, to be sure. It's complicated compared to most classic CISCs as well, by the intuitive definition of complexity. The point is, RISC and CISC aren't the only choices. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### Path: chonsp.franklin.ch!not-for-mail From: Neil Franklin Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 04 Jan 2003 22:51:20 +0100 Organization: My own Private Self Lines: 193 Message-ID: <6uel7sa8iv.fsf@chonsp.franklin.ch> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> NNTP-Posting-Host: chonsp.franklin.ch X-Trace: chonsp.franklin.ch 1041717086 563 10.0.3.2 (4 Jan 2003 21:51:26 GMT) X-Complaints-To: news@chonsp.franklin.ch NNTP-Posting-Date: 4 Jan 2003 21:51:26 GMT X-Newsreader: Gnus v5.7/Emacs 20.4 Xref: chonsp.franklin.ch alt.folklore.computers:125757 alt.sys.pdp10:12484 Morten Reistad writes: > According to Neil Franklin : > > > >jmfbahciv@aol.com writes: > > > >> >The PDP1/6/10 instruction set style had lots of things that > >> >anticipated RISC; > > > >Actually it simply predates the CISC "detour". The extended PDP10 > >(KL-10) made in midst of the CISC days had a few strongly CISCy > >additions (all the string stuff). > > There is still an element of orthogonality that seems to be > missing from other pre-1970 instruction sets I have seen. Yes. The one still sees how simple implementation lead to a very orthogonal system. The entire SET/AND/OR/XOR/... group is a wonderfull case: simply specify an 4bit truth table (in instruction bits 3-6) and have the 36 times 2 data bits being processed pick from them. Only 36 4:1 Multiplexers in there. > >> All right. Now I'm feeling uneducated :-). How does effective > >> address calculation affect the category? Gadzooks! > > >a) it forces an memory read of the indirect address, whose address of > >that address it is only known after the memory access direct before it. > >So the pipeline stalls for the full time until memory answers. A "slow" > >100MHz (=10ns per cycle) running from 100ns memory (that is full memory > >access cycle) falls to 10% speed if that happens constantly. A 200MHZ > >processor is already down to 5% speed with same memory. > > This is where it helped a lot with a fast memory cache. Yes. That would. Here cache can pay already in the small sizes used, even the 1/2kByte cache of the KS-10 would do this. > this was dramatic, even for the rather tiny cache in the 2065. But > consider that the cache is a fix for a problem introduced with the > instruction set. Of course it also accellerates all them other memory reads. > >> With hardware interfaces...are you talking about peripherals > >> or memory? > > > >I don't know either what he meant by this. Peripherals usually are not > >relevant for CPU instruction speed. Nor for CPU classification. > > I was thinking of the instruction set and associated complexness > of the i/o instructions on the PDP10. The actual IO instructions are just 2-address moves. With EA to address memory and the 7bit IO device number plus 2 bits from the opcode to address IO register. Can be implemented in many ways. > The processor has a programmer > visible bus structure with the "E" bus where there are channels > and a hardware interrupt system. > > This is all integrated into the CPU design in the PDP10. According > to a RISC philosphy it does not belong in the CPU. Well some sort of IO bus and interrupting has to be added. And the one in the PDP-10 does not look particularly bad AFAI have looked at it. Separate memory and IO buses are today the in thing (see AMDs Hammer with its direct to SDRAM memory interface and Hyperchannel IO). Any system that puts IO into an sub-range of memory adresses (PDP-1, 680x0, and early RISCs) gets here into "need to check address, if it is memory or IO" problems, nearly equivalent to the EA problems in the PDP-10. > >It is not about reduced count, but reduced complexity. This > >missunderstanding is actually quite widely spread. > > If you take a look into one of the emulators you will be struck with > how complex this intruction set really is. And it is not the odd > string or JFFO that contributes the bulk of it. String can even be simply left to trap&emulate. Just like 386es without 387s did it. Unless someone is into intensive use of them, they will not notice. Of couse traps are massive pipeline breaks, but if it does not happen too often... JFFO is no worse than shift operations. And seldom enough to be run through bit by bit. So it is cheap to do. > Having the halfword > instructions contributes. No worse to implement than SET/AND/OR/XOR/..., just multiplexers (3 layers of 2:1 behind each other). > The full set of MOVE,JUMP,SKIP,AOS,SOS,AOJ, > SOJ,CAM and CAI all as full classes of modifyable instructions > contribute. Also surprisingly simple. One 4-function ALU to compute D1-D2 or D1-0 or D1+1 or D1-1. Then an zero/non-zero detector (36bit wide OR gate). All the 8 equal/greater/less/.. test selection stuff (I assume that is what you mean with "modifyable") is a scant 2 2input ANDs, 1 2-input OR and 1 2-input XOR. > The Test instructions contribute even more. Identical to JUMP,SKIP,AOS,SOS,AOJ,SOJ,CAM and CAI, just an Mux and comparer and ALU. > It is > the sum of this that disqualifies the PDP10 as RISK. Even after all the above mentioned 4 of 8 instruction groups are implemented in my FPGA based clone[1], I only have problems with the I bit in EA computation. And with EA adressing both memory or registers. But those are each enough to fall out of RISC category. [1] http://neil.franklin.ch/Projects/PDP-10/pdp10.vv.html (that is a commented graphical representation of the chip layout so far) > > >Interestingly enough, this is where the emulator writers struggle > >> >with performance. > > > >Or rather the host processor running the emulator program. It goes > >into pipeline stalls, even if it has an pipeline friendly architecture. > > Coding the instruction dispatch and the effective address calculation > is a real challenge to get efficient, and both have lots of impact > on performance; as they are used at least once every instruction. For emulators: yes. Bit group extractions and multi-way jumps are terrible things. That is one reason to clone in hardware, where bits are just standing there, waiting for being evaluated. Getting bits there is what slams on the brakes. And EA is the big offender. > I have seen the emulator writes attack these from a number of different > angles as work progressed. Try hardware. That is what the originals used :-). > >> After I learn more about the three, perhaps we can get to this. > >> First, what is the emulator emulating? > > > >What ever the emulated target architecture does. > > These emulators emulate a machine not too different from a stock > 2060 or 1090. Actually I would assume more an 2040. Because I doubt they emulate the entire cache operation. Simply rely on the host CPU to do this for them. > >> >Otherwise the large number of regular registers, orthogonal > >> >instruction set and flat memory space were RISCy features. > >> > > > >Yes. PDP10 is quite good on many aspects. But the central RISC stuff > >is not one of them. That "I" bit is the worst anti-RISC thing one can do. > > But, again, as you said: it predates CISC. This instruction set > must be designed in the fifties (the first PDP1 came out in 1960 > as far as I recall.). The PDP-1 and PDP-6 are totally different. And PDP-6 is 1963/4. So I would assume design in 1961-2 space. Before then is near impossible as Alan Kotok (one of the lead designers) only went to DEC 1960/1. -- Neil Franklin, neil@franklin.ch.remove http://neil.franklin.ch/ Hacker, Unix Guru, El Eng HTL/BSc, Programmer, Archer, Blacksmith - hardware runs the world, software controls the hardware code generates the software, have you coded today? ###### From: John Sauter Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 04 Jan 2003 17:44:17 -0500 Organization: System Eyes Computer Store Reply-To: J_Sauter@Empire.Net Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> <6uel7sa8iv.fsf@chonsp.franklin.ch> X-Newsreader: Forte Agent 1.93/32.576 English (American) trialware MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Host: hydra141-89.empire.net X-Trace: news.destek.net 1041720259 198.144.141.89 (4 Jan 2003 17:44:19 -0500) Lines: 31 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newshub.northeast.verio.net!verio!news-out.cwix.com!newsfeed.cwix.com!news.destek.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125797 alt.sys.pdp10:12496 >The PDP1/6/10 instruction set style had lots of things that >anticipated RISC; [John Sauter] I missed the beginning of this discussion, but the PDP-1 was not very similar to the PDP-6 and PDP-10. The PDP-1 was an 18-bit machine with 4K words of addressable memory, extended to 16K in the PDP-1D, though addressing it was not simple. >Any system that puts IO into an sub-range of memory adresses (PDP-1, >680x0, and early RISCs) gets here into "need to check address, if it >is memory or IO" problems, nearly equivalent to the EA problems in >the PDP-10. [John Sauter] Here I belive you are thinking of the PDP-11. The PDP-1 had separate IOT instructions, not memory-mapped I/O. > But, again, as you said: it predates CISC. This instruction set > must be designed in the fifties (the first PDP1 came out in 1960 > as far as I recall.). > >The PDP-1 and PDP-6 are totally different. And PDP-6 is 1963/4. So I >would assume design in 1961-2 space. Before then is near impossible as >Alan Kotok (one of the lead designers) only went to DEC 1960/1. I believe the PDP-1 came out in 1959. I don't know when Alan Kotok arrived at DEC, but I was not aware of him until the mid-1960's, when the Stanford AI project bought a PDP-6 to replace our PDP-1. John Sauter (J_Sauter@Empire.Net) ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 5 Jan 2003 00:05:47 GMT Organization: TSS Inc. Lines: 61 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1041725147 77270 10.0.0.43 (5 Jan 2003 00:05:47 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 5 Jan 2003 00:05:47 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!msc1.onvoy!onvoy.com!arclight.uoregon.edu!logbridge.uoregon.edu!newsfeed.stanford.edu!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125783 alt.sys.pdp10:12489 In article , Rupert Pigott wrote: >However, back to the dispute in hand (and I should have said >this in the first place)... In common parlance, including most of >the websites and pulp you see the x86 is referred to as "CISC". >I'm pretty sure I've even seen INTEL refer to it as "CISC" too. Just because they designed tha architecture over the past 25 years doesn't mean they're right about this. There's this tendency for people to simply call anything they think is good "RISC" and anything they don't like "CISC". But there really is a difference between RISC and CISC philosophies. The CISC philosophy was to reduce the semantic distance between the programmer and the computer at the instruction set level. The most significant distinguishing feature of CISC is the attempt to allow the programmer to describe high level operations in the instructions themselves. Thus you have things like memory-memory operations, multiple instruction opcodes (so you can write "ra = rb[rc]" in one instruction), N-level indirection (so you can write "A = b.c^e.f" in one instruction), complex call operations (so you can handle Pascal's multi-frame call tree in one instruction), and so on. The x86 is a load-store architecture with a fairly minimal set of addressing modes, the most complex instructions I'm familiar with are the implied loops... but they're defined in a way that they can be treated as an explicit loop. It's got a complex memory architecture, and that's widely considered a mistake, but that came in wih the 8086 and very little 386 code has to deal with it. So, really, I don't think the x86 has many "CISC" features. Its complexity is in a whole other direction. >> To find something more like a classical RISC, you would need to >> look at something like the 1802. It's still an accumulator >> architecture, but it has 16 16-bit registers. Now the 1802 has a >Yeah, that looked like an interesting wee beastie. The TI9900(?) >was similar wasn't it ? Not as far as I know. It had more in common with the 6502... it didn't have many internal registers, but instead had a rich set of operations on a restricted chunk of memory. On the 6502 that was the zero page, on the 9900 it was pointed to by a register. I guess you could look at it (to bring things full circle) like a PDP-10 where you could relocate where the registers were mapped to memory, instead of having them always start at location 0. As far as that aspect goes, anyway, I don't know much about the 9900 instruction set. The 1802 registers were quite limited... they were basically all index registers... but they were definitely internal. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### Sender: phr2002-nospam@ruckus.brouhaha.com From: Paul Rubin Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> Date: 04 Jan 2003 17:55:05 -0800 Message-ID: <7xwulk4ayu.fsf@ruckus.brouhaha.com> Organization: Nightsong/Fort GNOX Lines: 11 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-Host: 209.66.107.17 X-Trace: 4 Jan 2003 18:31:23 -0800, 209.66.107.17 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news.stealth.net!news.stealth.net!logbridge.uoregon.edu!arclight.uoregon.edu!enews.sgi.com!news.spies.com!209.66.107.17 Xref: chonsp.franklin.ch alt.folklore.computers:125784 alt.sys.pdp10:12490 peter@taronga.com (Peter da Silva) writes: > The x86 is a load-store architecture with a fairly minimal set of > addressing modes, the most complex instructions I'm familiar with are > the implied loops... but they're defined in a way that they can be > treated as an explicit loop. It's got a complex memory architecture, > and that's widely considered a mistake, but that came in wih the 8086 > and very little 386 code has to deal with it. How about the floating point transcendental functions? The floating point register stack is still a monstrous bottleneck, btw. ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 5 Jan 2003 03:35:08 GMT Organization: TSS Inc. Lines: 13 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <7xwulk4ayu.fsf@ruckus.brouhaha.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1041737708 83645 10.0.0.43 (5 Jan 2003 03:35:08 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 5 Jan 2003 03:35:08 GMT X-No-Archive: yes X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!headwall.stanford.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125781 alt.sys.pdp10:12488 In article <7xwulk4ayu.fsf@ruckus.brouhaha.com>, Paul Rubin wrote: >The floating point register stack is still a monstrous bottleneck, btw. Oh yeh, the floating point instruction set on the x86 is pretty awful. But remember... that was originally a separate processor, as most micro FPUs were at the time, and really should be considered separately. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: bfranchuk@jetnet.ab.ca Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 04 Jan 2003 22:05:44 -0700 Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3E17BD28.2010407@jetnet.ab.ca> User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.0.1) Gecko/20021005 X-Accept-Language: en-us, en MIME-Version: 1.0 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> <3E1618F4.AB20363B@yahoo.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 24 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news.stealth.net!news.stealth.net!telocity-west!DIRECTV!sn-xit-03!sn-xit-06!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125761 alt.sys.pdp10:12485 Peter Flass wrote: > Neil Franklin wrote: > >>The PDP10 EA calculation hits on both these places. The culprint is >>the indirect bit (bit 13, whose number which is very appropriate :-)). >> > > So this is why indirect addressing has gone out of fashion. Pity. I > never thought much about it before. I was going to ask when the > pendulum was going to swing back towards ease-of-programming over sheer > effeciency, but since nearly everything is now programmed in HLLs, > ease-of-programming is probably no longer a consideration. It is multi-level indirection that is a problem on any classic machine. The real problem with the older architectures that were developed using core memory were often tightly coupled to the read then write back cycle of memory. Modern memory is allmost always 8 bits wide on a tristated bus. This limits any new development and ideas on anything not 32 or 64 bits wide. Bla-humbug! Ben. ###### Message-ID: <3E182540.D662D452@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 25 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc54 1041762653 12.241.15.59 (Sun, 05 Jan 2003 10:30:53 GMT) NNTP-Posting-Date: Sun, 05 Jan 2003 10:30:53 GMT Date: Sun, 05 Jan 2003 10:30:53 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.stealth.net!news.stealth.net!cyclone1.gnilink.net!wn11feed!wn14feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!rwcrnsc54.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125790 alt.sys.pdp10:12493 Rupert Pigott wrote: > > "Peter da Silva" wrote in message > news:av7cm6$230c$1@citadel.in.taronga.com... > > > > [snp...] [snip...] [snip...] > > > > To find something more like a classical RISC, you would need to > > look at something like the 1802. It's still an accumulator > > architecture, but it has 16 16-bit registers. Now the 1802 has a > > Yeah, that looked like an interesting wee beastie. The TI9900(?) > was similar wasn't it ? > While the TI9900 had sixteen 16-bit registers...the "register file" was taken from memory. The CPU had an internal register that gave the address of the beginning of the register file in main memory. The 9900 could be much faster if on-board RAM was used for allocating the register file, and TI built some models that had this. I do *not* think that the 9900 had the same bottlenecks as the 1802. -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sun, 5 Jan 2003 13:35:42 -0000 Lines: 16 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> <3E1618F4.AB20363B@yahoo.com> <3E1750EA.7021C56E@yahoo.com> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1041773741 15314 80.177.7.220 (5 Jan 2003 13:35:41 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Sun, 5 Jan 2003 13:35:41 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!newsfeed.icl.net!newsfeed.fjserv.net!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125779 alt.sys.pdp10:12487 "Peter Flass" wrote in message news:3E1750EA.7021C56E@yahoo.com... > I was thinking of cases such as when (IIRC) it's necessary to have an > instruction between a test and the corresponding conditional branch to > make the thing work properly. Maybe it's only familiarity, but Heh, widely acknowledged as a dumb idea and quite often rendered irrelevent by later Superscalar implementations.... You can always wang NOPs into those slots if your brain hurts too much... However you can't really do much about chronic register shortage. :) Cheers, Rupert ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: <28kg1vov264a0tb38n7au2u8gotlresc1i@4ax.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 34 Date: Sun, 05 Jan 2003 15:39:09 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news3.calgary.shaw.ca 1041781149 24.71.223.147 (Sun, 05 Jan 2003 08:39:09 MST) NNTP-Posting-Date: Sun, 05 Jan 2003 08:39:09 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.stealth.net!news.stealth.net!prodigy.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news3.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125789 alt.sys.pdp10:12492 On Sat, 4 Jan 2003 13:09:22 +0100, Morten Reistad wrote: >According to Neil Franklin : >>dropped the prime stuff, added pdp10 >> >>jmfbahciv@aol.com writes: >> >>> In article <9dn1va.bu7.ln@via.reistad.priv.no>, >>> Morten Reistad wrote: >>> >According to : >>> >>Yup. We do have different definitions. I had forgotten that my >>> >>idea of PDP-10 architecture having RISC-flavor was not politically >>> >>correct. :-) >>> > >>> >The PDP1/6/10 instruction set style had lots of things that >>> >anticipated RISC; PDP-1/9/15 were 18 bit machines; PDP-6/10 were 36 bit machines. >But, again, as you said: it predates CISC. This instruction set >must be designed in the fifties (the first PDP1 came out in 1960 >as far as I recall.). Different instruction sets for PDP-1 and PDP-6. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### NNTP-Posting-Date: Sun, 05 Jan 2003 13:23:55 -0600 From: "Timothy Stark" Newsgroups: alt.folklore.computers,alt.sys.pdp10 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E182540.D662D452@ev1.net> Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sun, 5 Jan 2003 14:23:55 -0500 X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 Message-ID: Lines: 22 NNTP-Posting-Host: 66.92.145.31 X-Trace: sv3-CbivqTY2mEqc2zit5dWXDLZ59VjWhcsdjBto+h8DOO92y8Xf/w2Mkhi9iTimtEb0vdTdrQ2Rz01Eo8g!43RmkkY+wEcgs2f2X9q7IkojxNw6Tb+E1NjVdTU7S6aj/gT6CM37tpRZ+DBzVIT6FXaq2YZVVhkM!zZtYSrzfQTC6ROsM X-Complaints-To: abuse@speakeasy.net X-DMCA-Complaints-To: abuse@speakeasy.net X-Abuse-and-DMCA-Info: Please be sure to forward a copy of ALL headers X-Abuse-and-DMCA-Info: Otherwise we will be unable to process your complaint properly X-Postfilter: 1.1 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.stealth.net!news.stealth.net!news-xfer.cox.net!cox.net!feed01.bs1!nntp.giganews.com!nntp3.aus1.giganews.com!nntp.speakeasy.net!news.speakeasy.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:125786 alt.sys.pdp10:12491 "Charles Richmond" wrote in message news:3E182540.D662D452@ev1.net... > While the TI9900 had sixteen 16-bit registers...the "register file" > was taken from memory. The CPU had an internal register that gave > the address of the beginning of the register file in main memory. > The 9900 could be much faster if on-board RAM was used for allocating > the register file, and TI built some models that had this. I do > *not* think that the 9900 had the same bottlenecks as the 1802. Yes, I know TMS9900 assembly language so well because I wrote assembly programs on my TI-99/4A home computer. Yes, it have sixteen 16-bit registers that are stored on anywhere in main storage by using a WP pointer. When it is turned on, TMS9900 will load WP and PC from first two 16-bit words of main storage then start execution. TMS9995 have 256 bytes on-chip RAM. I already implemented TMS9900 emulation into my ts10 emulator but I have figure out how to implement SDL support routines for 9918A video controller. Tim Stark ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 09 Jan 03 09:43:37 GMT Organization: UltraNet Communications, Inc. Lines: 23 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <7xwulk4ayu.fsf@ruckus.brouhaha.com> X-Trace: UmFuZG9tSVZo0BcqrFgo/FEEhBIa/KIDE/oIZr+q7xoDro6ocy/L7AFp/e3cTei1 X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 9 Jan 2003 10:18:17 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!news.webusenet.com!cyclone1.gnilink.net!wn11feed!worldnet.att.net!199.184.165.233!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-69 Xref: chonsp.franklin.ch alt.folklore.computers:126035 alt.sys.pdp10:12500 In article , lawrence.jones@eds.com wrote: >Paul Rubin wrote: >> >> The floating point register stack is still a monstrous bottleneck, btw. > >In what sense? It isn't really a stack any more, it's just a peculiar >way of addressing some of the many internal floating-point registers. I used to use double indexing to pick up a data description so I knew how to format the datum I was dumping to disk. I never played with indirection very much but the monitor guys did a lot. Just by setting a single bit you could move the contents of a word or the contents that word pointed to. So the list could be a mix of data and addresses; and the list could be modified on the fly. I get shivers down my spine just thinking about the power of that technique. /BAH Subtract a hundred and four for e-mail. ###### Sender: Larry Jones From: lawrence.jones@eds.com Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Newsgroups: alt.folklore.computers,alt.sys.pdp10 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <7xwulk4ayu.fsf@ruckus.brouhaha.com> User-Agent: tin/1.4.5-20010409 ("One More Nightmare") (UNIX) (BSD/OS/4.0.1 (i386)) Message-ID: Lines: 11 Date: Wed, 08 Jan 2003 23:19:54 GMT NNTP-Posting-Host: 65.27.189.87 X-Complaints-To: abuse@rr.com X-Trace: twister.neo.rr.com 1042067994 65.27.189.87 (Wed, 08 Jan 2003 18:19:54 EST) NNTP-Posting-Date: Wed, 08 Jan 2003 18:19:54 EST Organization: Road Runner High Speed Online -- Northeast Ohio Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news.stealth.net!news.stealth.net!news-east.rr.com!news-server.columbus.rr.com!cyclone3.kc.rr.com!news3.kc.rr.com!twister.neo.rr.com.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:126075 alt.sys.pdp10:12501 Paul Rubin wrote: > > The floating point register stack is still a monstrous bottleneck, btw. In what sense? It isn't really a stack any more, it's just a peculiar way of addressing some of the many internal floating-point registers. -Larry Jones Hey Doc, for 10 bucks I'll make sure you see those kids in the waiting room again real soon! -- Calvin ###### Sender: Larry Jones From: lawrence.jones@eds.com Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Newsgroups: alt.folklore.computers,alt.sys.pdp10 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> <3E1618F4.AB20363B@yahoo.com> User-Agent: tin/1.4.5-20010409 ("One More Nightmare") (UNIX) (BSD/OS/4.0.1 (i386)) Message-ID: Lines: 14 Date: Wed, 08 Jan 2003 23:19:54 GMT NNTP-Posting-Host: 65.27.189.87 X-Complaints-To: abuse@rr.com X-Trace: twister.neo.rr.com 1042067994 65.27.189.87 (Wed, 08 Jan 2003 18:19:54 EST) NNTP-Posting-Date: Wed, 08 Jan 2003 18:19:54 EST Organization: Road Runner High Speed Online -- Northeast Ohio Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!solnet.ch!solnet.ch!newsfeed.freenet.de!news.tu-darmstadt.de!news.belwue.de!news.stealth.net!news.stealth.net!news-east.rr.com!news-server.columbus.rr.com!cyclone3.kc.rr.com!news3.kc.rr.com!twister.neo.rr.com.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:126080 alt.sys.pdp10:12502 Peter Flass wrote: > > So this is why indirect addressing has gone out of fashion. Pity. I > never thought much about it before. With modern superscalar architectures and out-of-order execution of instructions, I don't think it makes any difference any more. Who cares how long it takes to compute the effective address for one instruction as long as you've got other instructions you can execute in the meantime? -Larry Jones Bad news, Mom. I promised my soul to the Devil this afternoon. -- Calvin ###### From: bfranchuk@jetnet.ab.ca Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 08 Jan 2003 17:37:21 -0700 Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3E1CC441.2070001@jetnet.ab.ca> User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.0.1) Gecko/20021005 X-Accept-Language: en-us, en MIME-Version: 1.0 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> <3E1618F4.AB20363B@yahoo.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 13 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!c03.atl99!news.webusenet.com!telocity-west!DIRECTV!sn-xit-03!sn-xit-01!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:126024 alt.sys.pdp10:12499 lawrence.jones@eds.com wrote: >>So this is why indirect addressing has gone out of fashion. Pity. I >>never thought much about it before. > With modern superscalar architectures and out-of-order execution of > instructions, I don't think it makes any difference any more. Who cares > how long it takes to compute the effective address for one instruction > as long as you've got other instructions you can execute in the > meantime? Yes but kills the whole idea of the RISC's, you have to access REAL memory. :) ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> <3E1618F4.AB20363B@yahoo.com> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 26 Date: Thu, 09 Jan 2003 02:36:21 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news3.calgary.shaw.ca 1042079781 24.71.223.147 (Wed, 08 Jan 2003 19:36:21 MST) NNTP-Posting-Date: Wed, 08 Jan 2003 19:36:21 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news.stealth.net!news.stealth.net!news.astraweb.com!news-small.astraweb.com!prodigy.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news3.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:126088 alt.sys.pdp10:12503 On Wed, 08 Jan 2003 23:19:54 GMT, lawrence.jones@eds.com wrote: >Peter Flass wrote: >> >> So this is why indirect addressing has gone out of fashion. Pity. I >> never thought much about it before. > >With modern superscalar architectures and out-of-order execution of >instructions, I don't think it makes any difference any more. Who cares >how long it takes to compute the effective address for one instruction >as long as you've got other instructions you can execute in the >meantime? That would be a major bottleneck -- you can't execute any following instructions until you've fetched and decoded that instruction to see what data it affects. OOO execution relies on the CPU knowing the data dependencies between instructions. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: <3uas1vc6t4r8h8jda9a3qkh0ggua85bnq3@4ax.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <9dn1va.bu7.ln@via.reistad.priv.no> <6uvg16uexm.fsf_-_@chonsp.franklin.ch> <3E1618F4.AB20363B@yahoo.com> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 29 Date: Fri, 10 Jan 2003 02:15:25 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news3.calgary.shaw.ca 1042164925 24.71.223.147 (Thu, 09 Jan 2003 19:15:25 MST) NNTP-Posting-Date: Thu, 09 Jan 2003 19:15:25 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news.stealth.net!news.stealth.net!news.astraweb.com!news-small.astraweb.com!news-xfer.cox.net!cox.net!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news3.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:126166 alt.sys.pdp10:12504 On Thu, 09 Jan 2003 19:19:55 GMT, lawrence.jones@eds.com wrote: >Brian Inglis wrote: >> >> That would be a major bottleneck -- you can't execute any >> following instructions until you've fetched and decoded that >> instruction to see what data it affects. OOO execution relies on >> the CPU knowing the data dependencies between instructions. > >True, but that doesn't necessarily imply a complete stall. Complicated >addressing modes usually refer to memory and modern architectures don't >map registers into the memory address space, so you can still execute >instructions that only address registers. Also, you usually know from >the instruction whether memory is being read or written, so you can >still execute other instructions that access memory the same way. But memory transfers are to or from registers in modern architectures, so the registers used and all other memory transfers would have to be blocked until an indirect instruction completed, as any address could be affected. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 11 Jan 2003 21:43:38 GMT Organization: TSS Inc. Lines: 17 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3uas1vc6t4r8h8jda9a3qkh0ggua85bnq3@4ax.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1042321418 82566 10.0.0.43 (11 Jan 2003 21:43:38 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 11 Jan 2003 21:43:38 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:126279 alt.sys.pdp10:12508 In article <3uas1vc6t4r8h8jda9a3qkh0ggua85bnq3@4ax.com>, Brian Inglis wrote: >But memory transfers are to or from registers in modern >architectures, so the registers used and all other memory >transfers would have to be blocked until an indirect instruction >completed, as any address could be affected. If it's a read, you'd have to stall writes, but other reads could go through. If it's a write, you can let reads go through, and possibly other reversible operations, and undo them if you guess wrong. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: "John R. Strohm" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 19 Feb 2003 18:50:20 -0600 Organization: Airnews.net! at Internet America Lines: 34 Message-ID: <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> X-Orig-Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> Abuse-Reports-To: abuse at airmail.net to report improper postings NNTP-Proxy-Relay: library1-aux.airnews.net NNTP-Posting-Time: Wed Feb 19 19:22:59 2003 NNTP-Posting-Host: ![`4=1k-YAjoZTf (Encoded at Airnews!) X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.50.4522.1200 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news-FFM2.ecrc.net!newsfeed.icl.net!newsfeed.fjserv.net!news.maxwell.syr.edu!news.airnews.net!cabal12.airnews.net!usenet Xref: chonsp.franklin.ch alt.folklore.computers:129409 alt.sys.pdp10:12528 "Peter da Silva" wrote in message news:av7ssr$2bem$1@citadel.in.taronga.com... > The x86 is a load-store architecture with a fairly minimal set of > addressing modes, the most complex instructions I'm familiar with are > the implied loops... but they're defined in a way that they can be > treated as an explicit loop. It's got a complex memory architecture, > and that's widely considered a mistake, but that came in wih the 8086 > and very little 386 code has to deal with it. > > So, really, I don't think the x86 has many "CISC" features. Its complexity > is in a whole other direction. A coworker of mine a few years ago told me some things about the x86 that explained a lot of the cruft in the architecture and instruction set. At the time that Intel did the 8086 (and 8088), they didn't really have any microprocessor designers on staff. What they had was the original 4004 (4-bit) core, but the guy who designed that part was LONG gone. The 8080 was made by laying down two 4004 cores, and taping up a few more registers and interconnects. The 8086 was made from four 4004 cores, and 8080 stuff, and some more registers and cruft. If you look CAREFULLY at the 8086 architecture, you can see where they pasted the bags onto the 8080 design. I *THINK* he told me they were still in this mode up through the 80286. It wasn't until they started the 80386 that they got some real designers and started over. He owned a CAD firm at the time. His niche was automated test vector generation. You supply him with register transfer logic equations, and his CAD systems would spit back test vectors to test your device. The problem is that they had to have the equations, and Intel didn't have any. ###### From: Peter da Silva Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 19 Feb 2003 20:13:24 -0600 Organization: ABB Inc., Network Management Division Lines: 22 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> NNTP-Posting-Host: switchblade.dhcp.abb.abbnm.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Trace: jeeves.eng.abbnm.com 1045707311 45333 10.127.143.26 (20 Feb 2003 02:15:11 GMT) X-Complaints-To: usenet@jeeves.eng.abbnm.com NNTP-Posting-Date: Thu, 20 Feb 2003 02:15:11 +0000 (UTC) User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.1) Gecko/20020826 X-Accept-Language: en-us, en, ja Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!news.gtei.net!bloom-beacon.mit.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!newsboy.eng.abbnm.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129453 alt.sys.pdp10:12530 John R. Strohm wrote: > The 8080 > was made by laying down two 4004 cores, and taping up a few more registers > and interconnects. I find that hard to believe, given the vast differences between the 4004 and the 8080. It's amazing how they managed to paper over the 4-bit BCD data memory and binary 8-bit program memory and RAM status characters. And where did the rest of the original 16 registers go? > If you look CAREFULLY at the 8086 > architecture, you can see where they pasted the bags onto the 8080 design. You mean like reading the release notes where Intel explained how to assemble 8080 code for the 8086? That was a deeply hidden secret that nobody but programmers and engineers and managers and other people who were likely to read the release notes could unravel. Assembly code compatibility with the 8080 was a design goal. It says nothing about the implementation. ###### From: "John R. Strohm" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 19 Feb 2003 22:53:33 -0600 Organization: Airnews.net! at Internet America Lines: 35 Message-ID: <4C975ED26C1160AB.CF3B0CD5751CDC41.6458CD3453A4A393@lp.airnews.net> X-Orig-Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> Abuse-Reports-To: abuse at airmail.net to report improper postings NNTP-Proxy-Relay: library1-aux.airnews.net NNTP-Posting-Time: Wed Feb 19 23:21:55 2003 NNTP-Posting-Host: !ZM:g1k-X)gNaSE (Encoded at Airnews!) X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.50.4522.1200 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.airnews.net!cabal12.airnews.net!usenet Xref: chonsp.franklin.ch alt.folklore.computers:129407 alt.sys.pdp10:12527 It may just have been the 4-bit ALU core of the 4004 then. Like I said, I got this story from a reputable source. Intel called his company to get test vectors generated for (I think) the 80286. He had to tell them that his company couldn't help them, because they didn't have the equations for the part. "Peter da Silva" wrote in message news:b31dnf$1c8l$1@jeeves.eng.abbnm.com... > John R. Strohm wrote: > > The 8080 > > was made by laying down two 4004 cores, and taping up a few more registers > > and interconnects. > > I find that hard to believe, given the vast differences between > the 4004 and the 8080. It's amazing how they managed to paper over > the 4-bit BCD data memory and binary 8-bit program memory and RAM > status characters. And where did the rest of the original 16 > registers go? > > > If you look CAREFULLY at the 8086 > > architecture, you can see where they pasted the bags onto the 8080 design. > > You mean like reading the release notes where Intel explained how > to assemble 8080 code for the 8086? That was a deeply hidden > secret that nobody but programmers and engineers and managers and > other people who were likely to read the release notes could unravel. > > Assembly code compatibility with the 8080 was a design goal. It > says nothing about the implementation. > ###### From: Peter da Silva Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 20 Feb 2003 11:00:15 -0600 Organization: ABB Inc., Network Management Division Lines: 11 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <4C975ED26C1160AB.CF3B0CD5751CDC41.6458CD3453A4A393@lp.airnews.net> NNTP-Posting-Host: switchblade.dhcp.abb.abbnm.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Trace: jeeves.eng.abbnm.com 1045760522 62582 10.127.143.26 (20 Feb 2003 17:02:02 GMT) X-Complaints-To: usenet@jeeves.eng.abbnm.com NNTP-Posting-Date: Thu, 20 Feb 2003 17:02:02 +0000 (UTC) User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.1) Gecko/20020826 X-Accept-Language: en-us, en, ja Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.stanford.edu!news-proxy.abbnm.com!newsboy.eng.abbnm.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129449 alt.sys.pdp10:12529 John R. Strohm wrote: > Like I said, I got this story from a reputable source. Intel called his > company to get test vectors generated for (I think) the 80286. He had to > tell them that his company couldn't help them, because they didn't have the > equations for the part. That just means they didn't have the equations for the part. Not that there was actually any silicon left over from the 4004. I mean, the 4004 was P-channel MOS... would the same design even work in a different technology? ###### Message-ID: <3E550E1C.995BB979@iquest.net> From: John Dyson Reply-To: dyson@iquest.net X-Mailer: Mozilla 4.8 [en] (X11; U; Linux 2.4.2 i386) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <4C975ED26C1160AB.CF3B0CD5751CDC41.6458CD3453A4A393@lp.airnews.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 17 Date: Thu, 20 Feb 2003 12:19:24 -0500 NNTP-Posting-Host: 206.246.160.137 X-Complaints-To: news-abuse@iquest.net X-Trace: news.iquest.net 1045761003 206.246.160.137 (Thu, 20 Feb 2003 12:10:03 EST) NNTP-Posting-Date: Thu, 20 Feb 2003 12:10:03 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news.tele.dk!news.tele.dk!small.news.tele.dk!newsfeed.icl.net!newsfeed.fjserv.net!newsfeed1.cidera.com!Cidera!news.iquest.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129404 alt.sys.pdp10:12526 Peter da Silva wrote: > > John R. Strohm wrote: > > Like I said, I got this story from a reputable source. Intel called his > > company to get test vectors generated for (I think) the 80286. He had to > > tell them that his company couldn't help them, because they didn't have the > > equations for the part. > > That just means they didn't have the equations for the part. Not > that there was actually any silicon left over from the 4004. I > mean, the 4004 was P-channel MOS... would the same design even > work in a different technology? > Just imagine the retrofit for the segmentation, let alone the very differing rules. John ###### Path: chonsp.franklin.ch!not-for-mail From: Neil Franklin Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 20 Feb 2003 21:22:25 +0100 Organization: My own Private Self Lines: 71 Message-ID: <6un0kqaeim.fsf@chonsp.franklin.ch> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> NNTP-Posting-Host: chonsp.franklin.ch X-Trace: chonsp.franklin.ch 1045772548 437 10.0.3.2 (20 Feb 2003 20:22:28 GMT) X-Complaints-To: news@chonsp.franklin.ch NNTP-Posting-Date: 20 Feb 2003 20:22:28 GMT X-Newsreader: Gnus v5.7/Emacs 20.4 Xref: chonsp.franklin.ch alt.folklore.computers:129492 alt.sys.pdp10:12531 "John R. Strohm" writes: > "Peter da Silva" wrote in message > news:av7ssr$2bem$1@citadel.in.taronga.com... > > The x86 is a load-store architecture with a fairly minimal set of > > addressing modes, the most complex instructions I'm familiar with are > > the implied loops... but they're defined in a way that they can be > > > > So, really, I don't think the x86 has many "CISC" features. Its complexity > > is in a whole other direction. Yup. Mainly in totally non orthogonal instruction set. And in 3 different MMU modes (actually 2 layered MMUs, one with 2 modes). > At the time that Intel did the 8086 (and 8088), they didn't really have any > microprocessor designers on staff. What they had was the original 4004 > (4-bit) core, but the guy who designed that part was LONG gone. The 8080 > was made by laying down two 4004 cores, and taping up a few more registers > and interconnects. That story sounds bogus. For one it ignores the 8008. And that was an 8bit made while the 4004 designers were still with Intel. And that looks like 1 4004s bolted together, appart from 16*4bit->8*8bit registers. As far as one of the designers said in an interview they left around the introduction of 8080, because Intel did not want to continue with another revision. Z80 was what they then made. So 8085 would be the first "designerless" Intel chip. And yes, 8080, Z80 and 8085 has an "half carry" flag. But that is for supporting the DAA instruction, not because of the ALU being 4bit derived. > The 8086 was made from four 4004 cores, and 8080 stuff, > and some more registers and cruft. As it was made parallel to 8085, I would assume it to be 2 8080s put together. Actually it was specified to be assembly compatible with 8080. > If you look CAREFULLY at the 8086 > architecture, you can see where they pasted the bags onto the 8080 design. You mean the MMU? That was a paste-on on any pure 16bit CPU that wanted more than 64k addresses, without becoming an 16/32bit like the 68000. Remeber PDP11? Same story, but better executed. > I *THINK* he told me they were still in this mode up through the 80286. It > wasn't until they started the 80386 that they got some real designers and > started over. Even 286 was already after the failled 80432 project. That is where its MMU came from. So they had designers by then. > generation. You supply him with register transfer logic equations, and his > CAD systems would spit back test vectors to test your device. The problem > is that they had to have the equations, and Intel didn't have any. Because they still designed by hand in the 1970s? -- Neil Franklin, neil@franklin.ch.remove http://neil.franklin.ch/ Hacker, Unix Guru, El Eng HTL/BSc, Programmer, Archer, Blacksmith - hardware runs the world, software controls the hardware code generates the software, have you coded today? ###### From: Joe Pfeiffer Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 19 Feb 2003 21:07:39 -0700 Organization: New Mexico State University Lines: 24 Message-ID: <1by94bd27o.fsf@cs.nmsu.edu> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> NNTP-Posting-Host: viper.cs.nmsu.edu Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: bubba.NMSU.Edu 1045714058 1801 128.123.64.113 (20 Feb 2003 04:07:38 GMT) X-Complaints-To: usenet@bubba.NMSU.Edu NNTP-Posting-Date: 20 Feb 2003 04:07:38 GMT User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!msc1.onvoy!onvoy.com!hardy.tc.umn.edu!nunki.unm.edu!news.NMSU.Edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129518 alt.sys.pdp10:12533 "John R. Strohm" writes: > > At the time that Intel did the 8086 (and 8088), they didn't really have any > microprocessor designers on staff. What they had was the original 4004 > (4-bit) core, but the guy who designed that part was LONG gone. The 8080 > was made by laying down two 4004 cores, and taping up a few more registers > and interconnects. The 8086 was made from four 4004 cores, and 8080 stuff, > and some more registers and cruft. If you look CAREFULLY at the 8086 > architecture, you can see where they pasted the bags onto the 8080 design. I really find this story hard to believe... architecturally, yes, everything up through and including IA-32 (which means 386 through Pentium 4) is one layer of cruft on top of another. The idea that it was literally done by wiring the cores together.... that's just strectching credulity too far. Especially since the 4004 was a multi-chip calculator, and the 8008 was a reimplementation of Datapoint's (though their name wasn't Datapoint yet; I forget what it was) bit-serial terminal controller. -- Joseph J. Pfeiffer, Jr., Ph.D. Phone -- (505) 646-1605 Department of Computer Science FAX -- (505) 646-1002 New Mexico State University http://www.cs.nmsu.edu/~pfeiffer Southwestern NM Regional Science and Engr Fair: http://www.nmsu.edu/~scifair ###### From: aek@spies.com (Al Kossow) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 21 Feb 2003 16:05:04 +0000 (UTC) Organization: Spies In the Wire Lines: 9 Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> NNTP-Posting-Host: spies.com X-Trace: 21 Feb 2003 08:46:39 -0800, spies.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!nntp-relay.ihug.net!ihug.co.nz!logbridge.uoregon.edu!arclight.uoregon.edu!enews.sgi.com!news.spies.com!unknown!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129499 alt.sys.pdp10:12532 From article <1by94bd27o.fsf@cs.nmsu.edu>, by Joe Pfeiffer : > "John R. Strohm" writes: >> > Especially since the 4004 was a multi-chip calculator, and the 8008 > was a reimplementation of Datapoint's (though their name wasn't > Datapoint yet; I forget what it was) bit-serial terminal controller. Computer Terminals Corp (CTC) ###### Message-ID: <3E56AB27.162B9193@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 47 Date: Fri, 21 Feb 2003 23:21:25 GMT NNTP-Posting-Host: 12.90.171.208 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc04-news.ops.worldnet.att.net 1045869685 12.90.171.208 (Fri, 21 Feb 2003 23:21:25 GMT) NNTP-Posting-Date: Fri, 21 Feb 2003 23:21:25 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.mathworks.com!wn13feed!worldnet.att.net!bgtnsc04-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129638 alt.sys.pdp10:12535 bfranchuk@jetnet.ab.ca wrote: > Joe Pfeiffer wrote: > > > I really find this story hard to believe... architecturally, > > yes, everything up through and including IA-32 (which means 386 > > through Pentium 4) is one layer of cruft on top of another. The > > idea that it was literally done by wiring the cores together.... > > that's just strectching credulity too far. > > I suspect is true in the sense that they were using proven > ideas rather starting fresh every time. If a 4004 alu logic > design works why change the logic for a 8bit cpu. Mask design > was done mostly by hand and logic equations could be in the form > of a large schematic. > > Ben. > > PS while the 16 bit cpu intel made was crap is was only ment for > about a 3-4 year lifespan. The real cpu that intel was working on > a harware object oriented 32? bit cpu, that never worked for years > and by that time the 386 sales took off big time. I liked the 8086 architecture. You could do elegant things with it, not including real virtual memory. I didn't like the Micro$loth code that ran on it, nor the misbegotten ISA bus signal polarities, nor the foolish waste of critical fault interrupts. In fact I think the x86 world would be better off using segments in an intelligent manner. The thing I disapproved was the removal of the XTHL 8080 instruction equivalent, which made it impossible to build truly general re-entrant routines, and the lack of an illegal instruction trap. The 286 was an abortion, neither fish nor fowl. From then on it is a song of downward compatibility, and sales history have proven them right, not elegant. But the realization of compatibility is a major feat. An 8086 in 1978 was orders of magnitude more capable than a PDP8 or SDS 910 of 1964, and somewhat cheaper. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### From: "John R. Strohm" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 21 Feb 2003 18:06:52 -0600 Organization: Airnews.net! at Internet America Lines: 36 Message-ID: <93F83D24B0A3E18E.B3103076363F175A.58C7F3DFE7653EAB@lp.airnews.net> X-Orig-Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> Abuse-Reports-To: abuse at airmail.net to report improper postings NNTP-Proxy-Relay: library1-aux.airnews.net NNTP-Posting-Time: Fri Feb 21 18:09:24 2003 NNTP-Posting-Host: !YT;p1k-Y+dOt)^ (Encoded at Airnews!) X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.50.4522.1200 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.belwue.de!feed.news.nacamar.de!news.maxwell.syr.edu!news.airnews.net!cabal12.airnews.net!usenet Xref: chonsp.franklin.ch alt.folklore.computers:129602 alt.sys.pdp10:12534 "CBFalconer" wrote in message news:3E56AB27.162B9193@yahoo.com... > I liked the 8086 architecture. You could do elegant things with > it, not including real virtual memory. I didn't like the > Micro$loth code that ran on it, nor the misbegotten ISA bus signal > polarities, nor the foolish waste of critical fault interrupts. > > In fact I think the x86 world would be better off using segments > in an intelligent manner. The thing I disapproved was the removal > of the XTHL 8080 instruction equivalent, which made it impossible > to build truly general re-entrant routines, and the lack of an > illegal instruction trap. > > The 286 was an abortion, neither fish nor fowl. From then on it > is a song of downward compatibility, and sales history have proven > them right, not elegant. But the realization of compatibility is > a major feat. > > An 8086 in 1978 was orders of magnitude more capable than a PDP8 > or SDS 910 of 1964, and somewhat cheaper. 1978??? I think you're a few years early. And, in any event, by 1981, Motorola had the 68000 in volume production, and Zilog had the Z8000. The ONLY reason that the 8086 went anywhere is that IBM selected the 8088 for the IBM PC. This was a horrible mistake, but it was driven by their desire for an 8-bit data bus, so they could fall back to a Z-80 design if the 8088 didn't make it. This was a bogus plan: at the same time that IBM Boca Raton, their Siberia division, was designing the PC around the 8088, another division was selling the SS9000 laboratory computer, which was built around the 68000. ###### From: "Glen Herrmannsfeldt" Newsgroups: alt.folklore.computers,alt.sys.pdp10 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <93F83D24B0A3E18E.B3103076363F175A.58C7F3DFE7653EAB@lp.airnews.net> Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Lines: 54 X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.00.2919.6700 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6700 Message-ID: NNTP-Posting-Host: 12.207.204.17 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc54 1045880520 12.207.204.17 (Sat, 22 Feb 2003 02:22:00 GMT) NNTP-Posting-Date: Sat, 22 Feb 2003 02:22:00 GMT Organization: AT&T Broadband Date: Sat, 22 Feb 2003 02:22:00 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.telusplanet.net!newsfeed.bc.tac.net!cyclone.bc.net!logbridge.uoregon.edu!nntp-server.caltech.edu!attla2!ip.att.net!attbi_feed3!attbi.com!rwcrnsc54.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129698 alt.sys.pdp10:12542 "John R. Strohm" wrote in message news:93F83D24B0A3E18E.B3103076363F175A.58C7F3DFE7653EAB@lp.airnews.net... > "CBFalconer" wrote in message > news:3E56AB27.162B9193@yahoo.com... > > I liked the 8086 architecture. You could do elegant things with > > it, not including real virtual memory. I didn't like the > > Micro$loth code that ran on it, nor the misbegotten ISA bus signal > > polarities, nor the foolish waste of critical fault interrupts. > > > > In fact I think the x86 world would be better off using segments > > in an intelligent manner. The thing I disapproved was the removal > > of the XTHL 8080 instruction equivalent, which made it impossible > > to build truly general re-entrant routines, and the lack of an > > illegal instruction trap. > > > > The 286 was an abortion, neither fish nor fowl. From then on it > > is a song of downward compatibility, and sales history have proven > > them right, not elegant. But the realization of compatibility is > > a major feat. I always liked the 286. It fixed a number of faults with the 8086, allowed segment swapping virtual storage with relatively simple hardware. I ran OS/2 1.0 and 1.2 on one, doing much more than DOS could do at the time. > > An 8086 in 1978 was orders of magnitude more capable than a PDP8 > > or SDS 910 of 1964, and somewhat cheaper. > > 1978??? I think you're a few years early. I am pretty sure they were available in 1979, and they probably existed in 1978, at least inside Intel. > And, in any event, by 1981, Motorola had the 68000 in volume production, and > Zilog had the Z8000. > > The ONLY reason that the 8086 went anywhere is that IBM selected the 8088 > for the IBM PC. This was a horrible mistake, but it was driven by their > desire for an 8-bit data bus, so they could fall back to a Z-80 design if > the 8088 didn't make it. This was a bogus plan: at the same time that IBM > Boca Raton, their Siberia division, was designing the PC around the 8088, > another division was selling the SS9000 laboratory computer, which was built > around the 68000. Looking back, I agree. But IBM wanted to get into the lower priced market, and 68000 was pretty expensive at the time. The original PC had 64K and a cassette tape interface for storage to keep prices down. The Apple Lisa came out about the same time, but much more expensive. -- glen ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 22 Feb 2003 13:38:27 GMT Organization: TSS Inc. Lines: 21 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1045921107 17505 10.0.0.43 (22 Feb 2003 13:38:27 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 22 Feb 2003 13:38:27 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!feed2.news.rcn.net!rcn!logbridge.uoregon.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129841 alt.sys.pdp10:12548 In article <3E56AB27.162B9193@yahoo.com>, CBFalconer wrote: >In fact I think the x86 world would be better off using segments >in an intelligent manner. You can't. 64k is too small. Even the 256k segments Multics had available over 10 years earlier turned out to be too small. I was writing in PL/M, which used the segments as intelligently as possible... they were first class data types. And 64k was too small. >An 8086 in 1978 was orders of magnitude more capable than a PDP8 >or SDS 910 of 1964, and somewhat cheaper. An 8080 was more capable than a PDP-8, for Turing's sake. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### Message-ID: <3E5794E4.6060303@bellatlantic.net> From: bob smith User-Agent: Mozilla/5.0 (Windows; U; WinNT4.0; en-US; rv:1.0.1) Gecko/20020823 Netscape/7.0 X-Accept-Language: en-us, en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Lines: 27 Date: Sat, 22 Feb 2003 15:08:51 GMT NNTP-Posting-Host: 138.88.112.150 X-Complaints-To: abuse@verizon.net X-Trace: nwrddc01.gnilink.net 1045926531 138.88.112.150 (Sat, 22 Feb 2003 10:08:51 EST) NNTP-Posting-Date: Sat, 22 Feb 2003 10:08:51 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!news.webusenet.com!cyclone1.gnilink.net!spamkiller.gnilink.net!nwrddc01.gnilink.net.POSTED!53ab2750!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129776 alt.sys.pdp10:12544 Peter da Silva wrote: > In article <3E56AB27.162B9193@yahoo.com>, > CBFalconer wrote: > >>In fact I think the x86 world would be better off using segments >>in an intelligent manner. > > > You can't. 64k is too small. Even the 256k segments Multics had available > over 10 years earlier turned out to be too small. > > I was writing in PL/M, which used the segments as intelligently as > possible... they were first class data types. And 64k was too small. > > >>An 8086 in 1978 was orders of magnitude more capable than a PDP8 >>or SDS 910 of 1964, and somewhat cheaper. > > > An 8080 was more capable than a PDP-8, for Turing's sake. > I don't want to change the context but Pete, can you tell us what you mean by that statement? More capable of what? What do you mean by capable? bob ###### Message-ID: <3E5792EB.E798F1F5@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <93F83D24B0A3E18E.B3103076363F175A.58C7F3DFE7653EAB@lp.airnews.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 53 Date: Sat, 22 Feb 2003 15:34:24 GMT NNTP-Posting-Host: 12.90.174.28 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc05-news.ops.worldnet.att.net 1045928064 12.90.174.28 (Sat, 22 Feb 2003 15:34:24 GMT) NNTP-Posting-Date: Sat, 22 Feb 2003 15:34:24 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!nf3.bellglobal.com!wn14feed!worldnet.att.net!bgtnsc05-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129657 alt.sys.pdp10:12537 "John R. Strohm" wrote: > "CBFalconer" wrote in message > > ... snip ... > > > > An 8086 in 1978 was orders of magnitude more capable than a PDP8 > > or SDS 910 of 1964, and somewhat cheaper. > > 1978??? I think you're a few years early. Funny - we had a large box running MPM86 on a blindingly fast 5 or 10 Mhz 8086 in 1978 or 1979. I think it had 5 or 10 Megs of hard drive and several terminal ports. > > And, in any event, by 1981, Motorola had the 68000 in volume > production, and Zilog had the Z8000. > > The ONLY reason that the 8086 went anywhere is that IBM selected > the 8088 for the IBM PC. This was a horrible mistake, but it was > driven by their desire for an 8-bit data bus, so they could fall > back to a Z-80 design if the 8088 didn't make it. This was a > bogus plan: at the same time that IBM Boca Raton, their Siberia > division, was designing the PC around the 8088, another division > was selling the SS9000 laboratory computer, which was built > around the 68000. True enough, but 8 bit busses are cheaper than 16, and the performance competed nicely with 4 Mhz Z80 installations (until the bios got in the way). The added memory space was a big plus, as were position independent code based on segment registers. If the 64180 (z180) had been available then it would probably never have happened, and 68000's might well be ruling the world today. The CP/M to whatever migration path was all important then. The stupidos were the ISA bus choices, especially IRQs, and Micro$loths misuse of the segmenting capabilities, which were probably Seattle computers misuse in their quick and dirty adaptation. Once Micro$loth got hold of it they weren't intelligent enough to make suitable revisions, just hacks. Meanwhile their "endearing" sales practices justified everything. DR also made fundamental mistakes in the rushed CP/M86 design, and never got in on the gold rush. Nobody really figured out how to use segments to advantage. They should have studied the HP3000 architecture. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 36 Date: Sat, 22 Feb 2003 16:38:28 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news1.calgary.shaw.ca 1045931908 24.71.223.147 (Sat, 22 Feb 2003 09:38:28 MST) NNTP-Posting-Date: Sat, 22 Feb 2003 09:38:28 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!out.nntp.be!propagator-sanjose!news-in-sanjose!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news1.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129887 alt.sys.pdp10:12551 On 22 Feb 2003 13:38:27 GMT in alt.folklore.computers, peter@taronga.com (Peter da Silva) wrote: >In article <3E56AB27.162B9193@yahoo.com>, >CBFalconer wrote: >>In fact I think the x86 world would be better off using segments >>in an intelligent manner. > >You can't. 64k is too small. Even the 256k segments Multics had available >over 10 years earlier turned out to be too small. An address space of 64KB was usable in the 1970s: witness the popularity of the PDP-11 in that time frame; but by the 1980s, a larger address space or virtual memory was getting essential for many apps: overlays were getting ridiculous. >I was writing in PL/M, which used the segments as intelligently as >possible... they were first class data types. And 64k was too small. They could be used as base registers addressing 1MB, the problem was there were too few of them: 8+8 or 16+16 base+index registers would have been usable; 4+5 (CS+PC, SS+SP+BP, DS+SI, ES+DI) plus 4 AC/index (AX, BX, CX, DX) registers was not enough even at that time in that architecture; a true GP register instruction set allowing any segment with any other register in all instruction address fields would have helped out somewhat, compared to the boneheaded segment override prefix hack (pej.) Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### From: bfranchuk@jetnet.ab.ca Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 22 Feb 2003 10:53:04 -0700 Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3E57B900.4030601@jetnet.ab.ca> User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.0.1) Gecko/20021005 X-Accept-Language: en-us, en MIME-Version: 1.0 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 16 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!feed.news.nacamar.de!news.maxwell.syr.edu!sn-xit-03!sn-xit-01!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129672 alt.sys.pdp10:12538 Peter da Silva wrote: > An 8080 was more capable than a PDP-8, for Turing's sake. HA HA HA. 8080 was slower and less efficent than a non-CMOS 8. Dec how ever did not understand the personal computer concept well as they expected main-frames forever, thus the pdp-8 lost out to the 8080 cause it was cheaper and more accessible to people. The design of the PDP 8 and 8080 however reflect different thoughts on what a computer was for the time.At one point a PDP-8 clone had 200 ns memory making the fastest 8 in production. I wonder what the PDP-11 would have been like if it was a 18 bit machine. Ben. ###### Message-ID: <3E57BFAD.57606AB@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <3E5794E4.6060303@bellatlantic.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 33 Date: Sat, 22 Feb 2003 18:37:29 GMT NNTP-Posting-Host: 12.90.169.48 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc05-news.ops.worldnet.att.net 1045939049 12.90.169.48 (Sat, 22 Feb 2003 18:37:29 GMT) NNTP-Posting-Date: Sat, 22 Feb 2003 18:37:29 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!colt.net!nycmny1-snf1.gtei.net!nycmny1-snh1.gtei.net!chcgil2-snh1.gtei.net!news.gtei.net!wn12feed!wn14feed!worldnet.att.net!bgtnsc05-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129924 alt.sys.pdp10:12552 bob smith wrote: > Peter da Silva wrote: > > CBFalconer wrote: > > ... snip ... > > > >>An 8086 in 1978 was orders of magnitude more capable than a PDP8 > >>or SDS 910 of 1964, and somewhat cheaper. > > > > > > An 8080 was more capable than a PDP-8, for Turing's sake. > > > > I don't want to change the context but Pete, can you tell > us what you mean by that statement? > More capable of what? What do you mean by capable? PDP8 8080 Memory space 4k 64k word size 12 8 register/arith size 12 16 memory cycle 5 uS? 500 nS instr. cycle 40 uS (PDP8S) 3 uS registers, inc. IP 2 ? 6 or 7 byte + 2 word IIRC -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 22 Feb 2003 19:28:09 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 22 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <93F83D24B0A3E18E.B3103076363F175A.58C7F3DFE7653EAB@lp.airnews.net> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1045942089 42008 146.186.61.46 (22 Feb 2003 19:28:09 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Sat, 22 Feb 2003 19:28:09 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!nntp-relay.ihug.net!ihug.co.nz!logbridge.uoregon.edu!news.cse.psu.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129716 In article , Roland Hutchinson wrote: >Glen Herrmannsfeldt wrote: >> The original PC had 64K and a >> cassette tape interface for storage to keep prices down. >It was worse than that. >When it was introduced, there was a 16K configuration, at least on >paper. This may have been basically a ploy so that they could charge >you for the additional 48K that fit (four? DIPs at a time) into sockets >on the motherboard. Nope; they existed. Saw them; demonstrated them; sold them . . . hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### Message-ID: <3E57D6D4.7070302@bellatlantic.net> From: bob smith User-Agent: Mozilla/5.0 (Windows; U; WinNT4.0; en-US; rv:1.0.1) Gecko/20020823 Netscape/7.0 X-Accept-Language: en-us, en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <3E57B900.4030601@jetnet.ab.ca> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Lines: 31 Date: Sat, 22 Feb 2003 19:50:07 GMT NNTP-Posting-Host: 138.88.112.150 X-Complaints-To: abuse@verizon.net X-Trace: nwrddc02.gnilink.net 1045943407 138.88.112.150 (Sat, 22 Feb 2003 14:50:07 EST) NNTP-Posting-Date: Sat, 22 Feb 2003 14:50:07 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!colt.net!easynet-quince!easynet.net!cox.net!cyclone1.gnilink.net!spamkiller.gnilink.net!nwrddc02.gnilink.net.POSTED!53ab2750!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129782 alt.sys.pdp10:12545 bfranchuk@jetnet.ab.ca wrote: > Peter da Silva wrote: > >> An 8080 was more capable than a PDP-8, for Turing's sake. > > HA HA HA. 8080 was slower and less efficent than a non-CMOS > 8. Dec how ever did not understand the personal computer > concept well as they expected main-frames forever, thus > the pdp-8 lost out to the 8080 cause it was cheaper and > more accessible to people. The design > of the PDP 8 and 8080 however reflect different thoughts > on what a computer was for the time.At one point a PDP-8 > clone had 200 ns memory making the fastest 8 in production. > I wonder what the PDP-11 would have been like if it was a > 18 bit machine. Ben. > Ben, Ya, it could have been a PDP9 or 15!! It could have been an eight and a half! Sorry, could not resist. Why not compare the 8080 to 6800 or 6502 or better still 6500? Yes, there was a 6500. It did have to be pulled tho for some legal reason I will let someone else recount. bob > > ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 23 Feb 2003 02:32:02 GMT Organization: TSS Inc. Lines: 20 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E56AB27.162B9193@yahoo.com> <3E5794E4.6060303@bellatlantic.net> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1045967522 58030 10.0.0.43 (23 Feb 2003 02:32:02 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 23 Feb 2003 02:32:02 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!headwall.stanford.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129864 alt.sys.pdp10:12549 In article <3E5794E4.6060303@bellatlantic.net>, bob smith wrote: >> An 8080 was more capable than a PDP-8, for Turing's sake. >I don't want to change the context but Pete, can you tell >us what you mean by that statement? >More capable of what? What do you mean by capable? More capable a processor. The 8080 was not a complete system, so you can't compare things like the I/O system without specifying which I/O system you're talking about, but as a CPU it was faster and had a larger address space (64 kilobytes beats 28 kilowords even if the words are 50% larger) and it was a flat linear address space, not split into 4kw chunks. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 23 Feb 2003 02:39:33 GMT Organization: TSS Inc. Lines: 38 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1045967973 58560 10.0.0.43 (23 Feb 2003 02:39:33 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 23 Feb 2003 02:39:33 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!headwall.stanford.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129833 alt.sys.pdp10:12547 In article <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com>, Brian Inglis wrote: >On 22 Feb 2003 13:38:27 GMT in alt.folklore.computers, >peter@taronga.com (Peter da Silva) wrote: >>In article <3E56AB27.162B9193@yahoo.com>, >>CBFalconer wrote: >>>In fact I think the x86 world would be better off using segments >>>in an intelligent manner. >>You can't. 64k is too small. Even the 256k segments Multics had available >>over 10 years earlier turned out to be too small. >An address space of 64KB was usable in the 1970s: witness the >popularity of the PDP-11 in that time frame; but by the 1980s, a >larger address space or virtual memory was getting essential for >many apps: overlays were getting ridiculous. The address space of the PDP-11 was too small for a segment-oriented operating system like Multics when the PDP-11 was introduced. That is why UNIX didn't follow the Multics design, but used a stream model for I/O. Multics eventually had to abandon a pure segment model, and its segments were 4x as large. >>I was writing in PL/M, which used the segments as intelligently as >>possible... they were first class data types. And 64k was too small. >They could be used as base registers addressing 1MB, In a way, but that 1MB would still have to be split into 64K chunks: No object in that address space could be larger than 64K. By the time the 8086 was out, 64K was clearly too small. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: Mark Crispin Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sun, 23 Feb 2003 02:19:07 -0800 Organization: Networks and Distributed Computing Lines: 35 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> NNTP-Posting-Host: shiva0.cac.washington.edu Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Trace: nntp1.u.washington.edu 1045995550 15236 (None) 140.142.17.39 X-Complaints-To: help@cac.washington.edu In-Reply-To: Content-Length: 314159 (believe this at your own risk) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!feed2.news.rcn.net!rcn!logbridge.uoregon.edu!news.u.washington.edu!140.142.17.34.MISMATCH!news.u.washington.edu!shiva0.cac.washington.edu!mrc Xref: chonsp.franklin.ch alt.folklore.computers:129697 alt.sys.pdp10:12541 On Sat, 23 Feb 2003, Peter da Silva wrote: > By the time the 8086 was out, 64K was clearly too small. Note that the unextended PDP-10 address space was effectively 1MB (depending how you look at it, it was actually 1 1/8MB to 1 1/4MB), and it was considered too small back in the 1970s. By the way, the extended PDP-10 architecture defined a 30-bit address space; this was equivalent to a 32-bit address space on a byte-oriented machine. However, DEC only ever implemented it to 23-bits virtual and 22-bits physical. XKL finally implemented the full 30-bits. My understanding was that it was fairly straightforward to push things to 27-bits (instead of a 32 word section table, have a whole page for section table), but that the last 3 bits were a major hassle since the KL10 page pointers only allowed for 18 bit page numbers. The reason why the PDP-10 ended up with a 30-bit address space was the expectation that for most purposes, 4GB was good enough (and the continued viability of 32-bit processors today seems to bear that out), and they wanted to leave space in pointers for index register and indirect. Certain individuals in the VAX community seized on the "30" number and claimed that "even after you fix the address space, the PDP-10 still has less than a VAX." This is, of course, false, since this was words and not bytes. But it didn't stop them from claiming it. I wonder if the VAX ever fixed the ridiculously small page size that it had. As I recall, a VAX page was a mere 512 bytes! -- Mark -- http://staff.washington.edu/mrc Science does not emerge from voting, party politics, or public debate. ###### User-Agent: Microsoft-Entourage/10.1.1.2418 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) From: nospam Newsgroups: alt.folklore.computers,alt.sys.pdp10 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> Mime-version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit Lines: 19 Date: Sun, 23 Feb 2003 13:30:21 GMT NNTP-Posting-Host: 144.136.91.243 X-Complaints-To: news@bigpond.net.au X-Trace: news-server.bigpond.net.au 1046007021 144.136.91.243 (Mon, 24 Feb 2003 00:30:21 EST) NNTP-Posting-Date: Mon, 24 Feb 2003 00:30:21 EST Organization: BigPond Internet Services (http://www.bigpond.net.au) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!news.tele.dk!news.tele.dk!small.news.tele.dk!ken-transit.news.telstra.net!news.telstra.net!news-server.bigpond.net.au!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129876 alt.sys.pdp10:12550 in article Pine.LNX.4.55.0302230200210.8879@shiva0.cac.washington.edu, Mark Crispin at mrc@CAC.Washington.EDU wrote on 23/02/2003 21:19: > > Certain individuals in the VAX community seized on the "30" number and > claimed that "even after you fix the address space, the PDP-10 still has > less than a VAX." This is, of course, false, since this was words and not > bytes. But it didn't stop them from claiming it. Lots of crap went down then, how history repeats. > > I wonder if the VAX ever fixed the ridiculously small page size that it > had. As I recall, a VAX page was a mere 512 bytes! The VAX from first to last is a 512Byte page. On alpha VMS still talk about these pages of 512bytes now called page lets as the physical page is the Alphas 8k page. Remember the VAX is Virtual Address eXtention, the extension being that of a PDP-11 and 512bytes would seem huge to them fokes :) ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC Reply-To: Anne & Lynn Wheeler From: Anne & Lynn Wheeler Message-ID: Organization: Wheeler&Wheeler User-Agent: Gnus/5.090015 (Oort Gnus v0.15) Emacs/21.2 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> Cancel-Lock: sha1:7DrnmvSxvHNWWujGBPdF1Zt3fVU= MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Lines: 74 Date: Sun, 23 Feb 2003 20:37:57 GMT NNTP-Posting-Host: 165.247.91.77 X-Complaints-To: abuse@earthlink.net X-Trace: newsread1.prod.itd.earthlink.net 1046032677 165.247.91.77 (Sun, 23 Feb 2003 12:37:57 PST) NNTP-Posting-Date: Sun, 23 Feb 2003 12:37:57 PST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!cyclone.bc.net!sjc70.webusenet.com!news.webusenet.com!newsfeed2.earthlink.net!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!newsread1.prod.itd.earthlink.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129957 alt.sys.pdp10:12554 Brian Inglis writes: > The 8KB page size would have been about right to allow growth for > total system memory on the VAX, and would probably have made > emulating 11 mode more straightforward. > 2KB pages were used in small IBM 370 OSes with up to 16MB system > memory, and 2KB was too small for full memory systems; 4KB was > used in bigger IBM 370 OSes and was okay for 16MB up to 32MB > system memory; 8KB would have been about right for the VAX time > frame and target system memory sizes, and was eventually used in > the Alpha. 2kb was probably 512kbyte or less. VS1 (which had 2k pages) running under VM with ecps and handshaking ... typically ran faster on 1mbyte (or larger real machine) under VM than running standalone (even after taking out VM's fixed kernel requirements of up to 200k ... aka 20 percent of one mbyte real machine). In handshaking mode, VS1 was given a 16mbyte virtual address space by VM (VS1 treated it as if running on 16mbyte real machine) ... and VS1 didn't do any paging ... but let VM handle it all using 4k pages ... aka VS1 normally had a single 16mbyte virtual address space ... and in handshaking mode it was mapped one-to-one with the virtual machine 16mbyte address apace. later in 3081 time frame with 3380s (early '80s) ... "big" pages were introduced for both VM and MVS. The hardware translation was still all 4k pages, however "big" pages was change to both vm & mvs page I/O to do transfers in units of full-track ... aka 40k or ten pages at a time. when being replaced .... ten pages from a virtual address space were collected together and written out in one i/o. On a page fault for any of the ten 4k pages in a unit ... all ten pages would be read in one operation. the issue that dominated wasn't so much the packing in real storage but the amount transfered in a single unit ... given the i/o operational characteristics. by the time-frame of 1mbyte real storages, I/O was more of a thruput constraing than real storage ... and strategies (like caching) were starting to appear that attempted to leverage real storage for i/o bottleneck compensation. misc. past discussion of big pages: http://www.garlic.com/~lynn/2001k.html#60 Defrag in linux? - Newbie question http://www.garlic.com/~lynn/2002b.html#20 index searching http://www.garlic.com/~lynn/2002c.html#29 Page size (was: VAX, M68K complex instructions) http://www.garlic.com/~lynn/2002c.html#48 Swapper was Re: History of Login Names http://www.garlic.com/~lynn/2002e.html#8 What are some impressive page rates? http://www.garlic.com/~lynn/2002e.html#11 What are some impressive page rates? http://www.garlic.com/~lynn/2002f.html#20 Blade architectures http://www.garlic.com/~lynn/2002l.html#36 Do any architectures use instruction count instead of timer http://www.garlic.com/~lynn/2002m.html#4 Handling variable page sizes? http://www.garlic.com/~lynn/2002m.html#7 Handling variable page sizes? http://www.garlic.com/~lynn/2003b.html#69 Disk drives as commodities. Was Re: Yamhill past discussions of 360/67 w/105 4k pages (after fixed kernel requirements) and 3081k with >7000 4k pages (after fixed kernel requirements): http://www.garlic.com/~lynn/93.html#31 Big I/O or Kicking the Mainframe out the Door http://www.garlic.com/~lynn/95.html#10 Virtual Memory (A return to the past?) http://www.garlic.com/~lynn/98.html#46 The god old days(???) http://www.garlic.com/~lynn/99.html#4 IBM S/360 http://www.garlic.com/~lynn/2001f.html#62 any 70's era supercomputers that ran as slow as today's supercomputers? http://www.garlic.com/~lynn/2001l.html#40 MVS History (all parts) http://www.garlic.com/~lynn/2001l.html#61 MVS History (all parts) http://www.garlic.com/~lynn/2001m.html#23 Smallest Storage Capacity Hard Disk? http://www.garlic.com/~lynn/2002.html#5 index searching http://www.garlic.com/~lynn/2002b.html#11 Microcode? (& index searching) http://www.garlic.com/~lynn/2002b.html#20 index searching http://www.garlic.com/~lynn/2002e.html#8 What are some impressive page rates? http://www.garlic.com/~lynn/2002e.html#9 What are some impressive page rates? http://www.garlic.com/~lynn/2002i.html#16 AS/400 and MVS - clarification please -- Anne & Lynn Wheeler | lynn@garlic.com - http://www.garlic.com/~lynn/ Internet trivia, 20th anniv: http://www.garlic.com/~lynn/rfcietff.htm ##### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 23 Feb 2003 15:01:48 GMT Organization: TSS Inc. Lines: 13 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046012508 97889 10.0.0.43 (23 Feb 2003 15:01:48 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 23 Feb 2003 15:01:48 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!news.tele.dk!news.tele.dk!small.news.tele.dk!enews.sgi.com!news.spies.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129979 alt.sys.pdp10:12559 In article , Mark Crispin wrote: >I wonder if the VAX ever fixed the ridiculously small page size that it >had. As I recall, a VAX page was a mere 512 bytes! I suspect that was probably overreaction to the PDP-11's 4kw pages, which were clearly too large for a 32kw address space. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 29 Date: Sun, 23 Feb 2003 17:58:53 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news3.calgary.shaw.ca 1046023133 24.71.223.147 (Sun, 23 Feb 2003 10:58:53 MST) NNTP-Posting-Date: Sun, 23 Feb 2003 10:58:53 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!cyclone.bc.net!sjc70.webusenet.com!news.webusenet.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news3.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129989 alt.sys.pdp10:12560 On 23 Feb 2003 15:01:48 GMT in alt.folklore.computers, peter@taronga.com (Peter da Silva) wrote: >In article , >Mark Crispin wrote: >>I wonder if the VAX ever fixed the ridiculously small page size that it >>had. As I recall, a VAX page was a mere 512 bytes! > >I suspect that was probably overreaction to the PDP-11's 4kw pages, which >were clearly too large for a 32kw address space. The 8KB page size would have been about right to allow growth for total system memory on the VAX, and would probably have made emulating 11 mode more straightforward. 2KB pages were used in small IBM 370 OSes with up to 16MB system memory, and 2KB was too small for full memory systems; 4KB was used in bigger IBM 370 OSes and was okay for 16MB up to 32MB system memory; 8KB would have been about right for the VAX time frame and target system memory sizes, and was eventually used in the Alpha. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### From: bfranchuk@jetnet.ab.ca Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sun, 23 Feb 2003 11:32:30 -0700 Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3E5913BE.8050207@jetnet.ab.ca> User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.0.1) Gecko/20021005 X-Accept-Language: en-us, en MIME-Version: 1.0 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 14 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!sn-xit-03!sn-xit-06!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129944 alt.sys.pdp10:12553 Peter da Silva wrote: > By the time the 8086 was out, 64K was clearly too small. 64K bytes that is. 32K words were the typical max size of many computers between 1970 and 1980. 64kb data/code space are the mininum for general user programs. I really don't know what the next level program size because the next size up is 32 bit bloatware with RISC or BrainDead CPU's like 386+ forcing programs to be large.I suspect 24 bit addressing would cover most programing needs if bitmapped graphics did not take up so much memory. Ben. ###### Message-ID: <3E5924DC.1000604@bellatlantic.net> From: bob smith User-Agent: Mozilla/5.0 (Windows; U; WinNT4.0; en-US; rv:1.0.1) Gecko/20020823 Netscape/7.0 X-Accept-Language: en-us, en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Lines: 44 Date: Sun, 23 Feb 2003 19:35:09 GMT NNTP-Posting-Host: 138.88.112.150 X-Complaints-To: abuse@verizon.net X-Trace: nwrddc02.gnilink.net 1046028909 138.88.112.150 (Sun, 23 Feb 2003 14:35:09 EST) NNTP-Posting-Date: Sun, 23 Feb 2003 14:35:09 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed-east.nntpserver.com!nntpserver.com!chi1.webusenet.com!news.webusenet.com!cyclone1.gnilink.net!spamkiller.gnilink.net!nwrddc02.gnilink.net.POSTED!53ab2750!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129964 alt.sys.pdp10:12557 Mark Crispin wrote: > On Sat, 23 Feb 2003, Peter da Silva wrote: > >>By the time the 8086 was out, 64K was clearly too small. > > > Note that the unextended PDP-10 address space was effectively 1MB > (depending how you look at it, it was actually 1 1/8MB to 1 1/4MB), and it > was considered too small back in the 1970s. > > By the way, the extended PDP-10 architecture defined a 30-bit address > space; this was equivalent to a 32-bit address space on a byte-oriented > machine. However, DEC only ever implemented it to 23-bits virtual and > 22-bits physical. > > XKL finally implemented the full 30-bits. My understanding was that it > was fairly straightforward to push things to 27-bits (instead of a 32 word > section table, have a whole page for section table), but that the last 3 > bits were a major hassle since the KL10 page pointers only allowed for 18 > bit page numbers. > > The reason why the PDP-10 ended up with a 30-bit address space was the > expectation that for most purposes, 4GB was good enough (and the continued > viability of 32-bit processors today seems to bear that out), and they > wanted to leave space in pointers for index register and indirect. > > Certain individuals in the VAX community seized on the "30" number and > claimed that "even after you fix the address space, the PDP-10 still has > less than a VAX." This is, of course, false, since this was words and not > bytes. But it didn't stop them from claiming it. And some of the vaxinated central engineering folks would not allow 2080 to implement 32 bit addressing. Something about too much of a discriminator when compared to vax. bob > > I wonder if the VAX ever fixed the ridiculously small page size that it > had. As I recall, a VAX page was a mere 512 bytes! > > -- Mark -- > > http://staff.washington.edu/mrc > Science does not emerge from voting, party politics, or public debate. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Antti Louko Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> Message-ID: Lines: 51 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.1 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: 23 Feb 2003 23:55:40 +0200 NNTP-Posting-Host: 81.17.195.26 X-Complaints-To: newsmaster@saunalahti.com X-Trace: reader1.news.jippii.net 1046037345 81.17.195.26 (Sun, 23 Feb 2003 23:55:45 EET) NNTP-Posting-Date: Sun, 23 Feb 2003 23:55:45 EET Organization: Saunalahti Customer Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!psinet-eu-nl!eusc.inter.net!feed.news.nacamar.de!uio.no!newsfeed.kolumbus.fi!feeder2.news.jippii.net!feeder1.news.jippii.net!reader1.news.jippii.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:129966 alt.sys.pdp10:12558 Brian Inglis writes: > On 23 Feb 2003 15:01:48 GMT in alt.folklore.computers, > peter@taronga.com (Peter da Silva) wrote: > > >In article , > >Mark Crispin wrote: > >>I wonder if the VAX ever fixed the ridiculously small page size that it > >>had. As I recall, a VAX page was a mere 512 bytes! > > > >I suspect that was probably overreaction to the PDP-11's 4kw pages, which > >were clearly too large for a 32kw address space. > > The 8KB page size would have been about right to allow growth for > total system memory on the VAX, and would probably have made > emulating 11 mode more straightforward. A posting from year 1990 offers another (and quite CISCy) explanation: ---------------------------------------------------------------------- From: gideony@microsoft.UUCP (Gideon YUVAL) Newsgroups: comp.arch Subject: Re: Is handling off-alignment important? Message-ID: <56353@microsoft.UUCP> Date: 7 Aug 90 00:29:29 GMT References: <104037@convex.convex.com> <8840014@hpfcso.HP.COM> <2357@crdos1.crd.ge.COM> <25900@mimsy.umd.edu> Reply-To: gideony@microsoft.UUCP (Gideon YUVAL) Organization: Microsoft Corp., Redmond WA Lines: 20 In article <25900@mimsy.umd.edu> chris@mimsy.umd.edu (Chris Torek) writes: >In article <2357@crdos1.crd.ge.COM> davidsen@crdos1.crd.ge.COM >(Wm E Davidsen Jr) writes: .... ... ... ... ... ... ... ... ... ... ... > Consider the >following VAX gem: > addl3 *0(r1),*0(r2),*0(r3) .... > total 14 pages for one `simple' `addl3' instruction. >(Imagine the fun with a six-argument instruction like `index'!) I'm told this is why the VAX has a 512-byte page -- a 128KB box was planned (but never made), and in the worst-case (an instruction straddling 40 pages), would have deadlocked at <40 pages available. The page-size thus had to be less than one-fortieth of the page-list of a bare-minimum (1977!) system. -- Gideon Yuval, gideony@microsof.UUCP, 206-882-8080 (fax:206-883-8101;TWX:160520) ---------------------------------------------------------------------- ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Mon, 24 Feb 03 09:13:56 GMT Organization: UltraNet Communications, Inc. Lines: 31 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> X-Trace: UmFuZG9tSVZg7kN2etE7l68e5KxRsUGVXa8NSJ/iE/4ia30K00DImmm2rJQ+UARw X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 24 Feb 2003 09:56:57 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-175 Xref: chonsp.franklin.ch alt.folklore.computers:130025 alt.sys.pdp10:12565 In article , Mark Crispin wrote: >On Sun, 23 Feb 2003, John R. Levine wrote: >> Considering all the microcode bugs >> that showed up in the Vax, I get the impression that they designed the >> instruction set to match what they thought programmers or compiler >> writers would want without a whole lot of thought about how you would >> implement it other than in a naive low-performance version like the >> 730. > >This always seemed obvious to me. More specifically, the VAX instruction >set always looked to me like a PDP-11 assembly language programmer's idea >of what a compiler writer would want. I've heard that much of the VAX >instruction set was never used by compilers. That was the targeted market place at the time. It was the impetus behind the string instructions that were added into the KL. Customers weren't looking for computing in that era; they were looking for faster data processing. Look at all the manpower and money we put into data base thingies at that time. Pure computing was getting put into its niche. > >The PDP-10 instruction set, on the other hand, was an assembly language >programmer's idea of what an assembly language programmer would want! ;-) I was very, very spoiled. :-) /BAH Subtract a hundred and four for e-mail. ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: <3eqi5v40q3kcpl9cuhit4oedei2ifs8c6m@4ax.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 77 Date: Mon, 24 Feb 2003 01:01:43 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news2.calgary.shaw.ca 1046048503 24.71.223.147 (Sun, 23 Feb 2003 18:01:43 MST) NNTP-Posting-Date: Sun, 23 Feb 2003 18:01:43 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!c03.atl99!rip!news.webusenet.com!prodigy.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news2.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130122 alt.sys.pdp10:12571 On 23 Feb 2003 23:55:40 +0200 in alt.folklore.computers, Antti Louko wrote: >Brian Inglis writes: > >> On 23 Feb 2003 15:01:48 GMT in alt.folklore.computers, >> peter@taronga.com (Peter da Silva) wrote: >> >> >In article , >> >Mark Crispin wrote: >> >>I wonder if the VAX ever fixed the ridiculously small page size that it >> >>had. As I recall, a VAX page was a mere 512 bytes! >> > >> >I suspect that was probably overreaction to the PDP-11's 4kw pages, which >> >were clearly too large for a 32kw address space. >> >> The 8KB page size would have been about right to allow growth for >> total system memory on the VAX, and would probably have made >> emulating 11 mode more straightforward. > >A posting from year 1990 offers another (and quite CISCy) explanation: > >---------------------------------------------------------------------- >From: gideony@microsoft.UUCP (Gideon YUVAL) >Newsgroups: comp.arch >Subject: Re: Is handling off-alignment important? >Message-ID: <56353@microsoft.UUCP> >Date: 7 Aug 90 00:29:29 GMT >References: <104037@convex.convex.com> <8840014@hpfcso.HP.COM> <2357@crdos1.crd.ge.COM> <25900@mimsy.umd.edu> >Reply-To: gideony@microsoft.UUCP (Gideon YUVAL) >Organization: Microsoft Corp., Redmond WA >Lines: 20 > >In article <25900@mimsy.umd.edu> chris@mimsy.umd.edu (Chris Torek) writes: >>In article <2357@crdos1.crd.ge.COM> davidsen@crdos1.crd.ge.COM >>(Wm E Davidsen Jr) writes: >.... ... ... ... ... ... ... ... ... ... ... > Consider the >>following VAX gem: > >> addl3 *0(r1),*0(r2),*0(r3) >.... >> total 14 pages for one `simple' `addl3' instruction. > >>(Imagine the fun with a six-argument instruction like `index'!) > >I'm told this is why the VAX has a 512-byte page -- a 128KB box >was planned (but never made), and in the worst-case (an instruction >straddling 40 pages), would have deadlocked at <40 pages available. > >The page-size thus had to be less than one-fortieth of the page-list >of a bare-minimum (1977!) system. If they had known their existing PDP-11, -10, and -20 customer bases better, they could maybe have come up with a more realistic assessement. Most VAX-11 customers would likely be upgrading from fully loaded 11/70s with (2?MW) 4MB or 512 8KB pages of PDP-11 code and data, which would likely be fully occupied at peak load; native VAX applications could possibly double the size of application code; and VMS would probably be larger than any PDP-11 OS (max 64KB, 128KB with I/D space), so minimum VAX-11 system memory should start with the maximum 11/70 memory. Factor in what -10 and -20 users would need if they migrated, add a bit for expansion, allow for doubling every 18-24 months, and you can get a reasonable maximum memory size for the expected life of a system (say 64MB after 5 years for the first models). It might be interesting to compare estimates with what DEC offered on the first VAX models. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### From: shoppa@trailing-edge.com (Tim Shoppa) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 23 Feb 2003 18:44:55 -0800 Organization: http://groups.google.com/ Lines: 15 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> NNTP-Posting-Host: 63.73.218.144 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-Trace: posting.google.com 1046054695 8714 127.0.0.1 (24 Feb 2003 02:44:55 GMT) X-Complaints-To: groups-abuse@google.com NNTP-Posting-Date: 24 Feb 2003 02:44:55 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!newscore.univie.ac.at!newsfeed01.sul.t-online.de!t-online.de!feed.news.nacamar.de!news.maxwell.syr.edu!newsfeed.stanford.edu!postnews1.google.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130111 alt.sys.pdp10:12569 peter@taronga.com (Peter da Silva) wrote in message news:... > In article , > Mark Crispin wrote: > >I wonder if the VAX ever fixed the ridiculously small page size that it > >had. As I recall, a VAX page was a mere 512 bytes! > > I suspect that was probably overreaction to the PDP-11's 4kw pages, which > were clearly too large for a 32kw address space. I will agree that 4kw pages on a PDP-11 were uncomfortably large, but if you wrap your brain around the problem it could be made into a good fit. After a while you begin not to notice your own distorted thinking, IOW :-) Tim. ###### From: johnl@iecc.com (John R. Levine) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 23 Feb 2003 21:53:11 -0500 Organization: I.E.C.C., Trumansburg NY USA Lines: 22 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: xuxa.iecc.com X-Trace: xuxa.iecc.com 1046055191 17797 208.31.42.42 (24 Feb 2003 02:53:11 GMT) X-Complaints-To: abuse@iecc.com NNTP-Posting-Date: 24 Feb 2003 02:53:11 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!newscore.univie.ac.at!feed.news.nacamar.de!newsfeed.icl.net!newsfeed.fjserv.net!out.nntp.be!propagator-sanjose!news-in-sanjose!in.nntp.be!gail.ripco.com!news.lightlink.com!xuxa.iecc.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130012 alt.sys.pdp10:12563 >>I wonder if the VAX ever fixed the ridiculously small page size that it >>had. As I recall, a VAX page was a mere 512 bytes! > >I suspect that was probably overreaction to the PDP-11's 4kw pages, which >were clearly too large for a 32kw address space. I always assumed it was to match the block size of the disks that they were selling. Or they figured that if you wanted bigger logical pages, you could cluster physical ones, which is indeed what we did. I know about the amazing instruction that touches 40 pages, but I'm not sure I find that compelling. Considering all the microcode bugs that showed up in the Vax, I get the impression that they designed the instruction set to match what they thought programmers or compiler writers would want without a whole lot of thought about how you would implement it other than in a naive low-performance version like the 730. -- John R. Levine, IECC, POB 727, Trumansburg NY 14886 +1 607 387 6869 johnl@iecc.com, Village Trustee and Sewer Commissioner, http://iecc.com/johnl, Member, Provisional board, Coalition Against Unsolicited Commercial E-mail ###### From: Mark Crispin Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Mon, 24 Feb 2003 00:26:40 -0800 Organization: Networks and Distributed Computing Lines: 20 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: shiva0.cac.washington.edu Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Trace: nntp1.u.washington.edu 1046075203 16392 (None) 140.142.17.39 X-Complaints-To: help@cac.washington.edu In-Reply-To: Content-Length: 314159 (believe this at your own risk) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!news.gtei.net!newsfeed.mathworks.com!arclight.uoregon.edu!logbridge.uoregon.edu!news.u.washington.edu!140.142.17.34.MISMATCH!news.u.washington.edu!shiva0.cac.washington.edu!mrc Xref: chonsp.franklin.ch alt.folklore.computers:130019 alt.sys.pdp10:12564 On Sun, 23 Feb 2003, John R. Levine wrote: > Considering all the microcode bugs > that showed up in the Vax, I get the impression that they designed the > instruction set to match what they thought programmers or compiler > writers would want without a whole lot of thought about how you would > implement it other than in a naive low-performance version like the > 730. This always seemed obvious to me. More specifically, the VAX instruction set always looked to me like a PDP-11 assembly language programmer's idea of what a compiler writer would want. I've heard that much of the VAX instruction set was never used by compilers. The PDP-10 instruction set, on the other hand, was an assembly language programmer's idea of what an assembly language programmer would want! ;-) -- Mark -- http://staff.washington.edu/mrc Science does not emerge from voting, party politics, or public debate. ###### From: alan.nospam@glaramara.freeserve.co.uk (Alan J. Wylie) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 24 Feb 2003 11:45:50 +0000 Organization: very little Lines: 34 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3eqi5v40q3kcpl9cuhit4oedei2ifs8c6m@4ax.com> NNTP-Posting-Host: 217.135.53.15 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Trace: news6.svr.pol.co.uk 1046087494 4538 217.135.53.15 (24 Feb 2003 11:51:34 GMT) NNTP-Posting-Date: 24 Feb 2003 11:51:34 GMT X-Complaints-To: abuse@theplanet.net User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!newspeer.lavaseals.co.uk!diablo.theplanet.net!news.theplanet.net!bilbo!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130031 alt.sys.pdp10:12566 On Mon, 24 Feb 2003 01:01:43 GMT, Brian Inglis said: >>> (Imagine the fun with a six-argument instruction like `index'!) >> I'm told this is why the VAX has a 512-byte page -- a 128KB box >> was planned (but never made), and in the worst-case (an instruction >> straddling 40 pages), would have deadlocked at <40 pages available. >> >> The page-size thus had to be less than one-fortieth of the >> page-list of a bare-minimum (1977!) system. > If they had known their existing PDP-11, -10, and -20 customer bases > better, they could maybe have come up with a more realistic > assessement. Most VAX-11 customers would likely be upgrading from > fully loaded 11/70s with (2?MW) 4MB or 512 8KB pages of PDP-11 code > and data, which would likely be fully occupied at peak load; native > VAX applications could possibly double the size of application code; > and VMS would probably be larger than any PDP-11 OS (max 64KB, 128KB > with I/D space), so minimum VAX-11 system memory should start with > the maximum 11/70 memory. Factor in what -10 and -20 users would > need if they migrated, add a bit for expansion, allow for doubling > every 18-24 months, and you can get a reasonable maximum memory size > for the expected life of a system (say 64MB after 5 years for the > first models). It might be interesting to compare estimates with > what DEC offered on the first VAX models. The VAX 11-780 I used in 1979-1985 (serial number GAL 42) arrived with 512 Kb of memory. Some time later an extra 256Kb was added, at great expense. -- Alan J. Wylie http://www.glaramara.freeserve.co.uk/ I'm looking for a job. CV at http://www.glaramara.freeserve.co.uk/cv.html Unix/Linux/C/Internet/embedded and lots more. UK:Bradford/Manchester ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 24 Feb 2003 11:51:35 GMT Organization: TSS Inc. Lines: 20 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046087495 63663 10.0.0.43 (24 Feb 2003 11:51:35 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 24 Feb 2003 11:51:35 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news-hog.berkeley.edu!ucberkeley!enews.sgi.com!news.spies.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130073 alt.sys.pdp10:12567 In article , Mark Crispin wrote: >This always seemed obvious to me. More specifically, the VAX instruction >set always looked to me like a PDP-11 assembly language programmer's idea >of what a compiler writer would want. I've heard that much of the VAX >instruction set was never used by compilers. I think it was more a matter of a PDP-11 assembly language programmer's idea of what a macro assembler writer would want. Most of the DEC PDP-11 APIs were defined in terms of assembly macros, and the regularity of the instruction set encouraged very complex macros. The actual bytes and words these macros laid down were rarely if ever defined explicitly. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 References: <3E05FE9D.B3A12286@vax6k.openecs.org> Reply-To: sarr@umich.edu Organization: University of Michigan Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) X-Newsreader: trn 4.0-test72 (19 April 1999) From: sarr@tetris.gpcc.itd.umich.edu (Sarr J. Blumson) Lines: 18 Message-ID: Date: Mon, 24 Feb 2003 14:54:17 GMT NNTP-Posting-Host: 141.211.2.208 X-Trace: news.itd.umich.edu 1046098457 141.211.2.208 (Mon, 24 Feb 2003 09:54:17 EST) NNTP-Posting-Date: Mon, 24 Feb 2003 09:54:17 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!nntp-out.monmouth.com!newspeer.monmouth.com!news-peer-east1.sprintlink.net!news.sprintlink.net!newsfeed.mathworks.com!newsxfer.eecs.umich.edu!news.itd.umich.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130078 alt.sys.pdp10:12568 In article , Mark Crispin wrote: > >This always seemed obvious to me. More specifically, the VAX instruction >set always looked to me like a PDP-11 assembly language programmer's idea >of what a compiler writer would want. I've heard that much of the VAX >instruction set was never used by compilers. I've read (maybe in Hennessey & Patterson, but it's been a long time and I don't have a copy handy) that on a 780 writing out the simple instructions to do a CALL actually runs faster than the CALL instruction. This insight is apparently part of how RISC was born. -- -------- Sarr Blumson sarr@umich.edu voice: +1 734 998 9932 home: +1 734 665 9591 JSTOR, University of Michigan http://www-personal.umich.edu/~sarr/ ###### From: bfranchuk@jetnet.ab.ca Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Mon, 24 Feb 2003 12:08:25 -0700 Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3E5A6DA9.6030005@jetnet.ab.ca> User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.0.1) Gecko/20021005 X-Accept-Language: en-us, en MIME-Version: 1.0 References: <3E05FE9D.B3A12286@vax6k.openecs.org> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 14 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!news.tele.dk!news.tele.dk!small.news.tele.dk!news.maxwell.syr.edu!sn-xit-03!sn-xit-01!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130007 alt.sys.pdp10:12562 Sarr J. Blumson wrote: > I've read (maybe in Hennessey & Patterson, but it's been a long time and I > don't have a copy handy) that on a 780 writing out the simple instructions > to do a CALL actually runs faster than the CALL instruction. This insight > is apparently part of how RISC was born. > The fact was memory got cheaper and faster was the biggest factor for RISC development. RISC machines of today are just as complex as CISC machines.In the 70's memory optimized for several people doing small amounts data procesing. 16 people on slow terminals all running the same text editor might have only 64k of virtual memory in use. Ben. ###### From: bfranchuk@jetnet.ab.ca Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Mon, 24 Feb 2003 12:22:53 -0700 Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3E5A710D.6030403@jetnet.ab.ca> User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.0.1) Gecko/20021005 X-Accept-Language: en-us, en MIME-Version: 1.0 References: <3E05FE9D.B3A12286@vax6k.openecs.org> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 16 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!sn-xit-03!sn-xit-06!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130003 alt.sys.pdp10:12561 jmfbahciv@aol.com wrote: > I was very, very spoiled. :-) > /BAH Can I quote that. :) It is said only about 10? lines of bug free code can be produced a day with all the docs and other needed information. With the focus to high level langauges where your line of code does more work (we hope) how did the PDP-10 fair in terms of writing portable code and and program designs compared to high level languages then and today. Ben? ###### Message-ID: <3E5A96D5.2E999F75@yahoo.com> From: Peter Flass X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 16 Date: Mon, 24 Feb 2003 22:06:43 GMT NNTP-Posting-Host: 24.194.50.82 X-Complaints-To: abuse@rr.com X-Trace: twister.nyroc.rr.com 1046124403 24.194.50.82 (Mon, 24 Feb 2003 17:06:43 EST) NNTP-Posting-Date: Mon, 24 Feb 2003 17:06:43 EST Organization: Road Runner Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!newspeer.lavaseals.co.uk!diablo.theplanet.net!newsfeed1.cidera.com!Cidera!cyclone.nyroc.rr.com!cyclone-out.nyroc.rr.com!twister.nyroc.rr.com.POSTED!53ab2750!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130140 alt.sys.pdp10:12572 Brian Inglis wrote: > The 8KB page size would have been about right to allow growth for > total system memory on the VAX, and would probably have made > emulating 11 mode more straightforward. > 2KB pages were used in small IBM 370 OSes with up to 16MB system > memory, and 2KB was too small for full memory systems; 4KB was > used in bigger IBM 370 OSes and was okay for 16MB up to 32MB > system memory; 8KB would have been about right for the VAX time > frame and target system memory sizes, and was eventually used in > the Alpha. > Right -- you're trading off page table/cache usage vs. memory wastage at the end of a page. Small page tables are a "good thing". My problem is i've grown so used to thinking page=4K that it's a major brain rewiring effort to think anything else. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 25 Feb 03 09:48:08 GMT Organization: UltraNet Communications, Inc. Lines: 55 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> X-Trace: UmFuZG9tSVYawqfURTuxEOzlKAOsxgY8CZIKdgNXirZuOCR7pGYNcLMBgnsWxm5P X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 25 Feb 2003 10:31:22 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-214 Xref: chonsp.franklin.ch alt.folklore.computers:130226 alt.sys.pdp10:12582 In article <3E5A710D.6030403@jetnet.ab.ca>, bfranchuk@jetnet.ab.ca wrote: >jmfbahciv@aol.com wrote: >> I was very, very spoiled. :-) >> /BAH >Can I quote that. :) > >It is said only about 10? lines of bug free code >can be produced a day with all the docs and other >needed information. This is a dumb metric to use since that's not how the work is done. It's a nice metric to use to estimate how long a complete project will take. I can't conceive of being limited to writing 10 "lines" of code/day. In the PDP-10 OS biz, 10 lines implies 10 instructions. I'd have to go out of my way to put a bug in 10 instructions. > ..With the focus to high level >langauges where your line of code does more work >(we hope) how did the PDP-10 fair in terms of >writing portable code and and program designs >compared to high level languages then and today. As you should know, I can't talk about today...or even the last decade. The term portable code, as its used today, just didn't exist back then. We wouldn't have trusted a compiler to do the correct instruction generation. DEC's answer to portable was BLISS, rather than the C you guys use. The problems with using BLISS was logistics (other than the fact that syntax sucked, wrong code was generated, debugging tools were of the BASIC level). If there was a bug in BLISS, it could never be fixed, tested and shipped _before_ the monitor shipped. Having a dependency on MACRO-10, COMPIL, and LINK was a big enough headache. Note that we had lengthy, formal field test criteria before we were allowed to ship to the field. Putting something into field test was not a trivial task (some days we thought not even God could give the approvals). I did find a way to get CUSPs into field test without the hassle...I think I can tell that secret now. We did try not to do the same thing twice between TOPS-10 and TOPS-20, despite the internal politics. Most of that portabled-ness came from design of the product, not just recompiling code on a different platform. I don't see the design care getting done anymore; it may happen (since I'm way off on the sidelines of the work) but I don't get any whiffs of it. So, this a very roundabout way to tell you, I don't think I can answer your question. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 25 Feb 03 11:50:47 GMT Organization: UltraNet Communications, Inc. Lines: 22 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> X-Trace: UmFuZG9tSVarfEYIfTS7pEZ3flf7XxSlTr3jeog2acyxukfVGRa6LNEfok63Hsa0 X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 25 Feb 2003 12:33:58 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-214 Xref: chonsp.franklin.ch alt.folklore.computers:130224 alt.sys.pdp10:12581 In article , Trevor.Jenkins@suneidesis.com (Trevor Jenkins) wrote: >On 24 Feb 2003 18:24:16 -0800, Tim Shoppa wrote: > >> Too many people hear "macro" and they think of the pitifully weak C >> preprocessor; good Macro assemblers are far far more expressive, allowing >> you to essentially create a new language specifically tailored for your >> problem. > >First real macro processor I used was Peter Brown's ML/1, of course, on a >DEC-10 in the early 80s. And, when I hear macro, I think of either TECO or of JMF's little projects. He would write a macro (MACRO-10's way of generating code based on an argument), debug it for a week, then take it to David Nixon to get it acting correctly. /BAH Subtract a hundred and four for e-mail. ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 25 Feb 03 12:57:08 -0800 Organization: http://extra.newsguy.com Lines: 32 Message-ID: <1281.186T405T7773222@kltpzyxm.invalid> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> NNTP-Posting-Host: p-596.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!cyclone.bc.net!logbridge.uoregon.edu!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews2 Xref: chonsp.franklin.ch alt.folklore.computers:130242 alt.sys.pdp10:12583 In article peter@taronga.com (Peter da Silva) writes: >In article , wrote: > >>I can't conceive of being limited to writing 10 "lines" of >>code/day. In the PDP-10 OS biz, 10 lines implies 10 instructions. >>I'd have to go out of my way to put a bug in 10 instructions. > >The OS/MVS utility IEFBR14 was one instruction long, and required four >revisions before it was shipped, and a year later there was a new bug >reported in it. The fix required a doubling of the instruction count. :-) >>As you should know, I can't talk about today...or even >>the last decade. The term portable code, as its used today, >>just didn't exist back then. > >In the '70s? It sure did. FORTRAN was the language of choice for >portable code back then, but the first OS port I know of started >in 1976, and that was in C. Ah, but Barb said "portable _as it's used today_" (emphasis mine). Nowadays, "portable" means "runs on both Windows 95 and NT". :-p -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: shoppa@trailing-edge.com (Tim Shoppa) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 24 Feb 2003 18:24:16 -0800 Organization: http://groups.google.com/ Lines: 15 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> NNTP-Posting-Host: 63.73.218.144 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-Trace: posting.google.com 1046139857 25148 127.0.0.1 (25 Feb 2003 02:24:17 GMT) X-Complaints-To: groups-abuse@google.com NNTP-Posting-Date: 25 Feb 2003 02:24:17 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!out.nntp.be!propagator2-SanJose!in.nntp.be!feedwest.aleron.net!aleron.net!sjc70.webusenet.com!news.webusenet.com!sn-xit-02!sn-xit-06!sn-xit-08!supernews.com!postnews1.google.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130289 alt.sys.pdp10:12588 bfranchuk@jetnet.ab.ca wrote in message news:<3E5A710D.6030403@jetnet.ab.ca>... > With the focus to high level > langauges where your line of code does more work > (we hope) You just have to properly divide your languages between "high" and "low" level. In "high" level go your Macro assemblers, LISP, Perl etc. In "low" level go languages like C and C++. Too many people hear "macro" and they think of the pitifully weak C preprocessor; good Macro assemblers are far far more expressive, allowing you to essentially create a new language specifically tailored for your problem. Tim. ###### Message-ID: <3E5B170F.90501@beagle-ears.com> Date: Mon, 24 Feb 2003 23:11:11 -0800 From: Lars Poulsen User-Agent: Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.0.2) Gecko/20021120 Netscape/7.01 X-Accept-Language: en, da MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <3E57B900.4030601@jetnet.ab.ca> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Lines: 21 NNTP-Posting-Host: 207.154.106.6 X-Trace: azure.impulse.net 1046157057 190 207.154.106.6 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!feed.news.nacamar.de!news.maxwell.syr.edu!newsfeed-east.nntpserver.com!nntpserver.com!newsfeed.mountaincable.net!news-out.nuthinbutnews.com!propagator2-sterling!news-in-sterling.newsfeed.com!sienna.impulse.net!azure.impulse.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130178 alt.sys.pdp10:12575 bfranchuk@jetnet.ab.ca wrote: > on what a computer was for the time.At one point a PDP-8 > clone had 200 ns memory making the fastest 8 in production. The fastest PDP-8 was the PDP-11/60 with a PDP-8 emulator loaded into its writable control store. I think that would have been around 1979 ? (I remember a DECUS presentation where they hauled a smelly rastafari guy out to face the crowd and told us he had done that as a demo of just how fast and flexible the 11/60 was. He looked none too pleased with the attention. And 200 ns would have been the 11/60's cache, which was about the right size to play main store for the PDP-8. I liked the 11/60. One of my customers bought one for his typesetting company in order to run the Atex software for his APS-5 photosetter. A lot of thought had gone into RAMP (reliability and maitainability program) features on the 11/60. It was a joy to service. -- / Lars Poulsen +1-805-569-5277 http://www.beagle-ears.com/lars/ 125 South Ontare Rd, Santa Barbara, CA 93105 USA lars@beagle-ears.com ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <93F83D24B0A3E18E.B3103076363F175A.58C7F3DFE7653EAB@lp.airnews.net> Organization: Chez Inwap X-Newsreader: trn 4.0-test76 (Apr 2, 2001) From: inwap@inwap.com (Joe Smith) Originator: inwap@inwap.com (Joe Smith) Lines: 16 Message-ID: <4xG6a.4624$io.186657@iad-read.news.verio.net> Date: Tue, 25 Feb 2003 09:17:52 GMT NNTP-Posting-Host: 130.94.177.175 X-Complaints-To: abuse@verio.net X-Trace: iad-read.news.verio.net 1046164672 130.94.177.175 (Tue, 25 Feb 2003 09:17:52 GMT) NNTP-Posting-Date: Tue, 25 Feb 2003 09:17:52 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!news.tele.dk!news.tele.dk!small.news.tele.dk!newsfeed.icl.net!newsfeed.fjserv.net!iad-peer.news.verio.net!news.verio.net!iad-read.news.verio.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130283 alt.sys.pdp10:12586 In article <93F83D24B0A3E18E.B3103076363F175A.58C7F3DFE7653EAB@lp.airnews.net>, John R. Strohm wrote: >"CBFalconer" wrote in message >news:3E56AB27.162B9193@yahoo.com... >> >> An 8086 in 1978 was orders of magnitude more capable than a PDP8 >> or SDS 910 of 1964, and somewhat cheaper. > >1978??? I think you're a few years early. 1978 is correct. Go to Intel's Microprocessor Hall of Fame for the dates. http://www.intel.com/intel/intelis/museum/exhibit/hist_micro/hof/hof_main.htm -Joe -- See http://www.inwap.com/ for PDP-10 and "ReBoot" pages. ###### From: Trevor.Jenkins@suneidesis.com (Trevor Jenkins) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 25 Feb 2003 10:25:29 GMT Organization: suneidesis Lines: 19 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> NNTP-Posting-Host: 213.208.87.130 X-Trace: fu-berlin.de 1046168729 56351537 213.208.87.130 (16 [129086]) X-Orig-Path: Trevor.Jenkins User-Agent: slrn/0.9.6.2 (Linux) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!newscore.univie.ac.at!fu-berlin.de!uni-berlin.de!213.208.87.130!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130179 alt.sys.pdp10:12576 On 24 Feb 2003 18:24:16 -0800, Tim Shoppa wrote: > Too many people hear "macro" and they think of the pitifully weak C > preprocessor; good Macro assemblers are far far more expressive, allowing > you to essentially create a new language specifically tailored for your > problem. First real macro processor I used was Peter Brown's ML/1, of course, on a DEC-10 in the early 80s. Regards, Trevor British Sign Language is not inarticulate handwaving; it's a living language. Support the campaign for formal recognition by the British government now! Details at http://www.fdp.org.uk/ -- <>< Re: deemed! ###### From: Trevor.Jenkins@suneidesis.com (Trevor Jenkins) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 25 Feb 2003 13:28:21 GMT Organization: suneidesis Lines: 36 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> NNTP-Posting-Host: 213.208.87.130 X-Trace: fu-berlin.de 1046179701 56413283 213.208.87.130 (16 [129086]) X-Orig-Path: Trevor.Jenkins User-Agent: slrn/0.9.6.2 (Linux) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!news.tele.dk!news.tele.dk!small.news.tele.dk!fu-berlin.de!uni-berlin.de!213.208.87.130!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130183 alt.sys.pdp10:12577 On Tue, 25 Feb 03 11:50:47 GMT, jmfbahciv@aol.com wrote: > In article , > Trevor.Jenkins@suneidesis.com (Trevor Jenkins) wrote: > >On 24 Feb 2003 18:24:16 -0800, Tim Shoppa > wrote: > > > >> Too many people hear "macro" and they think of the pitifully weak C > >> preprocessor; good Macro assemblers are far far more expressive, allowing > >> you to essentially create a new language specifically tailored for your > >> problem. > > > >First real macro processor I used was Peter Brown's ML/1, of course, on a > >DEC-10 in the early 80s. > > And, when I hear macro, I think of either TECO or of JMF's > little projects. He would write a macro (MACRO-10's way of > generating code based on an argument), debug it for a week, > then take it to David Nixon to get it acting correctly. Ah teco. Yes, wrote lots of lovely "macros" in that too. My best being one for removing tags from huge SGML-encoded documents and it substituted entities as required. But that one was on VAX/VMS. Thankfully not something that I needed to do twice as the macro was highly convoluted. My TOPS-10 ones were simialr and did rather affect the system performance but what did I care I was the systems programmer --- got to blame the student users. :-) Regards, Trevor British Sign Language is not inarticulate handwaving; it's a living language. Support the campaign for formal recognition by the British government now! Details at http://www.fdp.org.uk/ -- <>< Re: deemed! ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 31 Date: Tue, 25 Feb 2003 15:25:02 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news1.calgary.shaw.ca 1046186702 24.71.223.147 (Tue, 25 Feb 2003 08:25:02 MST) NNTP-Posting-Date: Tue, 25 Feb 2003 08:25:02 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news-fra1.dfn.de!news-stu1.dfn.de!news-nue1.dfn.de!uni-erlangen.de!newsfeed.arcor-online.net!nntp-relay.ihug.net!ihug.co.nz!out.nntp.be!propagator2-SanJose!in.nntp.be!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news1.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130159 alt.sys.pdp10:12573 On Mon, 24 Feb 2003 12:22:53 -0700 in alt.folklore.computers, bfranchuk@jetnet.ab.ca wrote: >jmfbahciv@aol.com wrote: >> I was very, very spoiled. :-) >> /BAH >Can I quote that. :) > >It is said only about 10? lines of bug free code >can be produced a day with all the docs and other >needed information. Maybe on average across all "software engineers" and the duration of a project at NASA or other bureaucratic development shop with critical safety and heavy documentation requirements. But other studies have shown that there are orders of magnitude differences between programmers in both producing code and defects. Some people can produce 1000 lines of defect free code per day, others only produce 100 lines and they'll contain 10 defects, and some will produce 10 lines with 10 defects. If you've had to maintain others' code, you've probably come across code with this range of characteristics. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### From: bfranchuk@jetnet.ab.ca Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 25 Feb 2003 12:16:25 -0700 Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3E5BC109.2050003@jetnet.ab.ca> User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.0.1) Gecko/20021005 X-Accept-Language: en-us, en MIME-Version: 1.0 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 41 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!lnewspeer00.lnd.ops.eu.uu.net!emea.uu.net!newsfeed01.sul.t-online.de!t-online.de!news.belwue.de!news-stu1.dfn.de!news-han1.dfn.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!sn-xit-03!sn-xit-01!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130168 alt.sys.pdp10:12574 jmfbahciv@aol.com wrote: > This is a dumb metric to use since that's not how the > work is done. It's a nice metric to use to estimate how > long a complete project will take. > > I can't conceive of being limited to writing 10 "lines" of > code/day. In the PDP-10 OS biz, 10 lines implies 10 instructions. > I'd have to go out of my way to put a bug in 10 instructions. Lines of code is indeed a stupid way to judge program development. But I don't see software design much improved at all, from both a user and super-user point of view since the 70's. Coding is not the problem but the whole development cycle is. 1000 lines of code per day to me just means a fast typest. > As you should know, I can't talk about today...or even > the last decade. The term portable code, as its used today, > just didn't exist back then. We wouldn't have trusted a compiler > to do the correct instruction generation. DEC's answer to portable > was BLISS, rather than the C you guys use. The problems with > using BLISS was logistics (other than the fact that syntax sucked, > wrong code was generated, debugging tools were of the BASIC level). > If there was a bug in BLISS, it could never be fixed, tested and > shipped _before_ the monitor shipped. Having a dependency on > MACRO-10, COMPIL, and LINK was a big enough headache. Classic C I like because I don't like information hiding as newer languages do because you loose track of just what the program is doing. Now days for many linux/unix programs finding the right dynamic libraries and scripts and other build information is a pain. GUI libraries and compilers change every few months I find. > So, this a very roundabout way to tell you, I don't think I can > answer your question. This is a good answer. Thank you. Ben. ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 25 Feb 2003 20:12:45 GMT Organization: TSS Inc. Lines: 22 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046203965 67125 10.0.0.43 (25 Feb 2003 20:12:45 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 25 Feb 2003 20:12:45 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!mtvwca1-snh1.ops.genuity.net!paloalto-snf1.gtei.net!news.gtei.net!news.compaq.com!nntp-relay.ihug.net!ihug.co.nz!news-out.nuthinbutnews.com!propagator2-sterling!news-in-sterling.newsfeed.com!news-in.nuthinbutnews.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130271 alt.sys.pdp10:12584 In article , wrote: >I can't conceive of being limited to writing 10 "lines" of >code/day. In the PDP-10 OS biz, 10 lines implies 10 instructions. >I'd have to go out of my way to put a bug in 10 instructions. The OS/MVS utility IEFBR14 was one instruction long, and required four revisions before it was shipped, and a year later there was a new bug reported in it. >As you should know, I can't talk about today...or even >the last decade. The term portable code, as its used today, >just didn't exist back then. In the '70s? It sure did. FORTRAN was the language of choice for portable code back then, but the first OS port I know of started in 1976, and that was in C. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### Message-ID: <3E5BED2B.C90073B6@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 15 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: sccrnsc04 1046204782 12.241.15.59 (Tue, 25 Feb 2003 20:26:22 GMT) NNTP-Posting-Date: Tue, 25 Feb 2003 20:26:22 GMT Date: Tue, 25 Feb 2003 20:26:22 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!lnewspeer00.lnd.ops.eu.uu.net!emea.uu.net!newsfeed.icl.net!newsfeed.fjserv.net!c03.atl99!chi1.webusenet.com!news.webusenet.com!cyclone1.gnilink.net!wn14feed!wn13feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!sccrnsc04.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130295 alt.sys.pdp10:12589 jmfbahciv@aol.com wrote: > > [snip...] [snip...] [sinp...] > > So, this a very roundabout way to tell you, I don't think I can > answer your question. > CIRCUMLOCUTION,* n. A literary trick whereby the writer who has nothing to say breaks it gently to the reader. --Ambroise Bierce, _The Devil's Dictionary -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### Sender: eric@ruckus.brouhaha.com From: Eric Smith Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <3E57B900.4030601@jetnet.ab.ca> <3E5B170F.90501@beagle-ears.com> Organization: Eric Conspiracy Secret Labs X-Eric-Conspiracy: There is no conspiracy. Date: 25 Feb 2003 12:58:12 -0800 Message-ID: Lines: 14 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-Host: 209.66.107.17 X-Trace: 25 Feb 2003 13:43:34 -0800, 209.66.107.17 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!news.gtei.net!news-out.visi.com!hermes.visi.com!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!news.he.net!news.kjsl.com!news.spies.com!209.66.107.17 Xref: chonsp.franklin.ch alt.folklore.computers:130275 alt.sys.pdp10:12585 Lars Poulsen writes: > The fastest PDP-8 was the PDP-11/60 with a PDP-8 emulator loaded > into its writable control store. I think that would have been around > 1979 ? (I remember a DECUS presentation where they hauled a smelly > rastafari guy out to face the crowd and told us he had done that > as a demo of just how fast and flexible the 11/60 was. He looked > none too pleased with the attention. I don't know about the DECUS presentation, but the PDP-8 emulator on the 11/60 was developed by Richie Lary. I met him in 2001, and can't imagine anyone thinking of him as a "smelly rastafari guy". He was also involved in writing the microcode for the 11/780 and the F11, much of the software for the HSCxx, and a lot of other neat stuff. ###### From: Mark Crispin Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 25 Feb 2003 13:08:56 -0800 Organization: Networks and Distributed Computing Lines: 16 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> NNTP-Posting-Host: shiva1.cac.washington.edu Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Trace: nntp1.u.washington.edu 1046207339 17298 (None) 140.142.17.38 X-Complaints-To: help@cac.washington.edu In-Reply-To: <3E5BC109.2050003@jetnet.ab.ca> Content-Length: 314159 (believe this at your own risk) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!earthquake.efnet.com!efnet.com!cyclone.bc.net!logbridge.uoregon.edu!news.u.washington.edu!140.142.17.34.MISMATCH!news.u.washington.edu!shiva1.cac.washington.edu!mrc Xref: chonsp.franklin.ch alt.folklore.computers:130185 alt.sys.pdp10:12578 On Tue, 25 Feb 2003 bfranchuk@jetnet.ab.ca wrote: > But I don't see software design much improved at all, from both > a user and super-user point of view since the 70's. Don't say that. The Linux world is slowly re-discovering software design techniques which were well-known to us in the PDP-10 world. In another 10 years or so, they may reach the point where we were in the early 1980s. [OK, I'm an optimist.] At least we're no longer hearing "we can't add that to UNIX, since the code to do it won't fit in a 32K PDP-11" any more. -- Mark -- http://staff.washington.edu/mrc Science does not emerge from voting, party politics, or public debate. ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC Date: Wed, 26 Feb 2003 15:13:43 -0700 Lines: 11 Message-ID: <1fqygmr.wlt0j0loz280N%lars@bearnip.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1281.186T405T7773222@kltpzyxm.invalid> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1046297624 57211970 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130515 alt.sys.pdp10:12629 Peter da Silva wrote: > Charlie Gibbs wrote: > > Ah, but Barb said "portable _as it's used today_" (emphasis mine). > > Nowadays, "portable" means "runs on both Windows 95 and NT". :-p > > You are a nasty cynical man. I can't help but approve. I see nothing cynical about it. I have met people who said something along those lines _and meant it_. ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC Date: Wed, 26 Feb 2003 15:54:55 -0700 Lines: 22 Message-ID: <1fqzvng.1jp56h21apsuawN%lars@bearnip.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1046300096 58346592 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130527 alt.sys.pdp10:12632 Mark Crispin wrote: > But to some people, it is offensive if code is not written exclusively for > the newest and most gonzo version of Linux. The mere fact of code for > other platforms enrages them, much less *compromises* in order to be > portable to both. Mhm. By now most of the users of my little project have grokked that I do not run Linux, and that in fact the main development does not happen on a x86 box at all - but there have been times when error reports consisted of not much more than "The program crashed. You can get the coredump from $URL." Well, at least they had the presence of mind to save the coredump for later analysis. > Speaking of GNU EMACS, does anyone know how to turn off that damn "DOS > mode" in the newer Linux versions? It's infuriated me to the point that I > was going to switch to vi, only I discovered that vi also has that hack. vim knows the 'set fileformat=[unix|mac|dos]' command to change the lineend interpretation. It probably can also be setup to automatically force everything into the same format, but I never needed that much automatism. ###### Sender: eric@ruckus.brouhaha.com From: Eric Smith Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1281.186T405T7773222@kltpzyxm.invalid> <1fqygmr.wlt0j0loz280N%lars@bearnip.com> Organization: Eric Conspiracy Secret Labs X-Eric-Conspiracy: There is no conspiracy. Date: 26 Feb 2003 15:01:16 -0800 Message-ID: Lines: 9 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-Host: 209.66.107.17 X-Trace: 26 Feb 2003 15:46:50 -0800, 209.66.107.17 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.he.net!news.kjsl.com!news.spies.com!209.66.107.17 Xref: chonsp.franklin.ch alt.folklore.computers:130502 alt.sys.pdp10:12626 Peter da Silva wrote: > You are a nasty cynical man. I can't help but approve. lars@bearnip.com (Lars Duening) writes: > I see nothing cynical about it. I have met people who said something > along those lines _and meant it_. Just because you actually mean it does not prevent a statement from being cynical. ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC Date: 27 Feb 2003 01:09:03 GMT Organization: TSS Inc. Lines: 19 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1281.186T405T7773222@kltpzyxm.invalid> <1fqygmr.wlt0j0loz280N%lars@bearnip.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046308143 59529 10.0.0.43 (27 Feb 2003 01:09:03 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 27 Feb 2003 01:09:03 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!nntp.theplanet.net!inewsm1.nntp.theplanet.net!195.40.4.120.MISMATCH!easynet-quince!easynet.net!news-lond.gip.net!news.gsl.net!gip.net!news.tele.dk!news.tele.dk!small.news.tele.dk!enews.sgi.com!news.spies.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130489 alt.sys.pdp10:12618 In article <1fqygmr.wlt0j0loz280N%lars@bearnip.com>, Lars Duening wrote: >Peter da Silva wrote: >> Charlie Gibbs wrote: >> > Ah, but Barb said "portable _as it's used today_" (emphasis mine). >> > Nowadays, "portable" means "runs on both Windows 95 and NT". :-p >> You are a nasty cynical man. I can't help but approve. >I see nothing cynical about it. I have met people who said something >along those lines _and meant it_. That's what happens when you cut funding for mental health programs. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 26 Feb 03 10:33:35 GMT Organization: UltraNet Communications, Inc. Lines: 37 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> X-Trace: UmFuZG9tSVYs7PjNUhUbI6zRdU+yoq+II1FaR5wjUW7xHvLeiVbSWG4xlHmICt5+ X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 26 Feb 2003 11:16:59 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-218 Xref: chonsp.franklin.ch alt.folklore.computers:130394 alt.sys.pdp10:12604 In article , peter@taronga.com (Peter da Silva) wrote: >In article , wrote: >>I can't conceive of being limited to writing 10 "lines" of >>code/day. In the PDP-10 OS biz, 10 lines implies 10 instructions. >>I'd have to go out of my way to put a bug in 10 instructions. > >The OS/MVS utility IEFBR14 was one instruction long, and required four >revisions before it was shipped, and a year later there was a new bug >reported in it. > >>As you should know, I can't talk about today...or even >>the last decade. The term portable code, as its used today, >>just didn't exist back then. > >In the '70s? It sure did. FORTRAN was the language of choice for portable >code back then, but the first OS port I know of started in 1976, and that >was in C. > I didn't write that clearly enough :-). I didn't mean there wasn't portable code (I learned how to code using cards). I meant that portable wasn't the religion du jour that it is now. If you watch closely, portable is getting to mean (to newbies who should know better) that any executable can run on any hardware platform without help. This is due to the nefarious practice of portraying source and executable files as one. I first met this concept using BASIC on the PDP-10 KA; I considered it a very bad practice. (The practice was to do the compilation, linking, and EXE formulation with one command.) /BAH /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 26 Feb 03 10:34:56 GMT Organization: UltraNet Communications, Inc. Lines: 37 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <1281.186T405T7773222@kltpzyxm.invalid> X-Trace: UmFuZG9tSVZCYj0lJekz0/MS8wKs8HCFc5zIUabwtIQEhuuPhaqdtslbN4h72Y5A X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 26 Feb 2003 11:18:21 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!enews.sgi.com!logbridge.uoregon.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-218 Xref: chonsp.franklin.ch alt.folklore.computers:130397 alt.sys.pdp10:12605 In article <1281.186T405T7773222@kltpzyxm.invalid>, "Charlie Gibbs" wrote: >In article peter@taronga.com >(Peter da Silva) writes: > >>In article , wrote: >> >>>I can't conceive of being limited to writing 10 "lines" of >>>code/day. In the PDP-10 OS biz, 10 lines implies 10 instructions. >>>I'd have to go out of my way to put a bug in 10 instructions. >> >>The OS/MVS utility IEFBR14 was one instruction long, and required four >>revisions before it was shipped, and a year later there was a new bug >>reported in it. > >The fix required a doubling of the instruction count. :-) I wouldn't call that a bug; we used to call that missing code. > >>>As you should know, I can't talk about today...or even >>>the last decade. The term portable code, as its used today, >>>just didn't exist back then. >> >>In the '70s? It sure did. FORTRAN was the language of choice for >>portable code back then, but the first OS port I know of started >>in 1976, and that was in C. > >Ah, but Barb said "portable _as it's used today_" (emphasis mine). >Nowadays, "portable" means "runs on both Windows 95 and NT". :-p I think it's worse than that :-(. See my other post. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 26 Feb 03 10:38:20 GMT Organization: UltraNet Communications, Inc. Lines: 25 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BED2B.C90073B6@ev1.net> X-Trace: UmFuZG9tSVZSFEDFvC/4SQ/fme948vHSZlFpbKmt7lVhAShQPMQfUy7yMV0j6MXE X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 26 Feb 2003 11:21:45 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-218 Xref: chonsp.franklin.ch alt.folklore.computers:130373 alt.sys.pdp10:12598 In article <3E5BED2B.C90073B6@ev1.net>, Charles Richmond wrote: >jmfbahciv@aol.com wrote: >> >> [snip...] [snip...] [sinp...] >> >> So, this a very roundabout way to tell you, I don't think I can >> answer your question. >> >CIRCUMLOCUTION,* n. A literary trick whereby the writer >who has nothing to say breaks it gently to the reader. > --Ambroise Bierce, _The Devil's Dictionary > Smarty pants. I started out thinking I was going to answer the question. I finished up realizing I didn't answer the question. However, there was enough useful stuff in what I wrote that I had to acknowledge it wasn't an answer. So, rather than delete everything, I added the acknowledgement. Granted I should have written the "I don't know." at the beginning so that he had a choice of not wading through the text. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 26 Feb 03 10:40:04 GMT Organization: UltraNet Communications, Inc. Lines: 28 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <3E57B900.4030601@jetnet.ab.ca> <3E5B170F.90501@beagle-ears.com> X-Trace: UmFuZG9tSVZkPVmjanAGhiKDTb7CFOOygNoxP5xVhOCZPwuFkjXqMGyD2LmCGBYv X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 26 Feb 2003 11:23:29 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-218 Xref: chonsp.franklin.ch alt.folklore.computers:130380 alt.sys.pdp10:12600 In article , Eric Smith wrote: >Lars Poulsen writes: >> The fastest PDP-8 was the PDP-11/60 with a PDP-8 emulator loaded >> into its writable control store. I think that would have been around >> 1979 ? (I remember a DECUS presentation where they hauled a smelly >> rastafari guy out to face the crowd and told us he had done that >> as a demo of just how fast and flexible the 11/60 was. He looked >> none too pleased with the attention. > >I don't know about the DECUS presentation, but the PDP-8 emulator on >the 11/60 was developed by Richie Lary. I met him in 2001, and can't >imagine anyone thinking of him as a "smelly rastafari guy". He cut his hair???!!! I'm so disappointed. > >He was also involved in writing the microcode for the 11/780 and the >F11, much of the software for the HSCxx, and a lot of other neat stuff. He did more than that but I can't remember what. Tape Prep did quite a bit of work for him. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 26 Feb 03 10:43:28 GMT Organization: UltraNet Communications, Inc. Lines: 27 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> X-Trace: UmFuZG9tSVaIaXzz/K4GkGh2krhlwkrSdXw+xH2gnZb7XxzbQSLfNlwCW5JAMfM2 X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 26 Feb 2003 11:26:52 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-218 Xref: chonsp.franklin.ch alt.folklore.computers:130389 alt.sys.pdp10:12602 In article , Mark Crispin wrote: >On Tue, 25 Feb 2003 bfranchuk@jetnet.ab.ca wrote: >> But I don't see software design much improved at all, from both >> a user and super-user point of view since the 70's. > >Don't say that. The Linux world is slowly re-discovering software design >techniques which were well-known to us in the PDP-10 world. In another 10 >years or so, they may reach the point where we were in the early 1980s. >[OK, I'm an optimist.] Perhaps not (being the optimist). :-) One of the reasons it's so damned important to get those emulators out there is so people can have access to different designs that had different goals and tradeoffs. > >At least we're no longer hearing "we can't add that to UNIX, since the >code to do it won't fit in a 32K PDP-11" any more. I sure would like apps people to know about the usefulness of GETSEGs. IOW, you don't have to have the whole kitchen sink in core in order to open the refrigerator. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 26 Feb 03 13:22:59 GMT Organization: UltraNet Communications, Inc. Lines: 49 Message-ID: References: <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> <1046265903.251131@saucer.planet.gong> X-Trace: UmFuZG9tSVblTBjs0jZAclWa33o+LVIMQXqxZf6JFTxPakPEwcIk22btHaJ/i0F0 X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 26 Feb 2003 14:06:22 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-182-136 Xref: chonsp.franklin.ch alt.folklore.computers:130409 alt.sys.pdp10:12606 In article <1046265903.251131@saucer.planet.gong>, "Rupert Pigott" wrote: > wrote in message news:b3i89s$i0m$5@bob.news.rcn.net... >> In article , >> Mark Crispin wrote: >> >On Tue, 25 Feb 2003 bfranchuk@jetnet.ab.ca wrote: >> >> But I don't see software design much improved at all, from both >> >> a user and super-user point of view since the 70's. >> > >> >Don't say that. The Linux world is slowly re-discovering software design >> >techniques which were well-known to us in the PDP-10 world. In another >10 >> >years or so, they may reach the point where we were in the early 1980s. >> >[OK, I'm an optimist.] >> >> Perhaps not (being the optimist). :-) One of the reasons it's >> so damned important to get those emulators out there is so people >> can have access to different designs that had different goals and >> tradeoffs. > >I agree with that, sentiment. However, I'd like to add that >real hardware (even non-functional) is a good idea. Of course. I have never said that hardware isn't important. I have said that the hardware is uninteresting when the software isn't running on it. > .. When >you see an ISA on paper you tend to think "Oh, jeez these >folks were on crack", until you see the actual hardware >constraints they were working with. Preserving the knowledge >of the hardware & production constraints is vitally important >to understanding why machines were designed a certain way. > >The much maligned VAX architecture is a great example of >shifting manufacturing techology... Just look at the >range covered by the 11/7[358]0 and the uVAX II. :) > >The 3[679]0 range is another classic example too. The reason I don't talk about the hardware is because I know nothing about it, know that I know nothing about it, and defer to those experts who do know something about it. :-) /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 27 Feb 03 09:37:26 GMT Organization: UltraNet Communications, Inc. Lines: 49 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> X-Trace: UmFuZG9tSVYDSU/ho/j7NemXH1foVhxu1mRXlKO5+tDn+D+w4aHu0zs2uSHT/6ba X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 27 Feb 2003 10:21:02 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed.media.kyoto-u.ac.jp!newsfeed.gol.com!feed1.news.rcn.net!rcn!207-172-216-169 Xref: chonsp.franklin.ch alt.folklore.computers:130374 alt.sys.pdp10:12599 In article , peter@taronga.com (Peter da Silva) wrote: >In article , wrote: >>I didn't write that clearly enough :-). I didn't mean there >>wasn't portable code (I learned how to code using cards). I >>meant that portable wasn't the religion du jour that it is >>now. > >I don't see that it's a bigger issue now than 10 or 20 years ago. You're talking about portable sources that, when built, will execute with expected behaviour. I'm encountering kiddies who execpt the EXEs to be portable. IOW, they don't know there's a difference between files that are in ASCII and files that are in machine language. The monitor group spent a lot of tradeoffs trying to be backwards compatible. So I do know a little bit about some compatibility issues. It took three monitor releases for TOPS-10 to "convert" the world from the old SAVe file format to EXE format. That conversion made file management a lot easier; what was given up was the ability to see if an executable was sharable and/or had a high seg by just doing a DIR of the disk area. > >The expression "all the world's a vax" is about that old. That's >a term of derision, by the way, applied to people who write code >that can only be compiled on a vax because it assumes too much >about the hardware. > >Writing portable code was, if anything, a bigger deal back in the >'70s than it is now: back then you might be running on a 1s complement >36-bit word-oriented computer with 6-bit FEILDDATA characters this >week, and a 2s-complement 32-bit byte-oriented computer with 8-bit >EBCDIC next week, and 36-bit 2s-complement word-oriented computer >with 7-bit ASCII packed 5 characters to a word the week after. If the program had been written in a higher level language, the library and OTS would have taken care of the portability for you. That used to be one of the design decisions of a project. It just seems strange to me that people use a HLL and then proceed to do bitty-bytey programming. /BAH /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 27 Feb 03 10:40:31 GMT Organization: UltraNet Communications, Inc. Lines: 19 Message-ID: References: <1281.186T405T7773222@kltpzyxm.invalid> <20030226200030.21366.00002427@mb-mu.aol.com> X-Trace: UmFuZG9tSVYntjjr+p3EqsPD1w0+Xn0QwEgj/vXXRUl08sFKIu8hZBFS7eLI/PJY X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 27 Feb 2003 11:24:05 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.stanford.edu!logbridge.uoregon.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-169 Xref: chonsp.franklin.ch alt.folklore.computers:130408 In article <20030226200030.21366.00002427@mb-mu.aol.com>, rsclient@aol.com (Rsclient) wrote: >In article <1281.186T405T7773222@kltpzyxm.invalid>, "Charlie Gibbs" > writes: > >> >>Ah, but Barb said "portable _as it's used today_" (emphasis mine). >>Nowadays, "portable" means "runs on both Windows 95 and NT". :-p >> >>-- > >Sadly, nowadays it means runs on Win XP, Win ME, but not Win98, it's not >supported any more :-( That may be the attitude, but it's not how the kiddies think it. /BAH Subtract a hundred and four for e-mail. ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 26 Feb 2003 00:05:19 GMT Organization: TSS Inc. Lines: 21 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5BC109.2050003@jetnet.ab.ca> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046217919 79567 10.0.0.43 (26 Feb 2003 00:05:19 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 26 Feb 2003 00:05:19 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!enews.sgi.com!news.spies.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130495 alt.sys.pdp10:12620 In article , Mark Crispin wrote: >On Tue, 25 Feb 2003 bfranchuk@jetnet.ab.ca wrote: >> But I don't see software design much improved at all, from both >> a user and super-user point of view since the 70's. >Don't say that. The Linux world is slowly re-discovering software design >techniques which were well-known to us in the PDP-10 world. In another 10 >years or so, they may reach the point where we were in the early 1980s. >[OK, I'm an optimist.] >At least we're no longer hearing "we can't add that to UNIX, since the >code to do it won't fit in a 32K PDP-11" any more. Pity. There's already too much 32K-here 64K-there bloat. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: ararghNOSPAM@NOT.AT.enteract.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 25 Feb 2003 18:56:25 -0600 Organization: Not Really! Lines: 17 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> NNTP-Posting-Host: tcr196.dynip.ripco.net Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: e250.ripco.com 1046221016 4839 209.100.226.196 (26 Feb 2003 00:56:56 GMT) X-Complaints-To: usenet@ripco.com NNTP-Posting-Date: Wed, 26 Feb 2003 00:56:56 +0000 (UTC) X-Newsreader: Forte Agent 1.92/32.572 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.tpinternet.pl!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!out.nntp.be!propagator2-SanJose!in.nntp.be!gail.ripco.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130538 alt.sys.pdp10:12636 On 25 Feb 2003 20:12:45 GMT, peter@taronga.com (Peter da Silva) wrote: >The OS/MVS utility IEFBR14 was one instruction long, and required four >revisions before it was shipped, and a year later there was a new bug >reported in it. IEFBR14 dates from the OS/360 days. I know that it was in existance as early as 1970, because I used it then. And, just to annoy the operators, I wrote a utility called IEFBR15. Didn't use it much. :-) -- Arargh (at arargh dot com) http://www.arargh.com To reply by email, change the domain name, and remove the garbage. (Enteract can keep the spam, they are gone anyway) ###### From: "Glen Herrmannsfeldt" Newsgroups: alt.folklore.computers,alt.sys.pdp10 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Lines: 24 X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.00.2919.6700 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6700 Message-ID: NNTP-Posting-Host: 12.207.204.17 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc51.ops.asp.att.net 1046241384 12.207.204.17 (Wed, 26 Feb 2003 06:36:24 GMT) NNTP-Posting-Date: Wed, 26 Feb 2003 06:36:24 GMT Organization: AT&T Broadband Date: Wed, 26 Feb 2003 06:36:24 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!chcgil2-snh1.gtei.net!news.gtei.net!wn12feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!rwcrnsc51.ops.asp.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130546 alt.sys.pdp10:12637 wrote in message news:hm3o5vkk6kk41sk28bjlq7uemehkaoop14@4ax.com... > On 25 Feb 2003 20:12:45 GMT, peter@taronga.com (Peter da Silva) wrote: > > > >The OS/MVS utility IEFBR14 was one instruction long, and required four > >revisions before it was shipped, and a year later there was a new bug > >reported in it. > IEFBR14 dates from the OS/360 days. I know that it was in existance > as early as 1970, because I used it then. > > And, just to annoy the operators, I wrote a utility called IEFBR15. > Didn't use it much. :-) Not quite IEFBR15, but this came up recently in another newsgroup: SR 0,0 BALR 1,0 BCTR 0,1 -- glen ###### From: stanb45@dial.pipex.com (Stan Barr) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> <3E5C026F.5080708@jetnet.ab.ca> Organization: Metropolis Grafix Reply-To: stanb45@dial.pipex.com Message-ID: X-Newsreader: slrn (0.9.5.2 UNIX) Date: 26 Feb 2003 08:38:13 GMT Lines: 20 NNTP-Posting-Host: userbf33.uk.uudial.com X-Trace: 1046248693 news.dial.pipex.com 5372 62.188.142.54 X-Complaints-To: abuse@uk.uu.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news-FFM2.ecrc.net!news-mue1.dfn.de!news-koe1.dfn.de!uucp.muenster.de!dnewsfeed03.dtm.ops.eu.uu.net!dnewsifeed01.dtm.ops.eu.uu.net!lnewsifeed03.lnd.ops.eu.uu.net!lnewspost00.lnd.ops.eu.uu.net!emea.uu.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130391 alt.sys.pdp10:12603 On Tue, 25 Feb 2003 16:55:27 -0700, bfranchuk@jetnet.ab.ca wrote: >Mark Crispin wrote: > >> At least we're no longer hearing "we can't add that to UNIX, since the >> code to do it won't fit in a 32K PDP-11" any more. > >The real problem is fitting the linux kernal bootstrap because of the >640k limit on PC's. THANK YOU IBM! THANK YOU INTEL! Then run it on a Mac, as I've just started doing :-) (Older Macs that can't run OSX are getting pretty cheap these days...) -- Cheers, Stan Barr stanb45@dial.pipex.com **Remove the digits from email address** The future was never like this! ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5BC109.2050003@jetnet.ab.ca> <3E5C026F.5080708@jetnet.ab.ca> From: Morten Reistad X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: mrr@acer.reistad.priv.no (Morten Reistad) Message-ID: <6k4i3b.hhc2.ln@via.reistad.priv.no> Lines: 28 Date: Wed, 26 Feb 2003 11:24:06 +0100 NNTP-Posting-Host: 80.111.42.204 X-Complaints-To: abuse@chello.no X-Trace: amstwist00 1046257201 80.111.42.204 (Wed, 26 Feb 2003 12:00:01 MET) NNTP-Posting-Date: Wed, 26 Feb 2003 12:00:01 MET Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!amsnews01.chello.com!amstwist00.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130470 alt.sys.pdp10:12614 According to : >Mark Crispin wrote: > >> At least we're no longer hearing "we can't add that to UNIX, since the >> code to do it won't fit in a 32K PDP-11" any more. > >The real problem is fitting the linux kernal bootstrap because of the >640k limit on PC's. THANK YOU IBM! THANK YOU INTEL! >Ben. My crystal ball keep telling me that next year, the linux folks will add another stage to the bootstrap. But they never listen to the crystal ball. Seriously: I saw this as a hack even in 1994, when I first installed Linux. The level of shoehorning they have done with the 640k boot image size is just amazing. Go compare the bootstrap code from e.g. FreeBSD and Linux, and you will see. Besides, Linux is for heretics. It is not the one, true unix!. -- mrr ###### User-Agent: Microsoft-Entourage/10.1.1.2418 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) From: nospam Newsgroups: alt.folklore.computers,alt.sys.pdp10 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> Mime-version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit Lines: 23 Date: Wed, 26 Feb 2003 10:41:15 GMT NNTP-Posting-Host: 144.136.91.243 X-Complaints-To: news@bigpond.net.au X-Trace: news-server.bigpond.net.au 1046256075 144.136.91.243 (Wed, 26 Feb 2003 21:41:15 EST) NNTP-Posting-Date: Wed, 26 Feb 2003 21:41:15 EST Organization: BigPond Internet Services (http://www.bigpond.net.au) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.stanford.edu!logbridge.uoregon.edu!arclight.uoregon.edu!enews.sgi.com!news-out.spamkiller.net!propagator2-maxim!news-in.spamkiller.net!lon-transit.news.telstra.net!news.telstra.net!news-server.bigpond.net.au!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130506 alt.sys.pdp10:12628 in article Pine.LNX.4.55.0302251305540.16246@shiva1.cac.washington.edu, Mark Crispin at mrc@CAC.Washington.EDU wrote on 26/02/2003 08:08: > On Tue, 25 Feb 2003 bfranchuk@jetnet.ab.ca wrote: > [OK, I'm an optimist.] I wouldn't get too excited;) > > At least we're no longer hearing "we can't add that to UNIX, since the > code to do it won't fit in a 32K PDP-11" any more. I actually think that would be a good thing for linux, its already too bloated even for somebody that seems to like UNIX. An OS full of good ideas, unfortunately, every bodies good ideas - all at once, all welcome non rejected ;) > > -- Mark -- > > http://staff.washington.edu/mrc > Science does not emerge from voting, party politics, or public debate. ###### Message-ID: <3E5CA03F.28C8047E@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> <3E5C026F.5080708@jetnet.ab.ca> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 18 Date: Wed, 26 Feb 2003 11:16:57 GMT NNTP-Posting-Host: 12.90.169.110 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc05-news.ops.worldnet.att.net 1046258217 12.90.169.110 (Wed, 26 Feb 2003 11:16:57 GMT) NNTP-Posting-Date: Wed, 26 Feb 2003 11:16:57 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!news-xfer.cox.net!cox.net!cyclone1.gnilink.net!wn14feed!wn13feed!worldnet.att.net!bgtnsc05-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130315 alt.sys.pdp10:12591 bfranchuk@jetnet.ab.ca wrote: > Mark Crispin wrote: > > > At least we're no longer hearing "we can't add that to UNIX, > > since the code to do it won't fit in a 32K PDP-11" any more. > > The real problem is fitting the linux kernal bootstrap because > of the 640k limit on PC's. THANK YOU IBM! THANK YOU INTEL! What's the problem? You can't fit a disk reader to load into the rest of memory into 640k? The real problem is kernals that are too big to fit onto a floppy disk. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 26 Feb 2003 13:25:03 -0000 Organization: Titanic Enterprises Unlimited Lines: 36 Message-ID: <1046265903.251131@saucer.planet.gong> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1046265904 22142 80.177.7.220 (26 Feb 2003 13:25:04 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Wed, 26 Feb 2003 13:25:04 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Cache: nntpcache 3.0.1 (see http://www.nntpcache.org/) X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Cache-Post-Path: saucer.planet.gong!unknown@voodoo.planet.gong Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!news.gtei.net!newsfeed.mathworks.com!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130434 alt.sys.pdp10:12607 wrote in message news:b3i89s$i0m$5@bob.news.rcn.net... > In article , > Mark Crispin wrote: > >On Tue, 25 Feb 2003 bfranchuk@jetnet.ab.ca wrote: > >> But I don't see software design much improved at all, from both > >> a user and super-user point of view since the 70's. > > > >Don't say that. The Linux world is slowly re-discovering software design > >techniques which were well-known to us in the PDP-10 world. In another 10 > >years or so, they may reach the point where we were in the early 1980s. > >[OK, I'm an optimist.] > > Perhaps not (being the optimist). :-) One of the reasons it's > so damned important to get those emulators out there is so people > can have access to different designs that had different goals and > tradeoffs. I agree with that, sentiment. However, I'd like to add that real hardware (even non-functional) is a good idea. When you see an ISA on paper you tend to think "Oh, jeez these folks were on crack", until you see the actual hardware constraints they were working with. Preserving the knowledge of the hardware & production constraints is vitally important to understanding why machines were designed a certain way. The much maligned VAX architecture is a great example of shifting manufacturing techology... Just look at the range covered by the 11/7[358]0 and the uVAX II. :) The 3[679]0 range is another classic example too. Cheers, Rupert ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 26 Feb 2003 14:19:52 GMT Organization: TSS Inc. Lines: 33 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046269192 24511 10.0.0.43 (26 Feb 2003 14:19:52 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 26 Feb 2003 14:19:52 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!cyclone.bc.net!arclight.uoregon.edu!enews.sgi.com!news.spies.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130492 alt.sys.pdp10:12619 In article , wrote: >I didn't write that clearly enough :-). I didn't mean there >wasn't portable code (I learned how to code using cards). I >meant that portable wasn't the religion du jour that it is >now. I don't see that it's a bigger issue now than 10 or 20 years ago. The expression "all the world's a vax" is about that old. That's a term of derision, by the way, applied to people who write code that can only be compiled on a vax because it assumes too much about the hardware. Writing portable code was, if anything, a bigger deal back in the '70s than it is now: back then you might be running on a 1s complement 36-bit word-oriented computer with 6-bit FEILDDATA characters this week, and a 2s-complement 32-bit byte-oriented computer with 8-bit EBCDIC next week, and 36-bit 2s-complement word-oriented computer with 7-bit ASCII packed 5 characters to a word the week after. These days, as someone pointed out, if it runs on Windows 98 and Windows XP you're doing pretty well. Even if you care about real portability, you can assume 32 bit words (64 bit if you're really aggressive), 2s-complement, byte-addressed, 8-bit ASCII, and anything but little-endian is "legacy". And every current OS is basically derived from a UNIX variant or has had one grafted onto it, and your code is almost certainly written in a C derivitive. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 26 Feb 2003 14:24:15 GMT Organization: TSS Inc. Lines: 18 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046269455 24806 10.0.0.43 (26 Feb 2003 14:24:15 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 26 Feb 2003 14:24:15 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!eusc.inter.net!news.tele.dk!news.tele.dk!small.news.tele.dk!enews.sgi.com!news.spies.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130482 alt.sys.pdp10:12617 In article <6k4i3b.hhc2.ln@via.reistad.priv.no>, Morten Reistad wrote: >Seriously: I saw this as a hack even in 1994, when I first installed >Linux. The level of shoehorning they have done with the 640k boot image >size is just amazing. In 1994 I was still running a couple of Xenix boxes to support customers on legacy systems. The minimum memory requirement to boot Xenix-286 to multiuser was 192K. If someone had told me that you couldn't fit a boot loader into 640K I'd have told them they were nuts. What the hell is it doing? -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: pechter@shell.monmouth.com (Bill/Carolyn Pechter) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 26 Feb 2003 10:10:08 -0500 Organization: Lakewood MicroSystems Lines: 39 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> NNTP-Posting-Host: shell.monmouth.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!news.tele.dk!news.tele.dk!small.news.tele.dk!newspeer.monmouth.com!news.monmouth.com!shell.monmouth.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130451 alt.sys.pdp10:12608 In article <6k4i3b.hhc2.ln@via.reistad.priv.no>, Morten Reistad wrote: >My crystal ball keep telling me that next year, the linux folks >will add another stage to the bootstrap. But they never listen to >the crystal ball. > >Seriously: I saw this as a hack even in 1994, when I first installed >Linux. The level of shoehorning they have done with the 640k boot image >size is just amazing. > >Go compare the bootstrap code from e.g. FreeBSD and Linux, and you >will see. > >Besides, Linux is for heretics. It is not the one, true unix!. > >-- mrr > Very, very true... But I still haven't got FreeBSD running Simh's ethernet support for the VAX emulation... The house linux box does it fine. I'm thinking of punching a hole in my house firewall and letting it get routed to the VAX emulation under linux... How slick would that be... Be real slick if the VAX emulation could run in a FreeBSD jail... HACK IN TO THIS 8-)... most of the script kiddies couldn't log in on a VAX... Heck, this should be a project worth doing. An emulated VAX/VMS Honeypot. Looks like it works on NetBSD and OpenBSD, however. -- +---------------------------------------------------------------------------+ | Bill and/or Carolyn Pechter | pechter@shell.monmouth.com | | Bill Gates is a Persian cat and a monocle away from being a villain in | | a James Bond movie -- Dennis Miller | +---------------------------------------------------------------------------+ ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Organization: Honest Chris' Sysadmin Emporium X-Newsreader: trn 4.0-test76 (Apr 2, 2001) From: cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) Originator: cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) Date: Wed, 26 Feb 2003 15:50:46 +0000 Message-ID: Lines: 16 NNTP-Posting-Date: 26 Feb 2003 16:11:01 GMT NNTP-Posting-Host: 62.254.134.35 X-Trace: 1046275861 news.gradwell.net 66678 cbh/62.254.134.35 X-Complaints-To: news-abuse@gradwell.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!195.149.39.56.MISMATCH!news-peer-lilac.gradwell.net!news-peer.gradwell.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130381 alt.sys.pdp10:12601 According to Bill/Carolyn Pechter : > Be real slick if the VAX emulation could run in a FreeBSD jail... > HACK IN TO THIS 8-)... most of the script kiddies couldn't log in on a > VAX... Heck, this should be a project worth doing. An emulated VAX/VMS > Honeypot. I'm not sure about that since VMS was one of the favourite targets for crackers in the '80s. Lots of the UK's universities had loads of problems with some gits coming in over an X.25 pad and causing mayhem with their VMS boxes, and the culprits are still out there... Chris. -- "If the world was an orange it would be like much too small, y'know?" Neil, '84 Currently playing: The Stranglers - "The Raven" http://www.chrishedley.com My stuff, including genealogy, other things, etc ###### From: bfranchuk@jetnet.ab.ca Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 26 Feb 2003 10:13:02 -0700 Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3E5CF59E.2070409@jetnet.ab.ca> User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.0.1) Gecko/20021005 X-Accept-Language: en-us, en MIME-Version: 1.0 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> <3E5C026F.5080708@jetnet.ab.ca> <3E5CA03F.28C8047E@yahoo.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 26 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!sn-xit-02!sn-xit-04!sn-xit-06!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130348 alt.sys.pdp10:12592 CBFalconer wrote: > bfranchuk@jetnet.ab.ca wrote: > >>Mark Crispin wrote: >> >> >>>At least we're no longer hearing "we can't add that to UNIX, >>>since the code to do it won't fit in a 32K PDP-11" any more. >> >>The real problem is fitting the linux kernal bootstrap because >>of the 640k limit on PC's. THANK YOU IBM! THANK YOU INTEL! > > > What's the problem? You can't fit a disk reader to load into the > rest of memory into 640k? The real problem is kernals that are > too big to fit onto a floppy disk. > Well they now have gone to 2 disk floppy boots. The default linux boot was wrote years ago when all you had was the minux file system and and floppy disk and small HD as i/o. FreeBSD has a better bootstrap. I still like the idea of system builder/linker to build a bootstrap rather than having to recompile the kernal or a dynamic library. Ben. ###### From: Mark Crispin Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 26 Feb 2003 10:10:39 -0800 Organization: Networks and Distributed Computing Lines: 43 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: shiva0.cac.washington.edu Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Trace: nntp1.u.washington.edu 1046283042 20688 (None) 140.142.17.40 X-Complaints-To: help@cac.washington.edu In-Reply-To: Content-Length: 314159 (believe this at your own risk) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!headwall.stanford.edu!newsfeed.stanford.edu!logbridge.uoregon.edu!news.u.washington.edu!140.142.17.34.MISMATCH!news.u.washington.edu!shiva0.cac.washington.edu!mrc Xref: chonsp.franklin.ch alt.folklore.computers:130365 alt.sys.pdp10:12595 On Wed, 26 Feb 2003, Peter da Silva wrote: > Writing portable code was, if anything, a bigger deal back in the > '70s than it is now Tell me about it! I could write a book about the vicious flames that I have received because the IMAP toolkit is aggressively portable. You have to go back to the days before ftruncate() (or chsize()) before you find a variant of UNIX that you can't build it on -- and even then, that only impacts the local file drivers, the IMAP/POP3/NNTP/SMTP support still works fine. It builds under Win32 (all variants), VAX/VMS, and TOPS-20. At one time, it built under DOS and Mac OS; don't know if it still does that (in particular, the 32K segment sizes in Mac OS may be a problem). But to some people, it is offensive if code is not written exclusively for the newest and most gonzo version of Linux. The mere fact of code for other platforms enrages them, much less *compromises* in order to be portable to both. > These days, as someone pointed out, if it runs on Windows 98 and > Windows XP you're doing pretty well. Thank you. > Even if you care about real > portability, you can assume 32 bit words (64 bit if you're really > aggressive), 2s-complement, byte-addressed, 8-bit ASCII, and anything > but little-endian is "legacy". Well, I know that XKL succeeded in porting GNU EMACS to TOPS-20. Haven't used it much; one of the delights of using TOPS-20 is to be able to use real EMACS again. I remember Richard Stallman saying that he had no intention of writing code that would port to other than byte-addressed platforms, so XKL must have done a lot of bug-fixing. Speaking of GNU EMACS, does anyone know how to turn off that damn "DOS mode" in the newer Linux versions? It's infuriated me to the point that I was going to switch to vi, only I discovered that vi also has that hack. -- Mark -- http://staff.washington.edu/mrc Science does not emerge from voting, party politics, or public debate. ###### Message-ID: <3E5D32A5.89915AB7@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <1281.186T405T7773222@kltpzyxm.invalid> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 21 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: sccrnsc02 1046288104 12.241.15.59 (Wed, 26 Feb 2003 19:35:04 GMT) NNTP-Posting-Date: Wed, 26 Feb 2003 19:35:04 GMT Date: Wed, 26 Feb 2003 19:35:04 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!out.nntp.be!propagator2-SanJose!in.nntp.be!newsfeed1.easynews.com!easynews.com!easynews!cyclone1.gnilink.net!wn14feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!sccrnsc02.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130532 alt.sys.pdp10:12633 Peter da Silva wrote: > > Charlie Gibbs wrote: > [IEFBR14] > > The fix required a doubling of the instruction count. :-) > > It required increasing it by a factor of 6, once they got the name in. > > > Ah, but Barb said "portable _as it's used today_" (emphasis mine). > > Nowadays, "portable" means "runs on both Windows 95 and NT". :-p > > You are a nasty cynical man. I can't help but approve. > "The power of accurate observation is commonly called cynicism by those who have not got it." --George Bernard Shaw -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### Message-ID: <3E5D358F.92831F8@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BED2B.C90073B6@ev1.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 57 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: sccrnsc03 1046288850 12.241.15.59 (Wed, 26 Feb 2003 19:47:30 GMT) NNTP-Posting-Date: Wed, 26 Feb 2003 19:47:30 GMT Date: Wed, 26 Feb 2003 19:47:30 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!cyclone.bc.net!logbridge.uoregon.edu!nntp-server.caltech.edu!attla2!ip.att.net!attbi_feed3!attbi_feed4!attbi.com!sccrnsc03.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130522 alt.sys.pdp10:12631 jmfbahciv@aol.com wrote: > > In article <3E5BED2B.C90073B6@ev1.net>, > Charles Richmond wrote: > >jmfbahciv@aol.com wrote: > >> > >> [snip...] [snip...] [sinp...] > >> > >> So, this a very roundabout way to tell you, I don't think I can > >> answer your question. > >> > >CIRCUMLOCUTION,* n. A literary trick whereby the writer > >who has nothing to say breaks it gently to the reader. > > --Ambroise Bierce, _The Devil's Dictionary > > > Smarty pants. I started out thinking I was going to > answer the question. I finished up realizing I didn't answer > the question. However, there was enough useful stuff in what > I wrote that I had to acknowledge it wasn't an answer. So, > rather than delete everything, I added the acknowledgement. > Granted I should have written the "I don't know." at the > beginning so that he had a choice of not wading through the > text. > The word "roundabout" brought up "circumlocution" for me. You might enjoy reading _The Devil's Dictionary_...I find it a useful and entertaining commentary on things. A couple of other definitions I particularly enjoy: Christian: n. One who believes that the New Testament is a divinely inspired book admirably suited to the spiritual needs of his neighbor. One who follows the teachings of Christ in so far as they are not inconsistent with a life of sin. idiot: n. A member of a large and powerful tribe whose influence in human affairs has always been dominant and controlling. The Idiot's activity is not confined to any special field of thought or action, but "pervades and regulates the whole." He has the last word in everything; his decision is unappealable. He sets the fashions and opinion of taste, dictates the limitations of speech and circumscribes conduct with a dead-line. Much of _The Devil's Dictionary_ is online at: and you can download a copy at: -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### Message-ID: <3E5D1B07.63890CF4@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 18 Date: Wed, 26 Feb 2003 20:05:06 GMT NNTP-Posting-Host: 12.90.176.136 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc05-news.ops.worldnet.att.net 1046289906 12.90.176.136 (Wed, 26 Feb 2003 20:05:06 GMT) NNTP-Posting-Date: Wed, 26 Feb 2003 20:05:06 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!chcgil2-snh1.gtei.net!news.gtei.net!wn12feed!worldnet.att.net!bgtnsc05-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130313 alt.sys.pdp10:12590 Mark Crispin wrote: > ... snip ... > > Speaking of GNU EMACS, does anyone know how to turn off that damn "DOS > mode" in the newer Linux versions? It's infuriated me to the point that I > was going to switch to vi, only I discovered that vi also has that hack. Guessing here, but maybe it just discovered that the input file had line endings, and put itself in a mode to preserve that characteristic. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> From: Morten Reistad X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: mrr@acer.reistad.priv.no (Morten Reistad) Message-ID: Lines: 38 Date: Wed, 26 Feb 2003 21:55:48 +0100 NNTP-Posting-Host: 80.111.42.204 X-Complaints-To: abuse@chello.no X-Trace: amstwist00 1046293201 80.111.42.204 (Wed, 26 Feb 2003 22:00:01 MET) NNTP-Posting-Date: Wed, 26 Feb 2003 22:00:01 MET Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news2.euro.net!amsnews01.chello.com!amstwist00.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130476 alt.sys.pdp10:12615 According to Peter da Silva : >In article <6k4i3b.hhc2.ln@via.reistad.priv.no>, >Morten Reistad wrote: >>Seriously: I saw this as a hack even in 1994, when I first installed >>Linux. The level of shoehorning they have done with the 640k boot image >>size is just amazing. > >In 1994 I was still running a couple of Xenix boxes to support customers >on legacy systems. The minimum memory requirement to boot Xenix-286 to >multiuser was 192K. If someone had told me that you couldn't fit a boot >loader into 640K I'd have told them they were nuts. > >What the hell is it doing? Linux does not have a real boot-loader per se. A real boot-loader is able to understand the rudiments of the root file system, and load the images it need before starting everything up. With some basic loader options for choosing different images, as appropriate. BSD, SunOS, Solaris, SGI, QNX and others all have variations on this theme. Linux, OTOH, has just an image that is loaded in real-mode, from a dedicated micro-partition, and that is it. The Linux kernel has to be in there, all of it. The S.K. Boot-loader is just a page sized hack that reads a block chain from raw disk and stuffs it into memory. So, 640k (actually, 632 I think) is the limit for how much stuff can be loaded in the OS boot. They have had to use a compressed kernel, that loads, builds the rudiments of page tables, flips to '386 mode and decompresses the kernel before starting it. Therefore, MS-DOS is one of the things Linux is sorely missing; a good boot-loader. -- mrr ###### From: Mark Crispin Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 26 Feb 2003 13:05:10 -0800 Organization: Networks and Distributed Computing Lines: 20 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D1B07.63890CF4@yahoo.com> NNTP-Posting-Host: shiva0.cac.washington.edu Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Trace: nntp1.u.washington.edu 1046293512 19294 (None) 140.142.17.38 X-Complaints-To: help@cac.washington.edu In-Reply-To: <3E5D1B07.63890CF4@yahoo.com> Content-Length: 314159 (believe this at your own risk) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!cyclone.bc.net!logbridge.uoregon.edu!news.u.washington.edu!140.142.17.34.MISMATCH!news.u.washington.edu!shiva0.cac.washington.edu!mrc Xref: chonsp.franklin.ch alt.folklore.computers:130358 alt.sys.pdp10:12594 On Wed, 26 Feb 2003, CBFalconer wrote: > > Speaking of GNU EMACS, does anyone know how to turn off that damn "DOS > > mode" in the newer Linux versions? It's infuriated me to the point that I > > was going to switch to vi, only I discovered that vi also has that hack. > > Guessing here, but maybe it just discovered that the input file > had line endings, and put itself in a mode to preserve > that characteristic. Indeed, that's what it does. I don't want it to do that. Once upon a time, there was a notion that a user could say "I know what I'm doing, damnit" and instruct software not to try to protect the user from himself. This was the single aspect of the "UNIX philosophy" that I agreed with. -- Mark -- http://staff.washington.edu/mrc Science does not emerge from voting, party politics, or public debate. ###### Sender: eric@ruckus.brouhaha.com From: Eric Smith Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Organization: Eric Conspiracy Secret Labs X-Eric-Conspiracy: There is no conspiracy. Date: 26 Feb 2003 14:58:38 -0800 Message-ID: Lines: 22 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-Host: 209.66.107.17 X-Trace: 26 Feb 2003 15:44:12 -0800, 209.66.107.17 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!24.215.0.32.MISMATCH!newsfeed.mountaincable.net!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!news.he.net!news.kjsl.com!news.spies.com!209.66.107.17 Xref: chonsp.franklin.ch alt.folklore.computers:130504 alt.sys.pdp10:12627 Morten Reistad writes: > Linux does not have a real boot-loader per se. > > A real boot-loader is able to understand the rudiments of the > root file system, and load the images it need before starting > everything up. With some basic loader options for choosing > different images, as appropriate. BSD, SunOS, Solaris, SGI, QNX > and others all have variations on this theme. Some Linux/x86 systems now use GRUB (instead of LILO), which does what you describe. Most Linux/Alpha systems use MILO, which also does that. MILO is built using parts of the kernel source tree, so it actually has the full-blown ext2 file system, and possibly others as well. In principle there's no reason why MILO couldn't be ported to the x86 or other processors. Linux/MIPS also has a bootloader called MILO, but it's not based on the Linux/Alpha MILO. I've never used it so I don't know how sophisticated it is. ###### Message-ID: <3E5D4816.C974A441@yahoo.com> From: Peter Flass X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <1281.186T405T7773222@kltpzyxm.invalid> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 16 Date: Wed, 26 Feb 2003 23:07:23 GMT NNTP-Posting-Host: 24.194.50.82 X-Complaints-To: abuse@rr.com X-Trace: twister.nyroc.rr.com 1046300843 24.194.50.82 (Wed, 26 Feb 2003 18:07:23 EST) NNTP-Posting-Date: Wed, 26 Feb 2003 18:07:23 EST Organization: Road Runner Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!news-west.rr.com!cyclone.nyroc.rr.com!cyclone-out.nyroc.rr.com!twister.nyroc.rr.com.POSTED!53ab2750!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130557 alt.sys.pdp10:12639 Charlie Gibbs wrote: [snip] > >In the '70s? It sure did. FORTRAN was the language of choice for > >portable code back then, but the first OS port I know of started > >in 1976, and that was in C. > > Ah, but Barb said "portable _as it's used today_" (emphasis mine). > Nowadays, "portable" means "runs on both Windows 95 and NT". :-p Actually, Multics was designed to be portable, and in fact *was* ported from a GE-645 to a Honeywell ... something, a similar machine, but with some significant differences. The only reason it wasn't more widely ported is that most current hardware lacks the support for segmentation and rings that it requires. As I believe I mentioned before, the only reason unix has been so widely ported is that it only uses the least common denominator in hardware. ###### Lines: 13 X-Admin: news@aol.com From: rsclient@aol.com (Rsclient) Newsgroups: alt.folklore.computers Date: 27 Feb 2003 01:00:30 GMT References: <1281.186T405T7773222@kltpzyxm.invalid> Organization: AOL http://www.aol.com X-Newsreader: Session Scheduler (Queue Name: gng-mu) Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <20030226200030.21366.00002427@mb-mu.aol.com> Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!feedme.news.mediaways.net!fu-berlin.de!cox.net!cyclone1.gnilink.net!ngpeer.news.aol.com!audrey-m2.news.aol.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130431 In article <1281.186T405T7773222@kltpzyxm.invalid>, "Charlie Gibbs" writes: > >Ah, but Barb said "portable _as it's used today_" (emphasis mine). >Nowadays, "portable" means "runs on both Windows 95 and NT". :-p > >-- Sadly, nowadays it means runs on Win XP, Win ME, but not Win98, it's not supported any more :-( Peter ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 27 Feb 2003 01:04:10 GMT Organization: TSS Inc. Lines: 26 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <6k4i3b.hhc2.ln@via.reistad.priv.no> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046307850 59230 10.0.0.43 (27 Feb 2003 01:04:10 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 27 Feb 2003 01:04:10 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!c03.atl99!news.webusenet.com!telocity-west!DIRECTV!204.94.211.44!enews.sgi.com!news.spies.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130498 alt.sys.pdp10:12623 In article , Morten Reistad wrote: >A real boot-loader is able to understand the rudiments of the >root file system, and load the images it need before starting >everything up. I thought that was the whole point of LILO. >So, 640k (actually, 632 I think) is the limit for how much stuff can >be loaded in the OS boot. They have had to use a compressed kernel, >that loads, builds the rudiments of page tables, flips to '386 mode >and decompresses the kernel before starting it. Geeze. I thought they had it pretty much all in modules by now. I mean, it was back when Red Hat was at 4.1 that they went to modules for the install disk, which would imply to me that they were mature enough to use. Obviously I've been drinking the good koolaid. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 27 Feb 2003 01:11:06 GMT Organization: TSS Inc. Lines: 13 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046308266 59813 10.0.0.43 (27 Feb 2003 01:11:06 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 27 Feb 2003 01:11:06 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!eusc.inter.net!news.tele.dk!news.tele.dk!small.news.tele.dk!enews.sgi.com!news.spies.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130496 alt.sys.pdp10:12621 In article , Mark Crispin wrote: >Speaking of GNU EMACS, does anyone know how to turn off that damn "DOS >mode" in the newer Linux versions? It's infuriated me to the point that I >was going to switch to vi, only I discovered that vi also has that hack. "DOS MODE"? -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 27 Feb 2003 01:15:03 GMT Organization: TSS Inc. Lines: 36 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D1B07.63890CF4@yahoo.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046308503 59909 10.0.0.43 (27 Feb 2003 01:15:03 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 27 Feb 2003 01:15:03 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130497 alt.sys.pdp10:12622 In article , Mark Crispin wrote: >On Wed, 26 Feb 2003, CBFalconer wrote: >> > Speaking of GNU EMACS, does anyone know how to turn off that damn "DOS >> > mode" in the newer Linux versions? It's infuriated me to the point that I >> > was going to switch to vi, only I discovered that vi also has that hack. >> Guessing here, but maybe it just discovered that the input file >> had line endings, and put itself in a mode to preserve >> that characteristic. >Indeed, that's what it does. I don't want it to do that. Since when does vi do that? You must be running one of those vi clones there instead of the real thing (either version). But in any case, don't you need that to remain maximally compatible with TOPS-20? (at least that's the first OS *I* used where programs expected in files) >Once upon a time, there was a notion that a user could say "I know what >I'm doing, damnit" and instruct software not to try to protect the user >from himself. This was the single aspect of the "UNIX philosophy" that I >agreed with. You're confusing (I suspect) vim or elvis with UNIX software. They're both ports of PC clones of VI back to UNIX. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### Message-ID: <3E5D6393.6B738216@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D1B07.63890CF4@yahoo.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 29 Date: Thu, 27 Feb 2003 01:57:03 GMT NNTP-Posting-Host: 12.90.176.232 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc04-news.ops.worldnet.att.net 1046311023 12.90.176.232 (Thu, 27 Feb 2003 01:57:03 GMT) NNTP-Posting-Date: Thu, 27 Feb 2003 01:57:03 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!chcgil2-snh1.gtei.net!news.gtei.net!wn12feed!worldnet.att.net!bgtnsc04-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130556 alt.sys.pdp10:12638 Mark Crispin wrote: > > On Wed, 26 Feb 2003, CBFalconer wrote: > > > Speaking of GNU EMACS, does anyone know how to turn off that > > > damn "DOS mode" in the newer Linux versions? It's infuriated > > > me to the point that I was going to switch to vi, only I > > > discovered that vi also has that hack. > > > > Guessing here, but maybe it just discovered that the input file > > had line endings, and put itself in a mode to preserve > > that characteristic. > > Indeed, that's what it does. I don't want it to do that. > > Once upon a time, there was a notion that a user could say "I > know what I'm doing, damnit" and instruct software not to try to > protect the user from himself. This was the single aspect of the > "UNIX philosophy" that I agreed with. So? It's preserving characteristics of the file. If you don't want that, pass the file through a filter to convert line endings first. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### From: Mark Crispin Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 26 Feb 2003 20:21:40 -0800 Organization: Networks and Distributed Computing Lines: 23 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D1B07.63890CF4@yahoo.com> NNTP-Posting-Host: shiva0.cac.washington.edu Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Trace: nntp1.u.washington.edu 1046319702 15488 (None) 140.142.17.38 X-Complaints-To: help@cac.washington.edu In-Reply-To: Content-Length: 314159 (believe this at your own risk) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!cyclone.bc.net!logbridge.uoregon.edu!news.u.washington.edu!140.142.17.34.MISMATCH!news.u.washington.edu!shiva0.cac.washington.edu!mrc Xref: chonsp.franklin.ch alt.folklore.computers:130366 alt.sys.pdp10:12596 On Wed, 27 Feb 2003, Peter da Silva wrote: > Since when does vi do that? You must be running one of those vi > clones there instead of the real thing (either version). Indeed, it appears that what is installed as "vi" on this Linux system is actually vim. > But in any case, don't you need that to remain maximally compatible > with TOPS-20? No. In TOPS-20 EMACS, you could still get at the CR and LF separately even though ^M was normally not displayed. Also, TOPS-20 EMACS didn't go and "fix" newlines in binary files either. This stupid stupid new version of GNU EMACS does not have that ability. Fortunately, I was just emailed the incantation to turn that behavior off: (setq inhibit-eol-conversion t) -- Mark -- http://staff.washington.edu/mrc Science does not emerge from voting, party politics, or public debate. ###### From: Mark Crispin Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 26 Feb 2003 20:28:59 -0800 Organization: Networks and Distributed Computing Lines: 17 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D1B07.63890CF4@yahoo.com> <3E5D6393.6B738216@yahoo.com> NNTP-Posting-Host: shiva0.cac.washington.edu Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Trace: nntp1.u.washington.edu 1046320142 14670 (None) 140.142.17.39 X-Complaints-To: help@cac.washington.edu In-Reply-To: <3E5D6393.6B738216@yahoo.com> Content-Length: 314159 (believe this at your own risk) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!cyclone.bc.net!logbridge.uoregon.edu!news.u.washington.edu!140.142.17.34.MISMATCH!news.u.washington.edu!shiva0.cac.washington.edu!mrc Xref: chonsp.franklin.ch alt.folklore.computers:130368 alt.sys.pdp10:12597 On Thu, 27 Feb 2003, CBFalconer wrote: > So? It's preserving characteristics of the file. If you don't > want that, pass the file through a filter to convert line endings > first. Once upon a time, EMACS was usable by default without requiring one to pass through filters first. That's the whole point. Why should I have to use a filter instead of using the EMACS commands that I have used for the past 25 years? Another objectionable "improvement" in this stupid new version of GNU EMACS is removing/changing ^X^P. -- Mark -- http://staff.washington.edu/mrc Science does not emerge from voting, party politics, or public debate. ###### Sender: phr2003-nospam@ruckus.brouhaha.com From: Paul Rubin Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D1B07.63890CF4@yahoo.com> Date: 26 Feb 2003 21:08:45 -0800 Message-ID: <7x1y1utin6.fsf@ruckus.brouhaha.com> Organization: Nightsong/Fort GNOX Lines: 10 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii NNTP-Posting-Host: 209.66.107.17 X-Trace: 26 Feb 2003 21:54:22 -0800, 209.66.107.17 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!newsfeed.icl.net!newsfeed.fjserv.net!news-out.newsfeeds.com!propagator2-maxim!news-in.spamkiller.net!news.he.net!news.kjsl.com!news.spies.com!209.66.107.17 Xref: chonsp.franklin.ch alt.folklore.computers:130499 alt.sys.pdp10:12624 Mark Crispin writes: > > Guessing here, but maybe it just discovered that the input file > > had line endings, and put itself in a mode to preserve > > that characteristic. > > Indeed, that's what it does. I don't want it to do that. You can use M-x find-file-literally to open a file without any conversions. There's also a way to set the default, but I don't remember it. ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: <6o9r5vc2ekhvfnku4cd86fkso6uejbc2n1@4ax.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D1B07.63890CF4@yahoo.com> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 25 Date: Thu, 27 Feb 2003 06:30:43 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news2.calgary.shaw.ca 1046327443 24.71.223.147 (Wed, 26 Feb 2003 23:30:43 MST) NNTP-Posting-Date: Wed, 26 Feb 2003 23:30:43 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!c03.atl99!sjc70.webusenet.com!news.webusenet.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news2.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130537 alt.sys.pdp10:12635 On Wed, 26 Feb 2003 20:21:40 -0800 in alt.folklore.computers, Mark Crispin wrote: >On Wed, 27 Feb 2003, Peter da Silva wrote: >> Since when does vi do that? You must be running one of those vi >> clones there instead of the real thing (either version). > >Indeed, it appears that what is installed as "vi" on this Linux system is >actually vim. You can add command line option -b or "set bin|binary" in EXINIT or .exrc to force unix type input without losing vi compatible mode: apparently all vim options are processed before vi compatible mode is set, whether that is by the command name, options on the command line, or in EXINIT, .exrc, VIMINIT, .vimrc. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> From: Morten Reistad X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: mrr@acer.reistad.priv.no (Morten Reistad) Message-ID: Lines: 34 Date: Thu, 27 Feb 2003 09:46:05 +0100 NNTP-Posting-Host: 80.111.42.204 X-Complaints-To: abuse@chello.no X-Trace: amstwist00 1046336401 80.111.42.204 (Thu, 27 Feb 2003 10:00:01 MET) NNTP-Posting-Date: Thu, 27 Feb 2003 10:00:01 MET Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!feedme.news.mediaways.net!itgate.net!news.it.colt.net!peernews3.colt.net!colt.net!news.maxwell.syr.edu!newsfeed-east.nntpserver.com!nntpserver.com!border1.nntp.aus1.giganews.com!nntp.giganews.com!amsnews01.chello.com!amstwist00.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130462 alt.sys.pdp10:12611 According to Peter da Silva : >In article , >Morten Reistad wrote: >>A real boot-loader is able to understand the rudiments of the >>root file system, and load the images it need before starting >>everything up. > >I thought that was the whole point of LILO. Lilo is just a loader, like we used to have in the old tape loaders. >>So, 640k (actually, 632 I think) is the limit for how much stuff can >>be loaded in the OS boot. They have had to use a compressed kernel, >>that loads, builds the rudiments of page tables, flips to '386 mode >>and decompresses the kernel before starting it. > >Geeze. > >I thought they had it pretty much all in modules by now. I mean, it was >back when Red Hat was at 4.1 that they went to modules for the install disk, >which would imply to me that they were mature enough to use. Of course, there is no room for more stuff in the kernel proper, so the overflow will have to go into modules, Leads to interesting scenarios if your root filesystem of choice is a rather large one, codewise; and is not the ext2 fs. (Like xfs or reiserfs), It IS possible to build an image small enough for boot, but it becomes a black art. >Obviously I've been drinking the good koolaid. Just stick to unix. Or, when feeling adventurous, try Plan9 or Qnx, -- mrr ###### From: Ben Hutchings Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 27 Feb 2003 10:23:07 +0000 Lines: 44 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <6k4i3b.hhc2.ln@via.reistad.priv.no> Reply-To: Ben Hutchings NNTP-Posting-Host: pc5-cmbg1-6-cust149.cmbg.cable.ntl.com (80.4.127.149) X-Trace: fu-berlin.de 1046341815 56760362 80.4.127.149 (16 [70929]) X-Orig-Path: 192.168.1.1!nobody User-Agent: slrn/0.9.7.4 (Linux) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!uni-erlangen.de!fu-berlin.de!uni-berlin.de!pc5-cmbg1-6-cust149.cmbg.cable.ntl.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130329 In article , Peter da Silva wrote: > In article , > Morten Reistad wrote: >>A real boot-loader is able to understand the rudiments of the >>root file system, and load the images it need before starting >>everything up. > > I thought that was the whole point of LILO. No. The LILO installer (the command called 'lilo') asks the kernel for a list of blocks occupied by the kernel image(s) listed in its configuration file and writes that to a file. Then it asks for the list of blocks occupied by the block list file, and includes it in the boot loader. Or something like that. So it needs to be reinstalled every time the kernel image(s) change, and there's no way to pick a different image at the boot prompt. GRUB only records a block list for its file-system module, so it only needs to be installed once. It provides a fairly good boot prompt from which you can boot any kernel image residing on a file-system of the selected type (or another boot-loader). I don't think it can support multiple file-systems in one installation, though. >>So, 640k (actually, 632 I think) is the limit for how much stuff can >>be loaded in the OS boot. They have had to use a compressed kernel, >>that loads, builds the rudiments of page tables, flips to '386 mode >>and decompresses the kernel before starting it. > > Geeze. > > I thought they had it pretty much all in modules by now. I mean, it was > back when Red Hat was at 4.1 that they went to modules for the install disk, > which would imply to me that they were mature enough to use. > > Obviously I've been drinking the good koolaid. Yes, they have modules, but (a) some people prefer to improve performance slightly by building everything they need into a single image, and (b) not everything can be modularised. -- Ben Hutchings | personal web site: http://womble.decadentplace.org.uk/ Lowery's Law: If it jams, force it. If it breaks, it needed replacing anyway. ###### From: Ben Hutchings Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 27 Feb 2003 13:11:27 +0000 Lines: 28 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> Reply-To: Ben Hutchings NNTP-Posting-Host: pc5-cmbg1-6-cust149.cmbg.cable.ntl.com (80.4.127.149) X-Trace: fu-berlin.de 1046351707 58684903 80.4.127.149 (16 [70929]) X-Orig-Path: 192.168.1.1!nobody User-Agent: slrn/0.9.7.4 (Linux) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!uni-erlangen.de!fu-berlin.de!uni-berlin.de!pc5-cmbg1-6-cust149.cmbg.cable.ntl.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130327 In article , nospam wrote: > in article Pine.LNX.4.55.0302251305540.16246@shiva1.cac.washington.edu, > Mark Crispin at mrc@CAC.Washington.EDU wrote on 26/02/2003 08:08: > >> On Tue, 25 Feb 2003 bfranchuk@jetnet.ab.ca wrote: >> [OK, I'm an optimist.] > I wouldn't get too excited;) >> >> At least we're no longer hearing "we can't add that to UNIX, since the >> code to do it won't fit in a 32K PDP-11" any more. > > I actually think that would be a good thing for linux, its already too > bloated even for somebody that seems to like UNIX. An OS full of good ideas, > unfortunately, every bodies good ideas - all at once, all welcome non > rejected ;) Linus is pretty strict about rejecting changes that don't really agree with the kernel architecture. The source is huge because it has support for a vast variety of hardware (both processors and peripherals), but most of that code doesn't get included in any given build. The core of it is not so large. Unfortunately any modern Unix (-clone) has to support a number of baroque and redundant APIs for compatibility. -- Ben Hutchings | personal web site: http://womble.decadentplace.org.uk/ Lowery's Law: If it jams, force it. If it breaks, it needed replacing anyway. ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 27 Feb 2003 14:14:29 GMT Organization: TSS Inc. Lines: 25 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046355269 1143 10.0.0.43 (27 Feb 2003 14:14:29 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 27 Feb 2003 14:14:29 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!mtvwca1-snh1.ops.genuity.net!paloalto-snf1.gtei.net!news.gtei.net!enews.sgi.com!news.spies.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130457 alt.sys.pdp10:12609 In article , Mark Crispin wrote: >On Wed, 27 Feb 2003, Peter da Silva wrote: >> Since when does vi do that? You must be running one of those vi >> clones there instead of the real thing (either version). >Indeed, it appears that what is installed as "vi" on this Linux system is >actually vim. I've had my runins with stealth VIM installs too. >This stupid stupid new version of GNU EMACS does not have that ability. What the hell is the point of an emacs that can't edit binary files? >Fortunately, I was just emailed the incantation to turn that behavior off: > (setq inhibit-eol-conversion t) Is there a (setq inhibit-stupid-defaults t) ? -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 27 Feb 2003 14:20:32 GMT Organization: TSS Inc. Lines: 9 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <6o9r5vc2ekhvfnku4cd86fkso6uejbc2n1@4ax.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046355632 1489 10.0.0.43 (27 Feb 2003 14:20:32 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 27 Feb 2003 14:20:32 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130481 alt.sys.pdp10:12616 Here's the closest thing to the real thing... http://www.bostic.com/vi/ -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 27 Feb 2003 14:50:28 GMT Organization: TSS Inc. Lines: 58 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046357428 3099 10.0.0.43 (27 Feb 2003 14:50:28 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 27 Feb 2003 14:50:28 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-fra1.dfn.de!npeer.de.kpn-eurorings.net!news.tele.dk!news.tele.dk!small.news.tele.dk!enews.sgi.com!news.spies.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130469 alt.sys.pdp10:12613 In article , wrote: >>I don't see that it's a bigger issue now than 10 or 20 years ago. >You're talking about portable sources that, when built, will >execute with expected behaviour. I'm encountering kiddies who >execpt the EXEs to be portable. People like that don't think in terms of "portable code", they think in terms of "of course you run Windows". >If the program had been written in a higher level language, the >library and OTS would have taken care of the portability for you. Um. You've been sheltered. I don't expect a substantial program in any language to compile first time when moving from one compiler to another on the same computer with the same OS, unless it was deliberately written for portability or it had already been ported. Doesn't matter if it's C, C++, Lisp, Smalltalk, Modula 2, or Fortran. >That used to be one of the design decisions of a project. It just >seems strange to me that people use a HLL and then proceed to do >bitty-bytey programming. Some languages, even high level languages, actually make that hard. PL/I has been a particular bugaboo, I spent a year working on software that converted PL/M to maintainable if often ugly C after intel abandoned the realtime OS business. The result worked, and we're still using the descendents of that software. That took care of the SCADA side of the business. For the applications side we had Fortran and PL/I on the Unisys 1100... the Fortran converted pretty easily, but the PL/I was full of things like "DCL FOO FIXED BIN(35);" which took human decision to figure out whether it really needed 36 bits or whether 32 would do. And it's not like the programmer was trying to do bit bashing, most of the time they were just writing high level code and not thinking about portability. We ended up just rewriting that stuff. So it was easier to convert low-level highly CPU-dependent code (in PL/M the 80x86 segment was a fundamental data type, and figuring out how it should be represented as a pointer was occasionally challenging, and the software was full of explicit bit and byte operations) from a 16-bit segmented little-endian architecture to a 32-bit flat big-endian one (and then to a 64-bit flat little-endian one), than to convert high level application code that didn't really care about bits from 36 bit word-addressed ones-complement to 64-bit byte-addressed twos-complement. Simply because the low level language didn't bother exposing details that didn't matter to the programmer. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 27 Feb 2003 15:01:31 GMT Organization: TSS Inc. Lines: 39 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046358091 3689 10.0.0.43 (27 Feb 2003 15:01:31 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 27 Feb 2003 15:01:31 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news-hog.berkeley.edu!ucberkeley!enews.sgi.com!news.spies.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130459 alt.sys.pdp10:12610 In article , Morten Reistad wrote: >>I thought they had it pretty much all in modules by now. I mean, it was >>back when Red Hat was at 4.1 that they went to modules for the install disk, >>which would imply to me that they were mature enough to use. >Of course, there is no room for more stuff in the kernel proper, >so the overflow will have to go into modules, Um, why not take things that you don't need for loading modules out of the kernel and stick them into modules? Like networking. Make networking a module, and only compile it in for the netboot kernel (and you can leave all your filesystems out of that) and you should have gobs of space. If it's that tight, put the modules and kernel in a little simple file system like SCO did, and only include that file system in the boot image. >>Obviously I've been drinking the good koolaid. >Just stick to unix. Or, when feeling adventurous, try Plan9 or Qnx, I have Plan9, haven't cracked it yet because I haven't had time to put together a network to run it on (I don't see the point of sticking it on a single computer). I just got an old Mac, so I'm going to play with BeOS, maybe AUX if I can get media. Gotta figure out which is the most recent BeOS to support the 7600. I keep wanting to boot the Sprite CD and screw around with that. It's been sitting staring at me for... getting on a decade... -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 27 Feb 2003 15:06:19 GMT Organization: TSS Inc. Lines: 21 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1281.186T405T7773222@kltpzyxm.invalid> <3E5D4816.C974A441@yahoo.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046358379 4210 10.0.0.43 (27 Feb 2003 15:06:19 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 27 Feb 2003 15:06:19 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newshub.northeast.verio.net!verio!hammer.uoregon.edu!enews.sgi.com!news.spies.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130466 alt.sys.pdp10:12612 In article , Dennis Ritchie wrote: >The sample source at www.multicians.org makes this pretty >visible. Wot a coincidence. I was just reminiscing about how much fun *I* had with 36-bit PL/I code. (peers at code) Aha! My old nemesis, "fixed bin(35)". Yep. I wouldn't want to port this. I used to gripe about how C should have provided this feature of PL/I, back when I was a newbie Berkeley student. Thank you ever so much for being smarter than that. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 34 Date: Thu, 27 Feb 2003 15:13:41 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news3.calgary.shaw.ca 1046358821 24.71.223.147 (Thu, 27 Feb 2003 08:13:41 MST) NNTP-Posting-Date: Thu, 27 Feb 2003 08:13:41 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!nntp-relay.ihug.net!ihug.co.nz!west.cox.net!cox.net!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news3.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130535 alt.sys.pdp10:12634 On Thu, 27 Feb 03 09:37:26 GMT in alt.folklore.computers, jmfbahciv@aol.com wrote: >In article , > peter@taronga.com (Peter da Silva) wrote: >>In article , wrote: >If the program had been written in a higher level language, the >library and OTS would have taken care of the portability for you. >That used to be one of the design decisions of a project. It just >seems strange to me that people use a HLL and then proceed to do >bitty-bytey programming. The quality of optimized C code generated by the best code generators is equal to that written by a good assembler programmer: registers are used as much as possible and the stack and memory are used as little as possible. A lot of assembler hacks have been built into code generators e.g. the code generator will generate the fastest sequence of instructions to get the end result if an integer expression contains some constants. Some of the factorings are non-trivial, and the instruction sequences are not obvious, like conversions of integer divides by a constant into integer multiplies by a "magic" constant which gives the same result. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### Sender: cstacy@BONK Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> From: cstacy@dtpq.com (Christopher C. Stacy) Message-ID: Lines: 29 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 27 Feb 2003 21:02:30 GMT NNTP-Posting-Host: 68.160.165.125 X-Complaints-To: abuse@verizon.net X-Trace: nwrddc02.gnilink.net 1046379750 68.160.165.125 (Thu, 27 Feb 2003 16:02:30 EST) NNTP-Posting-Date: Thu, 27 Feb 2003 16:02:30 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!c03.atl99!rip!news.webusenet.com!cox.net!cyclone1.gnilink.net!spamkiller.gnilink.net!nwrddc02.gnilink.net.POSTED!53ab2750!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130561 alt.sys.pdp10:12640 >>>>> On 27 Feb 2003 14:50:28 GMT, Peter da Silva ("Peter") writes: Peter> You've been sheltered. I don't expect a substantial program in Peter> any language to compile first time when moving from one Peter> compiler to another on the same computer with the same OS, Peter> unless it was deliberately written for portability or it had Peter> already been ported. Peter> Doesn't matter if it's C, C++, Lisp, Smalltalk, Modula 2, or Fortran. I most definitely do expect substantial Lisp programs to compile and run perfectly out of the box on all different platforms and compiler vendors. I do that all the time, using compilers from at least four different vendors, on assorted varieties of both Windows and Unix. However, if Lisp programs need to do non-ANSI things, then they do have to be written using portable libraries for that. The alternative is to code to the vendor-specific libraries - but even their libraries usually work across operating systems. One example offhand of a "substantial program" would be a complete web server; another would be all of the back-end dynamic-content applications that run on the web server, perhaps talking to an SQL database. Those particular examples use portable libraries (which are written in Lisp, but which are calling the OS-specific functions) for networking (that is, the "sockets" interface) and related multiprocessing/threading. But then, ANSI Common Lisp was specifically designed to be portable. ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Sat, 1 Mar 2003 02:43:04 -0700 Lines: 11 Message-ID: <1fr473e.19tmher19fx3diN%lars@bearnip.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1281.186T405T7773222@kltpzyxm.invalid> <3E5D4816.C974A441@yahoo.com> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1046511785 59533159 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130724 Charles Shannon Hendrix wrote: > This is how you know that people like Bill Gates are not true geeks, > computer scientists, etc. If I were a billionaire, it wouldn't > be wasted on a 12-meter yacht (maybe a tallship though) or a > huge mansion. > > It would be an underground fallout shelter with Multics machines, > PDP-11s, S/370s. Would an old missile silo complex suffice? ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Sat, 1 Mar 2003 02:43:05 -0700 Lines: 14 Message-ID: <1fr47ad.ujrbc01gjh24lN%lars@bearnip.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <3E57B900.4030601@jetnet.ab.ca> <3E5B170F.90501@beagle-ears.com> <3E5ECCE1.7040709@bellatlantic.net> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1046511786 59533159 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130720 Charles Shannon Hendrix wrote: > In article , jmfbahciv@aol.com wrote: > > >>Richie could always always find a way to bum code. > >>I think that was what we called it. > > > > Yup. That's what we called it, too. > > I remember a few people in the 6502 era calling it "page sixing" your > code, which was just a reference to cramming as much as you could into > page 6 memory. But page 6 was part of the video memory?! ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 28 Feb 03 10:51:10 GMT Organization: UltraNet Communications, Inc. Lines: 52 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> X-Trace: UmFuZG9tSVaFjG83orRkK4t/g8WD+/p9BmPjmtvHDc4GPe2U1cBZhn+jScEkTlBm X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 28 Feb 2003 11:34:56 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news.ip-plus.net!news.tesion.net!news.belwue.de!newsfeed01.sul.t-online.de!newsfeed00.sul.t-online.de!t-online.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-102-108 Xref: chonsp.franklin.ch alt.folklore.computers:130651 alt.sys.pdp10:12657 In article , peter@taronga.com (Peter da Silva) wrote: >In article , wrote: >>>I don't see that it's a bigger issue now than 10 or 20 years ago. > >>You're talking about portable sources that, when built, will >>execute with expected behaviour. I'm encountering kiddies who >>execpt the EXEs to be portable. > >People like that don't think in terms of "portable code", they think >in terms of "of course you run Windows". It's not all that clear anymore with that mess you guys call GUI. > >>If the program had been written in a higher level language, the >>library and OTS would have taken care of the portability for you. > >Um. > >You've been sheltered. That's what I've been learning. Sheltered and spoiled. I can't imagine a life without DDT. > .. I don't expect a substantial program in any language to >compile first time when moving from one compiler to another on the >same computer with the same OS, unless it was deliberately written >for portability or it had already been ported. > >Doesn't matter if it's C, C++, Lisp, Smalltalk, Modula 2, or Fortran. Since we didn't ship the first 5, I can't talk about them. I do know that we went to a hell of a lot of trouble documenting extensions in our FORTRAN manuals (and COBOL). The extensions to the standard were printed in blue (?I can't recall the color) characters with a caveat in the manual stating that these options might not compile under other compilers. There was also an effort for those extensions to be compatible (or rather, to do something smart) company-wide. I don't know what happened after the mid 80s in VMS land. I also know that the things that weren't documented in blue passed the Navy tests. I had assumed that, if a compiler passed those tests, then code written should have compiled elsewhere. But then, I've been accused of being too naive when I claim that common sense reigns. :-) /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 28 Feb 03 10:59:04 GMT Organization: UltraNet Communications, Inc. Lines: 53 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <3E57B900.4030601@jetnet.ab.ca> <3E5B170F.90501@beagle-ears.com> <3E5ECCE1.7040709@bellatlantic.net> X-Trace: UmFuZG9tSVa55A6aK14eroB/bExzRlOHwGvT28FfyVjtCcmLl65GUSXIUsoyOwGb X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 28 Feb 2003 11:42:50 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-102-108 Xref: chonsp.franklin.ch alt.folklore.computers:130659 alt.sys.pdp10:12659 In article <3E5ECCE1.7040709@bellatlantic.net>, bob smith wrote: > > >jmfbahciv@aol.com wrote: >> In article , >> Eric Smith wrote: >> >>>Lars Poulsen writes: >>> >>>>The fastest PDP-8 was the PDP-11/60 with a PDP-8 emulator loaded >>>>into its writable control store. I think that would have been around >>>>1979 ? (I remember a DECUS presentation where they hauled a smelly >>>>rastafari guy out to face the crowd and told us he had done that >>>>as a demo of just how fast and flexible the 11/60 was. He looked >>>>none too pleased with the attention. >>> >>>I don't know about the DECUS presentation, but the PDP-8 emulator on >>>the 11/60 was developed by Richie Lary. I met him in 2001, and can't >>>imagine anyone thinking of him as a "smelly rastafari guy". >> >> >> He cut his hair???!!! I'm so disappointed. >> >> >> >>>He was also involved in writing the microcode for the 11/780 and the >>>F11, much of the software for the HSCxx, and a lot of other neat stuff. >> >> >> He did more than that but I can't remember what. Tape Prep did >> quite a bit of work for him. >> >> /BAH >Richie could always always find a way to bum code. >I think that was what we called it. Yup. That's what we called it, too. > .. Richie came in >and helped on the DMC11, took a couple of days...I alwys >wondered if he put an 8 emulator in the DMC, it was the >same architecture as the 11/60 - the microcontorller/processor >was a harvard architecture little beast, about300NS cycle, 300 or >330, with the KMC being 180ns. It's possible ;-). He was working in the bad old days when programmers worked across product lines and OS snobbery hadn't become our straight jacket. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 28 Feb 03 13:04:35 GMT Organization: UltraNet Communications, Inc. Lines: 47 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> X-Trace: UmFuZG9tSVbd57jRwgI/OeoxMRu+4cGWBGzmMLM61Qbrpacfi/xOMlm5Qjxc40cX X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 28 Feb 2003 13:48:20 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!nntp-out.monmouth.com!newspeer.monmouth.com!nntp.abs.net!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-102-108 Xref: chonsp.franklin.ch alt.folklore.computers:130615 alt.sys.pdp10:12654 In article , peter@taronga.com (Peter da Silva) wrote: >In article , wrote: >>Since we didn't ship the first 5, I can't talk about them. I do >>know that we went to a hell of a lot of trouble documenting >>extensions in our FORTRAN manuals (and COBOL). The extensions >>to the standard were printed in blue (?I can't recall the color) >>characters with a caveat in the manual stating that these options >>might not compile under other compilers. > >And you think that stopped anyone from using them? > >Sheesh. Yes. > >I've been hanging out in comp.lang.smalltalk. Smalltalk is an absolutely >wonderful high level language with an abstraction mechanism second to >none. It's been available on various platforms for well over 20 years. It >has an ANSI standard and standard class libraries. > >They still don't seem to have agreed on a GUI API. Really?!! > ...And this is the language >where the whole idea of the windowing GUI originated! Even in the PDP-10 world nobody agreed on some things. But we still shipped software all the time we were arguing about it. I know people were surprised by extensions that didn't work cross-platform (we did have SPRs and customers used them to vent). I just wish I'd known about the acronym RTFM back then. On the more serious side, any I/O that was other than ASCII was known not to be transportable. I note that we're getting into that gray area of transportable vs. compatible. Our customers knew that bits stored here may not be read in quite the same way over there. /BAH Subtract a hundred and four for e-mail. ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 28 Feb 03 11:28:13 -0800 Organization: http://extra.newsguy.com Lines: 56 Message-ID: <2607.189T698T6883699@kltpzyxm.invalid> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5F0C8C.1080900@jetnet.ab.ca> NNTP-Posting-Host: p-135.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!news.gtei.net!newsfeed.mathworks.com!cyclone.swbell.net!cyclone-sf.pbi.net!129.250.175.17!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews2 Xref: chonsp.franklin.ch alt.folklore.computers:130675 alt.sys.pdp10:12661 In article peter@taronga.com (Peter da Silva) writes: >In article , >Morten Reistad wrote: > >>14W is not a lot to run a modern motherboard, hard disk, monitor etc. >>The "green" hard disks sold in malls here help a lot, combined with >>a mini-itx board. > >More seriously, if I was outfitting a boat I'd use a laptop. An >older one, even, because if battery life is any guide laptop power >requirements have gone through the roof. > >Plus an older laptop will be cheaper so when salt water does it in you >won't be too upset. Agreed. I wrote my scuba dive logging program (which includes facilities for detailed recording of observed species for survey work) so it runs under MS-DOS. It's lean and mean, and runs on any laptop you can find in the surplus stores. In effect, it runs on disposable hardware. (I can hear the Consumer Discipline Corps warming up the black helicopters already.) >>Me: How much power does it use, in use, blank and standby. > >Walk into the store with a multimeter and patch cord. When the sales >dude gives you the blank look, pull it out and measure it yourself. I went to our local cell phone dealer, wanting to see whether I could find a data cable for my wife's cell phone. I took my laptop along, because I would only be convinced by making it work on my hardware. (I did this when I got my own cell phone, which is why I still have that ancient analog-only Motorola MicroTAC 650. It's not sexy, but it does the work. My priorities are funny that way.) I plugged their cable into my wife's phone and my laptop's serial port, fired up minicom, and started typing. Nothing happened. "Oh, that's because you haven't installed the software," the sales droid said, offering me a CD-ROM. It took a while to convince him that not all laptops are running Windows; he probably had never seen Linux before. Thing got even more interesting when I pulled out my breakout box, cables, and adapters, and started stringing everything together on his counter. Their so-called "techs" were staring at me as if I was from another planet. I don't think any of them even knew what a breakout box was. I never did get the interface to work. To this day I'm sure they "know" it's because I simply didn't have the correct software. -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 01 Mar 03 12:03:41 GMT Organization: UltraNet Communications, Inc. Lines: 33 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> X-Trace: UmFuZG9tSVY2aEczWOzpKa/W6prJJOyh004601RCRYsX65xbr4vNBhCx5SRLQPC/ X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 1 Mar 2003 12:47:41 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!newsfeed.media.kyoto-u.ac.jp!newsfeed.gol.com!feed1.news.rcn.net!rcn!208-59-181-64 Xref: chonsp.franklin.ch alt.folklore.computers:130629 In article , Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: > >> now. If you watch closely, portable is getting to mean >> (to newbies who should know better) that any executable can >> run on any hardware platform without help. This is due > >They also seem to care little about the portability of data. What data? It's all got to be in core. > >"Gee, why don't you like our binary files? They are more compact!" I watched my nephew play with his computer (about five years ago). He had a very strange way of thinking about the thingies we call files. I didn't have time and was too sick to clarify his concept. But it smelled a lot like the approach JMF told me about...huh...now I can't recall the program..we were going to buy us one and study how the software arranged its data. Anyway, it was more like DBMS without the record indexing. So, I don't think that kids brought up on Windows thinks of file and data organizations the way we do. Ours is based on how we "view" the lay of the bits and theirs is based on how they "view" the lay of the window. And there's not much in common with the two. /BAH Subtract a hundred and four for e-mail. ###### From: "Paul Hardy" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 27 Feb 2003 15:53:22 -0000 Organization: Laser-Scan Ltd. Lines: 23 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3eqi5v40q3kcpl9cuhit4oedei2ifs8c6m@4ax.com> NNTP-Posting-Host: localhost X-Trace: relay.lsl.co.uk 1046361204 14555 127.0.0.1 (27 Feb 2003 15:53:24 GMT) X-Newsreader: Microsoft Outlook Express 5.50.4807.1700 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4910.0300 Originator: @lsnb18.int.lsl.co.uk ([192.168.150.4]) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!newspeer.lavaseals.co.uk!diablo.theplanet.net!psiuk-p2!psiuk-p1!uknet!psiuk-n!relay.lsl.co.uk!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130594 alt.sys.pdp10:12653 "Alan J. Wylie" wrote in message news:wgisv9hpg1.fsf@snoeze.glaramara.freeserve.co.uk... > > The VAX 11-780 I used in 1979-1985 (serial number GAL 42) arrived with > 512 Kb of memory. Some time later an extra 256Kb was added, at great > expense. Don't exaggerate. It was GA00047, and it arrived with 768KB. The minimum you could buy a VAX with at the time was 256KB, but an extra 512KB was ordered at a cost of £8177. (I was the original System Manager, and handed it over to Alan. I have the advantage of still being at the same company and hence have access to the original files!). -- Paul Hardy (PGH), Principal Consultant, Laser-Scan Ltd, Science Park, Milton Rd, CAMBRIDGE, CB4 0FY, GB. Tel: +44 (0)1223 420414 Fax: 420044, Email: Paul@LSL.co.uk, Web: http://www.Laser-Scan.com ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 28 Feb 2003 00:56:30 GMT Organization: TSS Inc. Lines: 19 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046393790 34602 10.0.0.43 (28 Feb 2003 00:56:30 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 28 Feb 2003 00:56:30 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130703 alt.sys.pdp10:12669 In article , Christopher C. Stacy wrote: >I most definitely do expect substantial Lisp programs to compile >and run perfectly out of the box on all different platforms and >compiler vendors. Fair enough, I haven't tried that for a decade or two, but it sure wasn't true in the period we're talking about... and IIRC that's one of the reasons Common Lisp exists. >But then, ANSI Common Lisp was specifically designed to be portable. Yeh, that's the one. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### Message-ID: <3E5ECCE1.7040709@bellatlantic.net> From: bob smith User-Agent: Mozilla/5.0 (Windows; U; WinNT4.0; en-US; rv:1.0.1) Gecko/20020823 Netscape/7.0 X-Accept-Language: en-us, en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <3E57B900.4030601@jetnet.ab.ca> <3E5B170F.90501@beagle-ears.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Lines: 45 Date: Fri, 28 Feb 2003 02:32:36 GMT NNTP-Posting-Host: 138.88.89.120 X-Complaints-To: abuse@verizon.net X-Trace: nwrddc01.gnilink.net 1046399556 138.88.89.120 (Thu, 27 Feb 2003 21:32:36 not) NNTP-Posting-Date: Thu, 27 Feb 2003 21:32:36 not Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!cox.net!cyclone1.gnilink.net!spamkiller.gnilink.net!nwrddc01.gnilink.net.POSTED!53ab2750!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130658 alt.sys.pdp10:12658 jmfbahciv@aol.com wrote: > In article , > Eric Smith wrote: > >>Lars Poulsen writes: >> >>>The fastest PDP-8 was the PDP-11/60 with a PDP-8 emulator loaded >>>into its writable control store. I think that would have been around >>>1979 ? (I remember a DECUS presentation where they hauled a smelly >>>rastafari guy out to face the crowd and told us he had done that >>>as a demo of just how fast and flexible the 11/60 was. He looked >>>none too pleased with the attention. >> >>I don't know about the DECUS presentation, but the PDP-8 emulator on >>the 11/60 was developed by Richie Lary. I met him in 2001, and can't >>imagine anyone thinking of him as a "smelly rastafari guy". > > > He cut his hair???!!! I'm so disappointed. > > > >>He was also involved in writing the microcode for the 11/780 and the >>F11, much of the software for the HSCxx, and a lot of other neat stuff. > > > He did more than that but I can't remember what. Tape Prep did > quite a bit of work for him. > > /BAH Richie could always always find a way to bum code. I think that was what we called it. Richie came in and helped on the DMC11, took a couple of days...I alwys wondered if he put an 8 emulator in the DMC, it was the same architecture as the 11/60 - the microcontorller/processor was a harvard architecture little beast, about300NS cycle, 300 or 330, with the KMC being 180ns. bob > > > Subtract a hundred and four for e-mail. ###### From: Peter da Silva Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 27 Feb 2003 21:42:31 -0600 Organization: ABB Inc., Network Management Division Lines: 29 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: switchblade.dhcp.abb.abbnm.com Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-Trace: jeeves.eng.abbnm.com 1046403861 77291 10.127.143.26 (28 Feb 2003 03:44:21 GMT) X-Complaints-To: usenet@jeeves.eng.abbnm.com NNTP-Posting-Date: Fri, 28 Feb 2003 03:44:21 +0000 (UTC) User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.1) Gecko/20020826 X-Accept-Language: en-us, en, ja Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.telusplanet.net!newsfeed.bc.tac.net!cyclone.bc.net!newsfeed.stanford.edu!news-proxy.abbnm.com!newsboy.eng.abbnm.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130689 alt.sys.pdp10:12664 Morten Reistad wrote: > If you get too aggressive on exporting things into modules, and the > bootloader cannot save you from your own stupidity it may come back > and bite you. Like when you suddenly have to boot from a CD that was > made iso9660, and you forgot to compile in static support for that. I know. That's why I mentioned RH 4.1, because that was the version where Red Hat started to use modules for boot, and it didn't support any of my SCSI controllers... all perfectly normal and conventional Adaptecs. Networking in Linux, however, should be pretty modularisable, since it's traditionally been developed independently of the kernel proper. In any case, I discussed this with our local Linux guy, and he says that LILO is now capable of loading kernels over 640K, so it's really a non-issue. > My project for computing is to bring a computer to my boat. I can > afford 1.2 amps of 12V to feed it. It will be on a diet. 1.2 amps of 12V? Luxury! My smallest computer has run for up to a month on a pair of AAAs. -- `-_-' Ar rug tú barróg ar do mhactíre inniu? 'U` Peter da Silva ###### From: bfranchuk@jetnet.ab.ca Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 28 Feb 2003 00:15:24 -0700 Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3E5F0C8C.1080900@jetnet.ab.ca> User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.0.1) Gecko/20021005 X-Accept-Language: en-us, en MIME-Version: 1.0 References: <3E05FE9D.B3A12286@vax6k.openecs.org> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 11 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!sn-xit-03!sn-xit-01!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130586 alt.sys.pdp10:12651 Peter da Silva wrote: >>My project for computing is to bring a computer to my boat. I can >>afford 1.2 amps of 12V to feed it. It will be on a diet. > 1.2 amps of 12V? Luxury! My smallest computer has run for up to a > month on a pair of AAAs. But he needs to use it as boat anchor too! :) Ben. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5F0C8C.1080900@jetnet.ab.ca> From: Morten Reistad X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: mrr@acer.reistad.priv.no (Morten Reistad) Message-ID: Lines: 69 Date: Fri, 28 Feb 2003 10:29:32 +0100 NNTP-Posting-Host: 80.111.42.204 X-Complaints-To: abuse@chello.no X-Trace: amstwist00 1046426401 80.111.42.204 (Fri, 28 Feb 2003 11:00:01 MET) NNTP-Posting-Date: Fri, 28 Feb 2003 11:00:01 MET Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!amsnews01.chello.com!amstwist00.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130684 alt.sys.pdp10:12662 According to : >Peter da Silva wrote: > >>>My project for computing is to bring a computer to my boat. I can >>>afford 1.2 amps of 12V to feed it. It will be on a diet. >> 1.2 amps of 12V? Luxury! My smallest computer has run for up to a >> month on a pair of AAAs. > >But he needs to use it as boat anchor too! I already have a large supply of boat anchors, thank you. But I do feel at odds with tbe world when out in shops to do outfitting. 14W is not a lot to run a modern motherboard, hard disk, monitor etc. The "green" hard disks sold in malls here help a lot, combined with a mini-itx board. Now for the monitor; I am out in the stores asking the "wrong" kind of questions: Sales dude: Here is the latest LCD small screen, 30ms update time, 1024x768 pixels, meegazillion colors .. Me: How much power does it use, in use, blank and standby. Sales dude looks at power supply : It says 80 watts here (looks at CE approval label) Me: That is the approval for the fuse, not the power consumption. etc. Similar when a sales dude insisted on me trying out their fancy navigation software on my laptop. It was running, so I gave it to him; it was the day I tried out the "enlightenment" look of window manager, all under FreeBSD. It was real entertainment watching him look for "start". "Oh, you didn't say it had to run Windows". There has been only one shop that has shown themselves very competent and helpful in every respect. That is the "fine woods" warehouse. When refurbishing a boat there is a large need for small tidbits of teak, mahogany and other threatened species of wood to replace the bits that has been waterlogged. This store has it all, and it even accepts returns for the bits I don't use; with a 15% price reduction. There are a couple of pretty competent carpenters that know the ins and outs of rare woods to handle customers (and you walk right in and look at the goods in the 7 meter high shelves). When you buy, a pencil-written ticket is issued, and you take this to their office next door. The amounts are issued into a large, manual ledger, and totals tallied up on old ADDO and FACIT adders. There is a PC in the room, but that is only used as a 3270 terminal to the network of woods suppliers. There is an ancient old modem running bisync blinkenlights there too. (it has a bright new LCD screen; the old screen gave up earlier this year).. It is the only time I have seen business done the "old way" for a very long time. -- mrr ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 28 Feb 2003 10:04:40 -0000 Organization: Titanic Enterprises Unlimited Lines: 25 Message-ID: <1046426680.806367@saucer.planet.gong> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3eqi5v40q3kcpl9cuhit4oedei2ifs8c6m@4ax.com> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1046426681 3072 80.177.7.220 (28 Feb 2003 10:04:41 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Fri, 28 Feb 2003 10:04:41 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Cache: nntpcache 3.0.1 (see http://www.nntpcache.org/) X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Cache-Post-Path: saucer.planet.gong!unknown@voodoo.planet.gong Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-fra1.dfn.de!news0.de.colt.net!peernews3.colt.net!colt.net!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130665 alt.sys.pdp10:12660 "Paul Hardy" wrote in message news:Qno*C04Lp@relay.lsl.co.uk... > > "Alan J. Wylie" wrote in message > news:wgisv9hpg1.fsf@snoeze.glaramara.freeserve.co.uk... > > > > The VAX 11-780 I used in 1979-1985 (serial number GAL 42) arrived with > > 512 Kb of memory. Some time later an extra 256Kb was added, at great > > expense. > > Don't exaggerate. It was GA00047, and it arrived with 768KB. > > The minimum you could buy a VAX with at the time was 256KB, but an extra > 512KB was ordered at a cost of £8177. > > (I was the original System Manager, and handed it over to Alan. I have the > advantage of still being at the same company and hence have access to the > original files!). You've been hanging around Catbert 'Evil HR Director' too long. Cheers, Rupert ###### From: alan.nospam@glaramara.freeserve.co.uk (Alan J. Wylie) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 28 Feb 2003 10:42:10 +0000 Organization: very little Lines: 24 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3eqi5v40q3kcpl9cuhit4oedei2ifs8c6m@4ax.com> NNTP-Posting-Host: 62.25.242.11 Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Trace: newsg1.svr.pol.co.uk 1046441535 24229 62.25.242.11 (28 Feb 2003 14:12:15 GMT) NNTP-Posting-Date: 28 Feb 2003 14:12:15 GMT X-Complaints-To: abuse@theplanet.net User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!newspeer.lavaseals.co.uk!diablo.theplanet.net!news.theplanet.net!bilbo!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130637 alt.sys.pdp10:12656 On Thu, 27 Feb 2003 15:53:22 -0000, "Paul Hardy" said: >> The VAX 11-780 I used in 1979-1985 (serial number GAL 42) arrived >> with 512 Kb of memory. Some time later an extra 256Kb was added, at >> great expense. > Don't exaggerate. It was GA00047, and it arrived with 768KB. My apologies - it was a long time ago - my memory must be suffering from parity errors. > The minimum you could buy a VAX with at the time was 256KB, but an > extra 512KB was ordered at a cost of £8177. To put that in context, my starting salary at Laser-Scan was £4000/year. BTW, there is a slow news feed somewhere - Paul's courtesy copy arrived yesterday afternoon. -- Alan J. Wylie http://www.glaramara.freeserve.co.uk/ I'm looking for a job. CV at http://www.glaramara.freeserve.co.uk/cv.html Unix/Linux/C/Internet/embedded and lots more. UK:Bradford/Manchester ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 28 Feb 2003 12:34:01 GMT Organization: TSS Inc. Lines: 23 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5F0C8C.1080900@jetnet.ab.ca> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046435641 72696 10.0.0.43 (28 Feb 2003 12:34:01 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 28 Feb 2003 12:34:01 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!cyclone.bc.net!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130714 alt.sys.pdp10:12671 In article , Morten Reistad wrote: >14W is not a lot to run a modern motherboard, hard disk, monitor etc. >The "green" hard disks sold in malls here help a lot, combined with >a mini-itx board. More seriously, if I was outfitting a boat I'd use a laptop. An older one, even, because if battery life is any guide laptop power requirements have gone through the roof. Plus an older laptop will be cheaper so when salt water does it in you won't be too upset. >Me: How much power does it use, in use, blank and standby. Walk into the store with a multimeter and patch cord. When the sales dude gives you the blank look, pull it out and measure it yourself. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 28 Feb 2003 12:39:47 GMT Organization: TSS Inc. Lines: 25 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046435987 72761 10.0.0.43 (28 Feb 2003 12:39:47 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 28 Feb 2003 12:39:47 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130705 alt.sys.pdp10:12670 In article , wrote: >Since we didn't ship the first 5, I can't talk about them. I do >know that we went to a hell of a lot of trouble documenting >extensions in our FORTRAN manuals (and COBOL). The extensions >to the standard were printed in blue (?I can't recall the color) >characters with a caveat in the manual stating that these options >might not compile under other compilers. And you think that stopped anyone from using them? Sheesh. I've been hanging out in comp.lang.smalltalk. Smalltalk is an absolutely wonderful high level language with an abstraction mechanism second to none. It's been available on various platforms for well over 20 years. It has an ANSI standard and standard class libraries. They still don't seem to have agreed on a GUI API. And this is the language where the whole idea of the windowing GUI originated! -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### Message-ID: <3E5FB200.673AA397@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 20 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: sccrnsc03 1046451780 12.241.15.59 (Fri, 28 Feb 2003 17:03:00 GMT) NNTP-Posting-Date: Fri, 28 Feb 2003 17:03:00 GMT Date: Fri, 28 Feb 2003 17:03:00 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!c03.atl99!sjc70.webusenet.com!sjc72.webusenet.com!news.webusenet.com!wn14feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi_feed4!attbi.com!sccrnsc03.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130739 alt.sys.pdp10:12676 Peter da Silva wrote: > > [snip...] [snip...] [snip...] > > I've been hanging out in comp.lang.smalltalk. Smalltalk is an absolutely > wonderful high level language with an abstraction mechanism second to > none. It's been available on various platforms for well over 20 years. It > has an ANSI standard and standard class libraries. > > They still don't seem to have agreed on a GUI API. And this is the language > where the whole idea of the windowing GUI originated! > One might email Doug Englebart and let him know that Smalltalk originated the GUI idea. Heck, you might have told him before he went to all the trouble of patenting the computer mouse. -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### From: Howard S Shubs Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 28 Feb 2003 13:00:55 -0500 Organization: ='SEQUENTIAL' Lines: 9 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> <3E5C026F.5080708@jetnet.ab.ca> NNTP-Posting-Host: p-082.newsdawg.com Mail-Copies-To: nobody User-Agent: MT-NewsWatcher/3.2 (PPC Mac OS X) X-Face: "S"r{U%bs].&Ud}Pc~~~0a]M:t5l>>EN\1Faw10M9NK1Xq59wo7-"s0S+[{etQorO /Nf-Ci"i9v'MT!R8)J]N[4|2&x1r^Iq&{SB"6dknr0=+6UFb.>+{zMn_1=rw&/V+"d@* ZS5\LoW_ Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!cyclone.swbell.net!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!howard Xref: chonsp.franklin.ch alt.folklore.computers:130731 alt.sys.pdp10:12673 In article , stanb45@dial.pipex.com (Stan Barr) wrote: > Then run it on a Mac, as I've just started doing :-) That was cruel. :-) -- Today, on Paper-view: The World Origami Championship ###### From: Sean Case Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Marginal References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> User-Agent: MT-NewsWatcher/3.1 (PPC) Message-ID: Lines: 19 Date: Sat, 01 Mar 2003 07:08:55 +1100 NNTP-Posting-Host: 61.8.1.45 X-Complaints-To: news@pacific.net.au X-Trace: nasal.pacific.net.au 1046462653 61.8.1.45 (Sat, 01 Mar 2003 07:04:13 EST) NNTP-Posting-Date: Sat, 01 Mar 2003 07:04:13 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!newsfeed.zip.com.au!nasal.pacific.net.au!gsc Xref: chonsp.franklin.ch alt.folklore.computers:130717 alt.sys.pdp10:12672 In article , Morten Reistad wrote: > A real boot-loader is able to understand the rudiments of the > root file system, and load the images it need before starting > everything up. With some basic loader options for choosing > different images, as appropriate. BSD, SunOS, Solaris, SGI, QNX > and others all have variations on this theme. Alternatively, marking an image as bootable could automatically copy a list of extents to a known location on disk. I'm sure that I've seen this done in a commercial Unix. Sean Case -- Sean Case gsc@zip.com.au Code is an illusion. Only assertions are real. ###### From: Peter da Silva Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 28 Feb 2003 14:39:43 -0600 Organization: ABB Inc., Network Management Division Lines: 18 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5FB200.673AA397@ev1.net> NNTP-Posting-Host: switchblade.dhcp.abb.abbnm.com Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-Trace: jeeves.eng.abbnm.com 1046464892 97450 10.127.143.26 (28 Feb 2003 20:41:32 GMT) X-Complaints-To: usenet@jeeves.eng.abbnm.com NNTP-Posting-Date: Fri, 28 Feb 2003 20:41:32 +0000 (UTC) User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.1) Gecko/20020826 X-Accept-Language: en-us, en, ja Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!headwall.stanford.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!newsboy.eng.abbnm.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130685 alt.sys.pdp10:12663 Charles Richmond wrote: > One might email Doug Englebart and let him know that Smalltalk > originated the GUI idea. I know about his work in hypertext and graphical interfaces, but I wasn't aware that he'd come up with windows as well. All the clips and movies I've seen of his software present a single document taking up the whole screen. I think I recall a split-screen display, but nothing like the Xerox windows. That's why I said "windowing gui" and not just "gui". Anyway, if I'm mistaken about that, then I apologise, and I'd really appreciate a link or something. -- `-_-' Ar rug tú barróg ar do mhactíre inniu? 'U` Peter da Silva ###### Message-ID: <3E5FDA7E.E69A472A@yahoo.com> From: Peter Flass X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 20 Date: Fri, 28 Feb 2003 21:57:06 GMT NNTP-Posting-Host: 24.194.50.82 X-Complaints-To: abuse@rr.com X-Trace: twister.nyroc.rr.com 1046469426 24.194.50.82 (Fri, 28 Feb 2003 16:57:06 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 16:57:06 EST Organization: Road Runner Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!news-west.rr.com!cyclone.nyroc.rr.com!cyclone-out.nyroc.rr.com!twister.nyroc.rr.com.POSTED!53ab2750!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130575 alt.sys.pdp10:12644 Peter da Silva wrote: > > Some languages, even high level languages, actually make that hard. > > PL/I has been a particular bugaboo, [snip] > For the applications side we had Fortran and PL/I on the > Unisys 1100... the Fortran converted pretty easily, but the PL/I was > full of things like "DCL FOO FIXED BIN(35);" which took human decision > to figure out whether it really needed 36 bits or whether 32 would do. > > And it's not like the programmer was trying to do bit bashing, most of the > time they were just writing high level code and not thinking about > portability. We ended up just rewriting that stuff. Worrying about it was really overkill. If you were porting C, or when you ported FORTRAN, you wouldn't worry about the size of an int (usually). You'd just take whatever the machine gives you. You'd get the same result with PL/I just converting all the FIXED BIN(35) to FIXED BIN(31). ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 16 Date: Fri, 28 Feb 2003 17:58:23 -0500 NNTP-Posting-Host: 209.96.179.247 X-Trace: sydney.visi.net 1046474796 209.96.179.247 (Fri, 28 Feb 2003 18:26:36 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 18:26:36 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-fra1.dfn.de!news0.de.colt.net!peernews3.colt.net!colt.net!peernews-us.colt.net!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!news.he.net!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130573 alt.sys.pdp10:12642 In article <3E5A710D.6030403@jetnet.ab.ca>, bfranchuk@jetnet.ab.ca wrote: > It is said only about 10? lines of bug free code > can be produced a day with all the docs and other > needed information. I never have believed generalizations like this. A lot depends on what you are writing, and how complex it is. Certainly I've been on projects where we'd put on a ton of code, and even after a year or more, our initial coding rate was far better than any of these rules-of-thumb I've heard. At other times, I think coming to work was a bug... ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 24 Date: Fri, 28 Feb 2003 18:01:15 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046481982 209.96.179.96 (Fri, 28 Feb 2003 20:26:22 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 20:26:22 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!feedme.news.mediaways.net!news-fra1.dfn.de!news-han1.dfn.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newshosting.com!news-xfer1.atl.newshosting.com!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130579 alt.sys.pdp10:12647 In article , Trevor Jenkins wrote: > Ah teco. Yes, wrote lots of lovely "macros" in that too. My best being one > for removing tags from huge SGML-encoded documents and it substituted > entities as required. But that one was on VAX/VMS. Thankfully not > something that I needed to do twice as the macro was highly convoluted. My > TOPS-10 ones were simialr and did rather affect the system performance but > what did I care I was the systems programmer --- got to blame the student > users. :-) I was insane enough one time to write some code in vi macros. It went through a file and did some intelligent processing. Mind you, it started out very simple, but I kept thinking... just one more buffer and a transformation, and... Now I just write scripts that vi calls on line ranges, using a more appropriate language. I used to have a vi macro that did the towers of hanoi solution, complete with vt100 graphics animation. Some people have tremendous amounts of time on their hands. ###### Sender: cstacy@BONK Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> From: cstacy@dtpq.com (Christopher C. Stacy) Message-ID: Lines: 49 User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 28 Feb 2003 23:32:47 GMT NNTP-Posting-Host: 68.160.165.125 X-Complaints-To: abuse@verizon.net X-Trace: nwrddc04.gnilink.net 1046475167 68.160.165.125 (Fri, 28 Feb 2003 18:32:47 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 18:32:47 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!news.webusenet.com!cyclone1.gnilink.net!spamkiller.gnilink.net!nwrddc04.gnilink.net.POSTED!53ab2750!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130593 alt.sys.pdp10:12652 >>>>> On Thu, 27 Feb 2003 16:58:05 -0800, Mark Crispin ("Mark") writes: Mark> On Thu, 27 Feb 2003, Christopher C. Stacy wrote: >> I most definitely do expect substantial Lisp programs to compile >> and run perfectly out of the box on all different platforms and >> compiler vendors. Mark> Will Maxima compile and run perfectly out of the box on TOPS-20 Mark> Common Lisp? There would be two issues with Macsyma. 1. Whether or not it is an ANSI Common Lisp program. Originally, it was in MACLISP, and the "Maxima" version of Macsyma was based on that, followed by some other version of Lisp, followed by an early pre-ANSI pre-standardized version of Common Lisp. I think they may have actually finally ported the program to ANSI Common Lisp, in which case it would work fine, except for the following issue. 2. What did they use to program the GUI? That's not part of the language standard, and I don't know what they did. Some options are to use X Windows (eg. "CLX"), or to program directly to the Microsoft Windows API, or to use some Common Lisp compiler vendor's GUI library. Given all that: I believe the answer to your original question is "Yes", it's all winning now, because I saw some announcements to that effect in recent months. But I haven't seen the code, don't know about the above issues, etc. So I'm not really sure. The Symbolics version was ported to ANSI Common Lisp a long time ago, and I worked on it a little bit when it was later owned by "Macsyma, Inc". At that time, it was portable and compiled fine on the two different Lisp implementations that we had in house. In fact, I developed on the Lisp Machine, and then those very sources went right over to the i-386 (if you can imagine that nightmare!) and were compiled there. That all just worked. We had seperate GUI implementations for the Symbolics and for Windows. Eventually, the Windows one was much fancier with "smart math notebooks" and animated 3D graphics and all kinds of cool shit. The part of the interface that received input lines you had just typed in, and then tokenized those strings and parsed them into expressions to be dealt with, was all portable. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 39 Date: Fri, 28 Feb 2003 18:42:52 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046481984 209.96.179.96 (Fri, 28 Feb 2003 20:26:24 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 20:26:24 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130583 alt.sys.pdp10:12650 In article , Mark Crispin wrote: > Don't say that. The Linux world is slowly re-discovering software design > techniques which were well-known to us in the PDP-10 world. In another 10 > years or so, they may reach the point where we were in the early 1980s. > [OK, I'm an optimist.] I don't see it as rediscovery as much as things have changed to allow it to happen more than it had been. The Linux/UNIX crowd is usually much more receptive to studying history, and doing the hard work needed for good software. I also suspect that today's favorite metric--how fast can you puke out the code--was probably not the way programmers were judged in those good old days. Systems and software were also simpler, and the industry was much more primitive. It's a whole new ball game now. It might be interesting for you to point out the techniques you think are missing and being rediscovered. There are people out here who are interested in hearing about it. > At least we're no longer hearing "we can't add that to UNIX, since the > code to do it won't fit in a 32K PDP-11" any more. Yeah... the there is an opposite problem: hey you have a 1GB memory, terabytes of storage, and a MHz CPU. Why not add this new feature we don't need but think would be really cool and shiny? Sometimes I long for an end to Moore's Law. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5BC109.2050003@jetnet.ab.ca> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 21 Date: Fri, 28 Feb 2003 18:55:17 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046481985 209.96.179.96 (Fri, 28 Feb 2003 20:26:25 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 20:26:25 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.belwue.de!news.tu-darmstadt.de!newsfeed.freenet.de!newsfeed.news2me.com!newshosting.com!news-xfer1.atl.newshosting.com!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130570 alt.sys.pdp10:12641 In article <6k4i3b.hhc2.ln@via.reistad.priv.no>, Morten Reistad wrote: > Seriously: I saw this as a hack even in 1994, when I first installed > Linux. The level of shoehorning they have done with the 640k boot image > size is just amazing. I don't know where you guys get this 640K thing. The Linux kernel does not have to fit in 640K. Mine is over 1MB compressed. LILO is only a two stage loader, but you can do multi-stage boots if you need to. > Go compare the bootstrap code from e.g. FreeBSD and Linux, and you > will see. > Besides, Linux is for heretics. It is not the one, true unix!. Which one is that? If you mean anything but NetBSD, shame on you... :) ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 19 Date: Fri, 28 Feb 2003 18:57:12 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046481986 209.96.179.96 (Fri, 28 Feb 2003 20:26:26 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 20:26:26 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news.he.net!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130581 alt.sys.pdp10:12649 In article , Peter da Silva wrote: > In 1994 I was still running a couple of Xenix boxes to support customers > on legacy systems. The minimum memory requirement to boot Xenix-286 to > multiuser was 192K. If someone had told me that you couldn't fit a boot > loader into 640K I'd have told them they were nuts. > > What the hell is it doing? Their information is wrong. The Linux loader does not need 640K. The boot loader is a few K, secondary only a little larger. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 49 Date: Fri, 28 Feb 2003 19:06:42 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046481988 209.96.179.96 (Fri, 28 Feb 2003 20:26:28 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 20:26:28 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!news.gtei.net!news-out.visi.com!hermes.visi.com!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130574 alt.sys.pdp10:12643 In article , Morten Reistad wrote: > Linux does not have a real boot-loader per se. > > A real boot-loader is able to understand the rudiments of the > root file system, and load the images it need before starting > everything up. I still use LILO, because it works fine for me. But, there is also grub and some commercial editors. > Linux, OTOH, has just an image that is loaded in real-mode, from > a dedicated micro-partition, and that is it. It is not a dedicated partition. LILO (if you use that) does have to know where the kernel is, but it's on the root partition. > So, 640k (actually, 632 I think) is the limit for how much stuff can > be loaded in the OS boot. My Linux kernel is 3MB. Even compressed I've booted 2MB kernels. My current kernel is about 1.1MB compressed. > They have had to use a compressed kernel, that loads, builds the > rudiments of page tables, flips to '386 mode and decompresses the > kernel before starting it. It's not limited to 640K. A PC BIOS can only boot an image up to 2.88MB, the size of the highest density 3.5" floppy. This is true even for a CD boot. The boot image on the CD is also limited to 2.88MB. Linux should have a real boot loader, but there are even issues with the *BSD loaders. It's not fun doing boot setups with a PC, because they have horribly broken BIOS. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 24 Date: Fri, 28 Feb 2003 19:10:54 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046481989 209.96.179.96 (Fri, 28 Feb 2003 20:26:29 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 20:26:29 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!news.gtei.net!news-out.visi.com!hermes.visi.com!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130578 alt.sys.pdp10:12646 In article , Peter da Silva wrote: > Um, why not take things that you don't need for loading modules out of > the kernel and stick them into modules? This is done by the distributions. They offer a variety of boot kernels depending on what you need. Building a boot loader that can understand all the different possible PC hardware and all the different filesystems offered by Linux is decidedly non-trivial. Most other systems have a boot loader because they don't support much hardware, and support very few boot filesystems. A PC is a hardware support nightmare. > I keep wanting to boot the Sprite CD and screw around with that. It's > been sitting staring at me for... getting on a decade... I hate that it died. Seems like so many projects just fade away... It should run nicely on a used SS5 or something. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 24 Date: Fri, 28 Feb 2003 19:16:14 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046481990 209.96.179.96 (Fri, 28 Feb 2003 20:26:30 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 20:26:30 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!colt.net!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130580 alt.sys.pdp10:12648 In article , Bill/Carolyn Pechter wrote: >>Besides, Linux is for heretics. It is not the one, true unix!. >> >>-- mrr >> > > > Very, very true... > > But I still haven't got FreeBSD running Simh's ethernet support Ahem... he said "the one, true unix!", not FreeBSD. NetBSD is the last true UNIX... :) > for the VAX emulation... The house linux box does it fine. > I'm thinking of punching a hole in my house firewall and letting it get > routed to the VAX emulation under linux... How slick would that be... That would be a nice trick. I think someone else is running a virtual PDP-10 like that already. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 15 Date: Fri, 28 Feb 2003 19:18:11 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046481992 209.96.179.96 (Fri, 28 Feb 2003 20:26:32 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 20:26:32 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!enews.sgi.com!newsfeed.mathworks.com!nycmny1-snh1.gtei.net!news.gtei.net!news-out.visi.com!hermes.visi.com!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130769 alt.sys.pdp10:12683 In article , nospam wrote: > I actually think that would be a good thing for linux, its already too > bloated even for somebody that seems to like UNIX. An OS full of good ideas, > unfortunately, every bodies good ideas - all at once, all welcome non > rejected ;) That's mostly support for a vast array of hardware though. My working Linux kernel hasn't changed much in years. You can greatly trim a kernel once your system is installed by building a new one sans all the hardware support you don't need any more. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E5BC109.2050003@jetnet.ab.ca> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 23 Date: Fri, 28 Feb 2003 19:23:49 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046481993 209.96.179.96 (Fri, 28 Feb 2003 20:26:33 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 20:26:33 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!news.gtei.net!newsfeed.mathworks.com!nntp.abs.net!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130577 alt.sys.pdp10:12645 In article , jmfbahciv@aol.com wrote: >>At least we're no longer hearing "we can't add that to UNIX, since the >>code to do it won't fit in a 32K PDP-11" any more. > > I sure would like apps people to know about the usefulness of > GETSEGs. IOW, you don't have to have the whole kitchen sink > in core in order to open the refrigerator. Linux loads by demand paging. Code doesn't get loaded unless it is called. I can't remember what kinds of modifications there are to this rule. Doing this manually in an application really doesn't have much benefit. The real key is leaning up the applications, and making their parts less interdependent. I don't see this happening. Some systems used to fully allocate the exec and also any swap space that process would need. ###### From: ararghNOSPAM@NOT.AT.enteract.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 28 Feb 2003 19:14:39 -0600 Organization: Not Really! Lines: 34 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5FDA7E.E69A472A@yahoo.com> NNTP-Posting-Host: tcr202.dynip.ripco.net Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: e250.ripco.com 1046481311 23737 209.100.226.202 (1 Mar 2003 01:15:11 GMT) X-Complaints-To: usenet@ripco.com NNTP-Posting-Date: Sat, 1 Mar 2003 01:15:11 +0000 (UTC) X-Newsreader: Forte Agent 1.92/32.572 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!gail.ripco.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130734 alt.sys.pdp10:12674 On Fri, 28 Feb 2003 21:57:06 GMT, Peter Flass wrote: >Peter da Silva wrote: >> >> Some languages, even high level languages, actually make that hard. >> >> PL/I has been a particular bugaboo, >[snip] >> For the applications side we had Fortran and PL/I on the >> Unisys 1100... the Fortran converted pretty easily, but the PL/I was >> full of things like "DCL FOO FIXED BIN(35);" which took human decision >> to figure out whether it really needed 36 bits or whether 32 would do. >> >> And it's not like the programmer was trying to do bit bashing, most of the >> time they were just writing high level code and not thinking about >> portability. We ended up just rewriting that stuff. > >Worrying about it was really overkill. If you were porting C, or when >you ported FORTRAN, you wouldn't worry about the size of an int >(usually). You'd just take whatever the machine gives you. You'd get >the same result with PL/I just converting all the FIXED BIN(35) to FIXED >BIN(31). Always assumming that the program doesn't expect to store values > 2^31 and then expect to get back the same number. :-) I have never done this type of conversion, but I would expect to spend FAR more time checking the program for varable size assumptions, than I would spend time in the conversion. -- Arargh (at arargh dot com) http://www.arargh.com To reply by email, change the domain name, and remove the garbage. (Enteract can keep the spam, they are gone anyway) ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <1281.186T405T7773222@kltpzyxm.invalid> <3E5D32A5.89915AB7@ev1.net> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 14 Date: Fri, 28 Feb 2003 21:29:03 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046489229 209.96.179.96 (Fri, 28 Feb 2003 22:27:09 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 22:27:09 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!c03.atl99!news.webusenet.com!news.maxwell.syr.edu!newsswitch.lcs.mit.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130762 alt.sys.pdp10:12680 In article <3E5D32A5.89915AB7@ev1.net>, Charles Richmond wrote: >> You are a nasty cynical man. I can't help but approve. >> > "The power of accurate observation is commonly called cynicism > by those who have not got it." > --George Bernard Shaw A cynic is an idealist whose rose colored glasses have been removed, snapped in two, and stomped into the ground, immediately improving his vision. http://www.i-cynic.com ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <1281.186T405T7773222@kltpzyxm.invalid> <3E5D4816.C974A441@yahoo.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 32 Date: Fri, 28 Feb 2003 21:39:32 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046489244 209.96.179.96 (Fri, 28 Feb 2003 22:27:24 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 22:27:24 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!newspump.monmouth.com!newspeer.monmouth.com!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130752 alt.sys.pdp10:12677 In article <3E5D4816.C974A441@yahoo.com>, Peter Flass wrote: > Actually, Multics was designed to be portable, and in fact *was* ported > from a GE-645 to a Honeywell ... something, a similar machine, but with > some significant differences. The only reason it wasn't more widely > ported is that most current hardware lacks the support for segmentation > and rings that it requires. I thought some modern CPUs had enough to handle the rings, and the segments could be done on top of existing MMUs, if the CPU is a purely flat system. In any case, with the power we have, I'd like to see it running again, even if something had to be done in software. This is how you know that people like Bill Gates are not true geeks, computer scientists, etc. If I were a billionaire, it wouldn't be wasted on a 12-meter yacht (maybe a tallship though) or a huge mansion. It would be an underground fallout shelter with Multics machines, PDP-11s, S/370s. Wealthy people are among the most dull and boring people on Earth. > As I believe I mentioned before, the only reason unix has been so > widely ported is that it only uses the least common denominator in > hardware. That's no longer true: all hardware is now equally common... :) ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 9 Date: Fri, 28 Feb 2003 21:43:19 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046489261 209.96.179.96 (Fri, 28 Feb 2003 22:27:41 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 22:27:41 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!feed2.news.rcn.net!rcn!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130755 alt.sys.pdp10:12679 In article , jmfbahciv@aol.com wrote: > now. If you watch closely, portable is getting to mean > (to newbies who should know better) that any executable can > run on any hardware platform without help. This is due They also seem to care little about the portability of data. "Gee, why don't you like our binary files? They are more compact!" ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 19 Date: Fri, 28 Feb 2003 21:56:30 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046489277 209.96.179.96 (Fri, 28 Feb 2003 22:27:57 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 22:27:57 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130763 alt.sys.pdp10:12681 In article , Peter da Silva wrote: >>Fortunately, I was just emailed the incantation to turn that behavior off: >> (setq inhibit-eol-conversion t) > > Is there a (setq inhibit-stupid-defaults t) ? Emacs has gotten really bad over the years about having new features default to true. For those of us with 10+ year old configuration files we like just fine the way they are, it's a mess. A few years ago I started trying to confirm to the "new way" and things are all messed up now. I keep having to turn stuff off! Bah! ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D1B07.63890CF4@yahoo.com> <3E5D6393.6B738216@yahoo.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 23 Date: Fri, 28 Feb 2003 22:01:10 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046492784 209.96.179.96 (Fri, 28 Feb 2003 23:26:24 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 23:26:24 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!feed2.news.rcn.net!rcn!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130765 alt.sys.pdp10:12682 In article , Mark Crispin wrote: >> So? It's preserving characteristics of the file. If you don't >> want that, pass the file through a filter to convert line endings >> first. > > Once upon a time, EMACS was usable by default without requiring one to > pass through filters first. That's the whole point. Why should I have to > use a filter instead of using the EMACS commands that I have used for > the past 25 years? You don't have to use filters or change anything as far as I know. I have the latest emacs and don't see the problem. I think it is reasonable to assume that if you load a file in DOS format, you want to keep it in that format. To me, quietly changing it to UNIX or other format behind my back would be far more annoying. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <3E57B900.4030601@jetnet.ab.ca> <3E5B170F.90501@beagle-ears.com> <3E5ECCE1.7040709@bellatlantic.net> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 11 Date: Fri, 28 Feb 2003 22:13:10 -0500 NNTP-Posting-Host: 209.96.179.96 X-Trace: sydney.visi.net 1046492785 209.96.179.96 (Fri, 28 Feb 2003 23:26:25 EST) NNTP-Posting-Date: Fri, 28 Feb 2003 23:26:25 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!peer.news.opaltelecom.net!newsfeed.icl.net!newsfeed.fjserv.net!newsfeed.icl.net!newsfeed.fjserv.net!newshosting.com!news-xfer2.atl.newshosting.com!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:130753 alt.sys.pdp10:12678 In article , jmfbahciv@aol.com wrote: >>Richie could always always find a way to bum code. >>I think that was what we called it. > > Yup. That's what we called it, too. I remember a few people in the 6502 era calling it "page sixing" your code, which was just a reference to cramming as much as you could into page 6 memory. ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 1 Mar 2003 03:24:41 GMT Organization: TSS Inc. Lines: 11 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <6k4i3b.hhc2.ln@via.reistad.priv.no> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046489081 19117 10.0.0.43 (1 Mar 2003 03:24:41 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 1 Mar 2003 03:24:41 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!nntp-relay.ihug.net!ihug.co.nz!news-out.nuthinbutnews.com!propagator2-sterling!news-in-sterling.newsfeed.com!news-in.nuthinbutnews.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130697 alt.sys.pdp10:12667 In article , Charles Shannon Hendrix wrote: >NetBSD is the last true UNIX... :) You misspelled "Plan 9". -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 1 Mar 2003 03:26:54 GMT Organization: TSS Inc. Lines: 18 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1281.186T405T7773222@kltpzyxm.invalid> <3E5D4816.C974A441@yahoo.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046489214 19640 10.0.0.43 (1 Mar 2003 03:26:54 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 1 Mar 2003 03:26:54 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!24.215.0.32.MISMATCH!newsfeed.mountaincable.net!news-out.nuthinbutnews.com!propagator2-sterling!news-in-sterling.newsfeed.com!news-in.nuthinbutnews.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130695 alt.sys.pdp10:12666 In article , Charles Shannon Hendrix wrote: >This is how you know that people like Bill Gates are not true geeks, >computer scientists, etc. If I were a billionaire, it wouldn't >be wasted on a 12-meter yacht (maybe a tallship though) or a >huge mansion. > >It would be an underground fallout shelter with Multics machines, >PDP-11s, S/370s. His huge mansion is partly underground, and he has an Alphaserver running a -10 emulator. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 1 Mar 2003 03:29:15 GMT Organization: TSS Inc. Lines: 19 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5FDA7E.E69A472A@yahoo.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046489355 19684 10.0.0.43 (1 Mar 2003 03:29:15 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 1 Mar 2003 03:29:15 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!nntp-out.monmouth.com!newspeer.monmouth.com!newsfeed.mathworks.com!nycmny1-snh1.gtei.net!news.gtei.net!bloom-beacon.mit.edu!micro-heart-of-gold.mit.edu!news-out.nuthinbutnews.com!propagator2-sterling!news-in-sterling.newsfeed.com!news-in.nuthinbutnews.com!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130692 alt.sys.pdp10:12665 In article <3E5FDA7E.E69A472A@yahoo.com>, Peter Flass wrote: >Worrying about it was really overkill. If you were porting C, or when >you ported FORTRAN, you wouldn't worry about the size of an int >(usually). You'd just take whatever the machine gives you. You'd get >the same result with PL/I just converting all the FIXED BIN(35) to FIXED >BIN(31). No, that didn't work. Some of them really had to be that big. Really, they should have just said "fixed binary" if they didn't care. But that (35) stuff is just too bleeding tempting for some people. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### Message-ID: <3E603711.F262F358@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5FDA7E.E69A472A@yahoo.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 28 Date: Sat, 01 Mar 2003 04:48:13 GMT NNTP-Posting-Host: 12.90.167.91 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc05-news.ops.worldnet.att.net 1046494093 12.90.167.91 (Sat, 01 Mar 2003 04:48:13 GMT) NNTP-Posting-Date: Sat, 01 Mar 2003 04:48:13 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!wn14feed!wn13feed!worldnet.att.net!bgtnsc05-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130772 alt.sys.pdp10:12684 ararghNOSPAM@NOT.AT.enteract.com wrote: > Peter Flass wrote: > ... snip ... > > > > Worrying about it was really overkill. If you were porting C, > > or when you ported FORTRAN, you wouldn't worry about the size > > of an int (usually). You'd just take whatever the machine > > gives you. You'd get the same result with PL/I just > > converting all the FIXED BIN(35) to FIXED BIN(31). > > Always assumming that the program doesn't expect to store values > > 2^31 and then expect to get back the same number. :-) > > I have never done this type of conversion, but I would expect to > spend FAR more time checking the program for varable size > assumptions, than I would spend time in the conversion. Showing the foolishness of such a type spec. in the first place. A Pascal sub-range would simply specify the min..max for a type, and all is well and portable. It is now up to the compiler to make any storage economies it wishes (or doesn't). -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### Message-ID: <3E605F87.A8035B31@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5FB200.673AA397@ev1.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 36 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc52.ops.asp.att.net 1046496204 12.241.15.59 (Sat, 01 Mar 2003 05:23:24 GMT) NNTP-Posting-Date: Sat, 01 Mar 2003 05:23:24 GMT Date: Sat, 01 Mar 2003 05:23:24 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!peernews-us.colt.net!newsfeed.news2me.com!cyclone1.gnilink.net!wn14feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!rwcrnsc52.ops.asp.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130737 alt.sys.pdp10:12675 Peter da Silva wrote: > > Charles Richmond wrote: > > One might email Doug Englebart and let him know that Smalltalk > > originated the GUI idea. > > I know about his work in hypertext and graphical interfaces, but I > wasn't aware that he'd come up with windows as well. All the clips > and movies I've seen of his software present a single document > taking up the whole screen. I think I recall a split-screen > display, but nothing like the Xerox windows. > Xerox windows pioneered the "overlapping" windows IIRC. Doug Englebart used "pained" windows. Apple pioneered direct manipulation of the windows. IOW, with Xerox windows, you had to pop up a menu and select "drag" before you could move a window. Apple had the "dragger bar" at the top...you just grabbed on with the mouse and dragged the window. > > That's why I said "windowing gui" and not just "gui". > See above... > > Anyway, if I'm mistaken about that, then I apologise, and I'd > really appreciate a link or something. > No, you have the right idea IMHO. But I do *not* think that you should omit Doug Englebart's contribution. IIRC, Englebart is where Alan Kay and the others at PARC got the beginnings of their GUI ideas from. And Englebart really did patent the mouse in 1970...US Patent No. 3,541,541. -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 1 Mar 2003 06:19:37 GMT Organization: TSS Inc. Lines: 30 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5FB200.673AA397@ev1.net> <3E605F87.A8035B31@ev1.net> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046499577 28836 10.0.0.43 (1 Mar 2003 06:19:37 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 1 Mar 2003 06:19:37 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!news.gtei.net!bloom-beacon.mit.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130698 alt.sys.pdp10:12668 In article <3E605F87.A8035B31@ev1.net>, Charles Richmond wrote: >Xerox windows pioneered the "overlapping" windows IIRC. Doug >Englebart used "pained" windows. I don't know about calling them 'windows'. Paned windows of the type I've seen are basically split-screen text, and that's one of those IBM 3270 type things that goes back to the very first video consoles. >Apple pioneered direct >manipulation of the windows. IOW, with Xerox windows, you >had to pop up a menu and select "drag" before you could move >a window. In Smalltalk, yes. Don't forget the Xerox Star, though. >No, you have the right idea IMHO. But I do *not* think that >you should omit Doug Englebart's contribution. Or Vannevar Bush, Ted Nelson, and all the rest. I don't think making a rhetorical point about Smalltalk's failure to take the step from being the first to standardise on a gui to having a standard gui really counts as "ignoring" every other point I might have made that simply didn't relate to the one I was trying to make. :) -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: stanb45@dial.pipex.com (Stan Barr) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <1281.186T405T7773222@kltpzyxm.invalid> <3E5D32A5.89915AB7@ev1.net> Organization: Metropolis Grafix Reply-To: stanb45@dial.pipex.com Message-ID: X-Newsreader: slrn (0.9.5.2 UNIX) Date: 01 Mar 2003 09:39:09 GMT Lines: 24 NNTP-Posting-Host: 62-241-190-242.dsl.pipex.com X-Trace: 1046511549 news.dial.pipex.com 6299 62.241.190.242 X-Complaints-To: abuse@uk.uu.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!solnet.ch!solnet.ch!newsfeed.freenet.de!feed.news.nacamar.de!easynet-melon!easynet.net!newsfeed.kabelfoon.nl!195.129.110.21.MISMATCH!bnewsfeed00.bru.ops.eu.uu.net!bnewsifeed00.bru.ops.eu.uu.net!lnewspost00.lnd.ops.eu.uu.net!emea.uu.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130635 On Fri, 28 Feb 2003 21:29:03 -0500, Charles Shannon Hendrix wrote: >In article <3E5D32A5.89915AB7@ev1.net>, Charles Richmond wrote: >>> You are a nasty cynical man. I can't help but approve. >>> >> "The power of accurate observation is commonly called cynicism >> by those who have not got it." >> --George Bernard Shaw > > A cynic is an idealist whose rose colored glasses have been removed, > snapped in two, and stomped into the ground, immediately improving > his vision. A cynic is what an idealist calls a realist. Sir Humphrey Appleby -- Cheers, Stan Barr stanb45@dial.pipex.com **Remove the digits from email address** The future was never like this! ###### Message-ID: <3E60AEEB.4FBB6FD2@yahoo.com> From: Peter Flass X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 24 Date: Sat, 01 Mar 2003 13:03:41 GMT NNTP-Posting-Host: 24.194.50.82 X-Complaints-To: abuse@rr.com X-Trace: twister.nyroc.rr.com 1046523821 24.194.50.82 (Sat, 01 Mar 2003 08:03:41 EST) NNTP-Posting-Date: Sat, 01 Mar 2003 08:03:41 EST Organization: Road Runner Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed-east.nntpserver.com!nntpserver.com!news-west.rr.com!cyclone.nyroc.rr.com!cyclone-out.nyroc.rr.com!twister.nyroc.rr.com.POSTED!53ab2750!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130582 Charles Shannon Hendrix wrote: > > In article <3E5A710D.6030403@jetnet.ab.ca>, bfranchuk@jetnet.ab.ca wrote: > > > It is said only about 10? lines of bug free code > > can be produced a day with all the docs and other > > needed information. > > I never have believed generalizations like this. > > A lot depends on what you are writing, and how complex it is. > > Certainly I've been on projects where we'd put on a ton of code, and > even after a year or more, our initial coding rate was far better than > any of these rules-of-thumb I've heard. > Back in my application programming days we were programming a new accounting system. My wife was taking evening courses in grad school. I'd drive her to class and drag along a COBOL pad and the program specs. By the end of an hour I'd have a program that usually only needed another hour or so of test and cleanup to work correctly -- say maybe 500-750 lines. OTOH I've done parsers that took weeks for the same number of LOC. Of course the 10 LOC *is* an average. ###### Message-ID: <3E60B143.71DC2839@yahoo.com> From: Peter Flass X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5FDA7E.E69A472A@yahoo.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 20 Date: Sat, 01 Mar 2003 13:13:42 GMT NNTP-Posting-Host: 24.194.50.82 X-Complaints-To: abuse@rr.com X-Trace: twister.nyroc.rr.com 1046524422 24.194.50.82 (Sat, 01 Mar 2003 08:13:42 EST) NNTP-Posting-Date: Sat, 01 Mar 2003 08:13:42 EST Organization: Road Runner Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed-west.nntpserver.com!hub1.meganetnews.com!nntpserver.com!news-west.rr.com!cyclone.nyroc.rr.com!cyclone-out.nyroc.rr.com!twister.nyroc.rr.com.POSTED!53ab2750!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130773 alt.sys.pdp10:12685 ararghNOSPAM@NOT.AT.enteract.com wrote: > >Worrying about it was really overkill. If you were porting C, or when > >you ported FORTRAN, you wouldn't worry about the size of an int > >(usually). You'd just take whatever the machine gives you. You'd get > >the same result with PL/I just converting all the FIXED BIN(35) to FIXED > >BIN(31). > Always assumming that the program doesn't expect to store values > > 2^31 and then expect to get back the same number. :-) > > I have never done this type of conversion, but I would expect to spend > FAR more time checking the program for varable size assumptions, than > I would spend time in the conversion. Right, but my point is that you have the same problem with C, FORTRAN, or any language. Just because PL/I gives you the ability to control the size of an int, or INTEGER, doesn't mean you need or want to use it. Actually, IBM blew it by making 16-bits the default for FIXED BINARY data. The default is supposed to be the machine's word size, like C's int datatype. PL/I's 'DECLARE a FIXED BINARY' should be the equivalent of C's 'int a'. ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 1 Mar 2003 16:55:04 -0000 Organization: Titanic Enterprises Unlimited Lines: 46 Message-ID: <1046537704.132366@saucer.planet.gong> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5FDA7E.E69A472A@yahoo.com> <3E60B143.71DC2839@yahoo.com> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1046537705 11318 80.177.7.220 (1 Mar 2003 16:55:05 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Sat, 1 Mar 2003 16:55:05 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Cache: nntpcache 3.0.1 (see http://www.nntpcache.org/) X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Cache-Post-Path: saucer.planet.gong!unknown@voodoo.planet.gong Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!feed.news.nacamar.de!easynet-quince!easynet.net!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130782 alt.sys.pdp10:12686 "Peter Flass" wrote in message news:3E60B143.71DC2839@yahoo.com... > ararghNOSPAM@NOT.AT.enteract.com wrote: > > >Worrying about it was really overkill. If you were porting C, or when > > >you ported FORTRAN, you wouldn't worry about the size of an int > > >(usually). You'd just take whatever the machine gives you. You'd get > > >the same result with PL/I just converting all the FIXED BIN(35) to FIXED > > >BIN(31). > > Always assumming that the program doesn't expect to store values > > > 2^31 and then expect to get back the same number. :-) > > > > I have never done this type of conversion, but I would expect to spend > > FAR more time checking the program for varable size assumptions, than > > I would spend time in the conversion. > > Right, but my point is that you have the same problem with C, FORTRAN, > or any language. Just because PL/I gives you the ability to control the > size of an int, or INTEGER, doesn't mean you need or want to use it. I've found that it's a good idea to establish hard & fast word sizes & subscripts simply because they are another form of documentation. They define the parameters that a piece of code is designed to handle. All too often this sort of stuff is missed in a porting effort which often causes lots of pain tracking the bugs. Porting code like that is about 1000x easier in my experience, simply because you know it will behave exactly the same after the port... Hmm... Well you know what I mean - ideal world and all that. :) > Actually, IBM blew it by making 16-bits the default for FIXED BINARY > data. The default is supposed to be the machine's word size, like C's > int datatype. PL/I's 'DECLARE a FIXED BINARY' should be the equivalent > of C's 'int a'. I don't have a big problem with that. Use the type that fits your problem, screw the machine word size... If you really want to do that use a macro-preprocessor to sub whatever size you want in (dunno enough about PL/I to offer a neater solution). Cheers, Rupert ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 1 Mar 2003 20:38:05 GMT Organization: TSS Inc. Lines: 24 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5FDA7E.E69A472A@yahoo.com> <3E60B143.71DC2839@yahoo.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046551085 76915 10.0.0.43 (1 Mar 2003 20:38:05 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 1 Mar 2003 20:38:05 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130793 alt.sys.pdp10:12687 In article <3E60B143.71DC2839@yahoo.com>, Peter Flass wrote: >Right, but my point is that you have the same problem with C, FORTRAN, >or any language. Yah, but C, FORTRAN, and other similar languages don't have that particular "press to detonate" button. They have others (the C preprocessor is particularly tempting for an overly creatve programmer), but few have as big an impact on portability. > The default is supposed to be the machine's word size, like C's >int datatype. PL/I's 'DECLARE a FIXED BINARY' should be the equivalent >of C's 'int a'. But in the '70s C's "int a" was also 16 bits. The difference is that in C, you didn't write "int a:32", you wrote "long a". -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: Ben Hutchings Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sun, 2 Mar 2003 03:24:50 +0000 Lines: 14 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Reply-To: Ben Hutchings NNTP-Posting-Host: pc5-cmbg1-6-cust149.cmbg.cable.ntl.com (80.4.127.149) X-Trace: fu-berlin.de 1046575805 60143878 80.4.127.149 (16 [70929]) X-Orig-Path: 192.168.1.1!nobody User-Agent: slrn/0.9.7.4 (Linux) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!uni-erlangen.de!fu-berlin.de!uni-berlin.de!pc5-cmbg1-6-cust149.cmbg.cable.ntl.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130874 In article , Charles Shannon Hendrix wrote: > A PC BIOS can only boot an image up to 2.88MB, the size of the highest > density 3.5" floppy. This is true even for a CD boot. The boot image > on the CD is also limited to 2.88MB. A PC BIOS only reads the first sector off the boot disk. It's up to the code squeezed into those 512 bytes (446 in the case of a hard disk, as 66 bytes are used for the partition table) to tell it to read additional sectors. There is a limit on the extent of a disk accessible through the BIOS routines, but it's much higher than 2.88MB. ###### User-Agent: Microsoft-Entourage/10.1.1.2418 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) From: nospam Newsgroups: alt.folklore.computers,alt.sys.pdp10 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5FDA7E.E69A472A@yahoo.com> <3E60B143.71DC2839@yahoo.com> Mime-version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit Lines: 37 Date: Sun, 02 Mar 2003 03:46:45 GMT NNTP-Posting-Host: 144.136.91.243 X-Complaints-To: news@bigpond.net.au X-Trace: news-server.bigpond.net.au 1046576805 144.136.91.243 (Sun, 02 Mar 2003 14:46:45 EST) NNTP-Posting-Date: Sun, 02 Mar 2003 14:46:45 EST Organization: BigPond Internet Services (http://www.bigpond.net.au) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!newsfeed01.tsnz.net!ken-transit.news.telstra.net!news.telstra.net!news-server.bigpond.net.au!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130850 alt.sys.pdp10:12690 in article 3E60B143.71DC2839@yahoo.com, Peter Flass at peter_flass@yahoo.com wrote on 02/03/2003 00:13: > ararghNOSPAM@NOT.AT.enteract.com wrote: >>> Worrying about it was really overkill. If you were porting C, or when >>> you ported FORTRAN, you wouldn't worry about the size of an int >>> (usually). You'd just take whatever the machine gives you. You'd get >>> the same result with PL/I just converting all the FIXED BIN(35) to FIXED >>> BIN(31). >> Always assumming that the program doesn't expect to store values > >> 2^31 and then expect to get back the same number. :-) >> >> I have never done this type of conversion, but I would expect to spend >> FAR more time checking the program for varable size assumptions, than >> I would spend time in the conversion. > > Right, but my point is that you have the same problem with C, FORTRAN, > or any language. Just because PL/I gives you the ability to control the > size of an int, or INTEGER, doesn't mean you need or want to use it. > Actually, IBM blew it by making 16-bits the default for FIXED BINARY > data. The default is supposed to be the machine's word size, like C's C, its all stuffed up. Int has become defacto standard 32bits. Most UNIX programmers esp those from the linux camp (aka young) think long is 32 bits. All VMS programmers use long instead of int cause they think its the same long that is referred to by system services/macro-32. Digitals C (UNIX) compiler assumes long is the word size of the alpha 64bits. Which all though I think is correct screwes up both the code from the unix camp who think long is 32bits and those VMS C programmers that think long is 32 bits. GNU and other compilers long long would be 64 bits. Its good coding style to use int32 int64 or other such typedefs instead of assuming a word length. Most of the programmers alive now I imagine have not programmed anything other than a 32bit machines, maybe an 8bit machines when they really young:) > int datatype. PL/I's 'DECLARE a FIXED BINARY' should be the equivalent > of C's 'int a'. ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 2 Mar 2003 04:29:43 GMT Organization: TSS Inc. Lines: 22 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E60B143.71DC2839@yahoo.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046579383 1716 10.0.0.43 (2 Mar 2003 04:29:43 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 2 Mar 2003 04:29:43 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!headwall.stanford.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130840 alt.sys.pdp10:12688 In article , nospam wrote: >C, its all stuffed up. Int has become defacto standard 32bits. Don't blame C, blame the VAX. To be precise, blame the decision not to increase the size of long on the VAX to 64 bits, like it should have been from the start. It's a long integer, something longer than integer. So what if it's 2 words. It was 2 words on the PDP-11 and that was no problem. On the alpha, int should be 64 bits, long should be 128, and short should remain 16. If you want 32 bits, than you should be using a derived type that is based on something ugly like: _hardware_dependent_int_32 -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### User-Agent: Microsoft-Entourage/10.1.1.2418 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) From: nospam Newsgroups: alt.folklore.computers,alt.sys.pdp10,alt.os.multics Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <1281.186T405T7773222@kltpzyxm.invalid> <3E5D4816.C974A441@yahoo.com> Mime-version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7bit Lines: 43 Date: Sun, 02 Mar 2003 05:40:22 GMT NNTP-Posting-Host: 144.136.91.243 X-Complaints-To: news@bigpond.net.au X-Trace: news-server.bigpond.net.au 1046583622 144.136.91.243 (Sun, 02 Mar 2003 16:40:22 EST) NNTP-Posting-Date: Sun, 02 Mar 2003 16:40:22 EST Organization: BigPond Internet Services (http://www.bigpond.net.au) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!newsfeed1.bredband.com!bredband!news.algonet.se!algonet!ken-transit.news.telstra.net!news.telstra.net!news-server.bigpond.net.au!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130848 alt.sys.pdp10:12689 in article slrnb607b4.4eu.shannon@news.widomaker.com, Charles Shannon Hendrix at shannon@news.widomaker.com wrote on 01/03/2003 13:39: > >> Actually, Multics was designed to be portable, and in fact *was* ported >> from a GE-645 to a Honeywell ... something, a similar machine, but with >> some significant differences. The only reason it wasn't more widely >> ported is that most current hardware lacks the support for segmentation >> and rings that it requires. > I thought some modern CPUs had enough to handle the rings, and the> > segments could be done on top of existing MMUs, if the CPU is a purely > flat system. I think emulator in software is going to be the ticket here, though from what I've read there is a big hole in hardware documentation and a lot of guesses from the source code are going to make this a harder task than it already is. > This is how you know that people like Bill Gates are not true geeks, > computer scientists, etc. If I were a billionaire, it wouldn't > be wasted on a 12-meter yacht (maybe a tallship though) or a > huge mansion. > > It would be an underground fallout shelter with Multics machines, > PDP-11s, S/370s. Not sure 11's would be high on my list, certainly a PDP-10 would be. I did once tell the wife that a concrete bunker would suite for the new home/office security and aircon over views etc. She never liked that idea and I'm sure too never live it down ;) > > Wealthy people are among the most dull and boring people on Earth. Well I kind of like sailing - so I'd not consider people that do are either dull or boring. Geek has never been my favourite word anyway. I thought it was used in a derogatory sense and didn't know it had become a desirable moniker. I would rather the dull and boring title esp if it comes with the yacht and cash :) Its Larry from oracle that sails, I don't think Mr Gates is really into sailing. I would argue that BG is the one TRUE Geek ;) and that geeks are dull and boring ;) Cheers Mark ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: <0pl36voeftkkppsh19q8oped2bankd1lc9@4ax.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 32 Date: Sun, 02 Mar 2003 10:09:53 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news3.calgary.shaw.ca 1046599793 24.71.223.147 (Sun, 02 Mar 2003 03:09:53 MST) NNTP-Posting-Date: Sun, 02 Mar 2003 03:09:53 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!nntp.theplanet.net!inewsm1.nntp.theplanet.net!195.40.4.120.MISMATCH!easynet-quince!easynet.net!cox.net!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news3.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130869 On Sun, 2 Mar 2003 03:24:50 +0000 in alt.folklore.computers, Ben Hutchings wrote: >In article , >Charles Shannon Hendrix wrote: > >> A PC BIOS can only boot an image up to 2.88MB, the size of the highest >> density 3.5" floppy. This is true even for a CD boot. The boot image >> on the CD is also limited to 2.88MB. > > >A PC BIOS only reads the first sector off the boot disk. It's up to >the code squeezed into those 512 bytes (446 in the case of a hard >disk, as 66 bytes are used for the partition table) to tell it to read >additional sectors. There is a limit on the extent of a disk >accessible through the BIOS routines, but it's much higher than >2.88MB. AFAIR the PC BIOS can only access the first 1024 cylinders of an EIDE drive, so the start of any bootable partition and the system bootstrap (e.g. IBMBIO, LILO, etc.) must reside in that space; but a system bootstrap like LILO can switch an EIDE drive to LBA mode and load the system from anywhere on disk. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5FDA7E.E69A472A@yahoo.com> <3E60B143.71DC2839@yahoo.com> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 57 Date: Sun, 02 Mar 2003 10:17:59 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news3.calgary.shaw.ca 1046600279 24.71.223.147 (Sun, 02 Mar 2003 03:17:59 MST) NNTP-Posting-Date: Sun, 02 Mar 2003 03:17:59 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!nntp.theplanet.net!inewsm1.nntp.theplanet.net!195.40.4.120.MISMATCH!easynet-quince!easynet.net!cox.net!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news3.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130866 alt.sys.pdp10:12692 On Sun, 02 Mar 2003 03:46:45 GMT in alt.folklore.computers, nospam wrote: >in article 3E60B143.71DC2839@yahoo.com, Peter Flass at peter_flass@yahoo.com >wrote on 02/03/2003 00:13: > >> ararghNOSPAM@NOT.AT.enteract.com wrote: >>>> Worrying about it was really overkill. If you were porting C, or when >>>> you ported FORTRAN, you wouldn't worry about the size of an int >>>> (usually). You'd just take whatever the machine gives you. You'd get >>>> the same result with PL/I just converting all the FIXED BIN(35) to FIXED >>>> BIN(31). >>> Always assumming that the program doesn't expect to store values > >>> 2^31 and then expect to get back the same number. :-) >>> >>> I have never done this type of conversion, but I would expect to spend >>> FAR more time checking the program for varable size assumptions, than >>> I would spend time in the conversion. >> >> Right, but my point is that you have the same problem with C, FORTRAN, >> or any language. Just because PL/I gives you the ability to control the >> size of an int, or INTEGER, doesn't mean you need or want to use it. >> Actually, IBM blew it by making 16-bits the default for FIXED BINARY >> data. The default is supposed to be the machine's word size, like C's >C, its all stuffed up. Int has become defacto standard 32bits. Most UNIX >programmers esp those from the linux camp (aka young) think long is 32 bits. >All VMS programmers use long instead of int cause they think its the same >long that is referred to by system services/macro-32. Digitals C (UNIX) >compiler assumes long is the word size of the alpha 64bits. Which all though >I think is correct screwes up both the code from the unix camp who think >long is 32bits and those VMS C programmers that think long is 32 bits. GNU >and other compilers long long would be 64 bits. Its good coding style to use >int32 int64 or other such typedefs instead of assuming a word length. it's good coding style not to make such assumptions as there is no portable way to guarantee that anything applies other than the Standard: char >= 8 bits; short >= 16 bits, int >= short, long >= int; all else is a possibly incorrect assumption. > Most of the programmers alive now I imagine have not programmed anything >other than a 32bit machines, maybe an 8bit machines when they really young:) Oh, I think a few probably programmed 16 bitters like the PDP-11 and the i86. >> int datatype. PL/I's 'DECLARE a FIXED BINARY' should be the equivalent >> of C's 'int a'. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### Message-ID: <3E6200D6.9100C120@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5F0C8C.1080900@jetnet.ab.ca> <2607.189T698T6883699@kltpzyxm.invalid> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 56 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: sccrnsc03 1046603037 12.241.15.59 (Sun, 02 Mar 2003 11:03:57 GMT) NNTP-Posting-Date: Sun, 02 Mar 2003 11:03:57 GMT Date: Sun, 02 Mar 2003 11:03:57 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!news.webusenet.com!cyclone1.gnilink.net!wn14feed!wn12feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi_feed4!attbi.com!sccrnsc03.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130860 alt.sys.pdp10:12691 Charlie Gibbs wrote: > > [snip...] [snip...] [snip...] > > I went to our local cell phone dealer, wanting to see whether I could > find a data cable for my wife's cell phone. I took my laptop along, > because I would only be convinced by making it work on my hardware. > (I did this when I got my own cell phone, which is why I still have > that ancient analog-only Motorola MicroTAC 650. It's not sexy, but > it does the work. My priorities are funny that way.) > > I plugged their cable into my wife's phone and my laptop's serial > port, fired up minicom, and started typing. Nothing happened. > "Oh, that's because you haven't installed the software," the sales > droid said, offering me a CD-ROM. It took a while to convince him > that not all laptops are running Windows; he probably had never seen > Linux before. Thing got even more interesting when I pulled out my > breakout box, cables, and adapters, and started stringing everything > together on his counter. Their so-called "techs" were staring at me > as if I was from another planet. I don't think any of them even knew > what a breakout box was. > In the movie "Sleepless in Seattle", someone at a meeting said that a woman over 30 had a better chance of being killed by a terrorist than she did getting married. And one woman said: "That's *not* true!!! But it feels like it's true..." Mi$uck Windows and PC's are *not* the only computer setup... but it feels like it is the only computer setup. If you don't do things like everyone else does them, then you are by definition a madman!!! After all, a sane person buys what is on the counter for sale...and keeps his mouth shut!!! ;-) Again I post the Emily Dickinson poem: Much madness is divinest sense- To a discerning Eye- Much Sense - the Starkest Madness- 'Tis the Majority In this as All, prevail- Assent - and you are sane- Demur - you're straightway dangerous- And handled with a Chain- > > I never did get the interface to work. To this day I'm sure they > "know" it's because I simply didn't have the correct software. > Do you mean that your ancient laptop did *not* cause any concern??? They probably figured that you have been listening to AM radio... like BAH has admitted that she does (when her stove is unplugged). -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### From: Ben Hutchings Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sun, 2 Mar 2003 13:40:48 +0000 Lines: 36 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <0pl36voeftkkppsh19q8oped2bankd1lc9@4ax.com> Reply-To: Ben Hutchings NNTP-Posting-Host: pc5-cmbg1-6-cust149.cmbg.cable.ntl.com (80.4.127.149) X-Trace: fu-berlin.de 1046612705 60671532 80.4.127.149 (16 [70929]) X-Orig-Path: 192.168.1.1!nobody User-Agent: slrn/0.9.7.4 (Linux) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!pc5-cmbg1-6-cust149.cmbg.cable.ntl.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:130870 In article <0pl36voeftkkppsh19q8oped2bankd1lc9@4ax.com>, Brian Inglis wrote: > On Sun, 2 Mar 2003 03:24:50 +0000 in alt.folklore.computers, Ben > Hutchings wrote: > >>In article , >>Charles Shannon Hendrix wrote: >> >>> A PC BIOS can only boot an image up to 2.88MB, the size of the highest >>> density 3.5" floppy. This is true even for a CD boot. The boot image >>> on the CD is also limited to 2.88MB. >> >> >>A PC BIOS only reads the first sector off the boot disk. It's up to >>the code squeezed into those 512 bytes (446 in the case of a hard >>disk, as 66 bytes are used for the partition table) to tell it to read >>additional sectors. There is a limit on the extent of a disk >>accessible through the BIOS routines, but it's much higher than >>2.88MB. > > AFAIR the PC BIOS can only access the first 1024 cylinders of an > EIDE drive, so the start of any bootable partition and the system > bootstrap (e.g. IBMBIO, LILO, etc.) must reside in that space; > but a system bootstrap like LILO can switch an EIDE drive to LBA > mode and load the system from anywhere on disk. LILO cannot control arbitary disk hardware directly, so it always uses BIOS calls to load and can't make any switch like that. The original INT 13 interface still allows about 8 GB of the disk to be read, assuming the BIOS does appropriate remapping of addresses (which they have been doing since the mid-90s). There is an "extended INT 13" interface that allows callers to give linear addresses. I don't know whether LILO knows how to use this extension. -- Ben Hutchings | personal web site: http://womble.decadentplace.org.uk/ Anthony's Law of Force: Don't force it, get a larger hammer. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> Reply-To: shannon@widomaker.com From: shannon@daydream.shannon.net (Charles Shannon Hendrix) Message-ID: <0kd04b.lsg.ln@escape.shannon.net> Lines: 15 Date: Mon, 3 Mar 2003 15:23:28 -0500 NNTP-Posting-Host: 206.246.249.97 X-Trace: sydney.visi.net 1046726900 206.246.249.97 (Mon, 03 Mar 2003 16:28:20 EST) NNTP-Posting-Date: Mon, 03 Mar 2003 16:28:20 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!nntp.abs.net!newsxfer.visi.net!sydney.visi.net!daydream.shannon.net!shannon Xref: chonsp.franklin.ch alt.folklore.computers:130942 alt.sys.pdp10:12694 In article , Peter da Silva wrote: > In article , > Charles Shannon Hendrix wrote: > >NetBSD is the last true UNIX... :) > > You misspelled "Plan 9". MMmmmm... how could I forget. -- UNIX/Perl/C/Pizza____________________s h a n n o n@wido !SPAM maker.com ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D4816.C974A441@yahoo.com> Reply-To: shannon@widomaker.com From: shannon@daydream.shannon.net (Charles Shannon Hendrix) Message-ID: Lines: 23 Date: Mon, 3 Mar 2003 15:24:53 -0500 NNTP-Posting-Host: 206.246.249.97 X-Trace: sydney.visi.net 1046726958 206.246.249.97 (Mon, 03 Mar 2003 16:29:18 EST) NNTP-Posting-Date: Mon, 03 Mar 2003 16:29:18 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsswitch.lcs.mit.edu!newsxfer.visi.net!sydney.visi.net!daydream.shannon.net!shannon Xref: chonsp.franklin.ch alt.folklore.computers:130946 alt.sys.pdp10:12695 In article , Peter da Silva wrote: > In article , > Charles Shannon Hendrix wrote: > >This is how you know that people like Bill Gates are not true geeks, > >computer scientists, etc. If I were a billionaire, it wouldn't > >be wasted on a 12-meter yacht (maybe a tallship though) or a > >huge mansion. > > > >It would be an underground fallout shelter with Multics machines, > >PDP-11s, S/370s. > > His huge mansion is partly underground, and he has an Alphaserver running > a -10 emulator. Didn't think Larry Ellison would be interested in that sort of thing. Oh, you mean Bill Gates? He ever use it? -- UNIX/Perl/C/Pizza____________________s h a n n o n@wido !SPAM maker.com ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10,alt.os.multics Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D4816.C974A441@yahoo.com> Reply-To: shannon@widomaker.com From: shannon@daydream.shannon.net (Charles Shannon Hendrix) Message-ID: <6ud04b.lah.ln@escape.shannon.net> Lines: 47 Date: Mon, 3 Mar 2003 15:28:54 -0500 NNTP-Posting-Host: 206.246.249.97 X-Trace: sydney.visi.net 1046727006 206.246.249.97 (Mon, 03 Mar 2003 16:30:06 EST) NNTP-Posting-Date: Mon, 03 Mar 2003 16:30:06 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!feedme.news.mediaways.net!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!daydream.shannon.net!shannon Xref: chonsp.franklin.ch alt.folklore.computers:130941 alt.sys.pdp10:12693 In article , nospam wrote: > I think emulator in software is going to be the ticket here, though from > what I've read there is a big hole in hardware documentation and a lot of > guesses from the source code are going to make this a harder task than it > already is. Seems like enough heads would be able to form a good collective memory. Why did the hardware documentation go away? > > It would be an underground fallout shelter with Multics machines, > > PDP-11s, S/370s. > Not sure 11's would be high on my list, certainly a PDP-10 would be. I did > once tell the wife that a concrete bunker would suite for the new > home/office security and aircon over views etc. She never liked that idea > and I'm sure too never live it down ;) My list wasn't complete of course. I left out the Crays, the Cyber 205 that would be a must, and a few others. Of course, I'd probably stick to stuff I wanted to play with. Toys are no fun just sitting in the closet. > > Wealthy people are among the most dull and boring people on Earth. > Well I kind of like sailing - so I'd not consider people that do are either > dull or boring. No, I said 12-meter yacht. I didn't say sailing was boring... :) With that kind of money, I'd have something more on the order of 100 meters, with a jolly roger on top. > Its Larry from oracle that sails, I don't think Mr Gates is really into > sailing. I would argue that BG is the one TRUE Geek ;) and that geeks are > dull and boring ;) I have never thought of Bill Gates as a geek. His geeksploits are greatly exaggerated. His partner... he's a geek. No, not the dancing monkey. -- UNIX/Perl/C/Pizza____________________s h a n n o n@wido !SPAM maker.com ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 04 Mar 03 09:24:57 GMT Organization: UltraNet Communications, Inc. Lines: 17 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <0kd04b.lsg.ln@escape.shannon.net> X-Trace: UmFuZG9tSVZifN7W4Jpf+fLcrhFZhNHP481hxJc8/PjZdCbVdkZo2yOKE+dSKI8X X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 4 Mar 2003 10:09:24 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!newsfeed.gol.com!feed1.news.rcn.net!rcn!208-59-181-237 Xref: chonsp.franklin.ch alt.folklore.computers:130993 alt.sys.pdp10:12696 In article <0kd04b.lsg.ln@escape.shannon.net>, shannon@daydream.shannon.net (Charles Shannon Hendrix) wrote: >In article , >Peter da Silva wrote: >> In article , >> Charles Shannon Hendrix wrote: >> >NetBSD is the last true UNIX... :) >> >> You misspelled "Plan 9". > >MMmmmm... how could I forget. Some of us only count in octal. /BAH Subtract a hundred and four for e-mail. ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 04 Mar 03 13:37:01 -0800 Organization: http://extra.newsguy.com Lines: 33 Message-ID: <994.193T61T8173778@kltpzyxm.invalid> References: <3E05FE9D.B3A12286@vax6k.openecs.org><0kd04b.lsg.ln@escape.shannon.net> <20030304211555.4731fa87.steveo@eircom.net> NNTP-Posting-Host: p-935.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!diablo.theplanet.net!news.maxwell.syr.edu!cyclone.bc.net!logbridge.uoregon.edu!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews3 Xref: chonsp.franklin.ch alt.folklore.computers:131039 alt.sys.pdp10:12699 In article <20030304211555.4731fa87.steveo@eircom.net> steveo@eircom.net (Steve O'Hara-Smith) writes: > On Tue, 04 Mar 03 09:24:57 GMT > jmfbahciv@aol.com wrote: > >> In article <0kd04b.lsg.ln@escape.shannon.net>, >> shannon@daydream.shannon.net (Charles Shannon Hendrix) wrote: >> >>> In article , >>> Peter da Silva wrote: >>> >>>> In article , >>>> Charles Shannon Hendrix wrote: >>>> >>>>> NetBSD is the last true UNIX... :) >>>> >>>> You misspelled "Plan 9". >>> >>> MMmmmm... how could I forget. >> >> Some of us only count in octal. > > So how come there was a PDP8 ? Dunno, but it thinks it's a PDP-10. -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 05 Mar 03 11:35:15 GMT Organization: UltraNet Communications, Inc. Lines: 31 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <0kd04b.lsg.ln@escape.shannon.net> <20030304211555.4731fa87.steveo@eircom.net> X-Trace: UmFuZG9tSVY0JTxTFHqeEMZOnKwvSe/AKtgP4H1okR3jfq4exOHBuUIs09L01TlR X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 5 Mar 2003 12:19:54 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-102-34 Xref: chonsp.franklin.ch alt.folklore.computers:131007 alt.sys.pdp10:12697 In article <20030304211555.4731fa87.steveo@eircom.net>, Steve O'Hara-Smith wrote: >On Tue, 04 Mar 03 09:24:57 GMT >jmfbahciv@aol.com wrote: > >JC> In article <0kd04b.lsg.ln@escape.shannon.net>, >JC> shannon@daydream.shannon.net (Charles Shannon Hendrix) wrote: >JC> >In article , >JC> >Peter da Silva wrote: >JC> >> In article , >JC> >> Charles Shannon Hendrix wrote: >JC> >> >NetBSD is the last true UNIX... :) >JC> >> >JC> >> You misspelled "Plan 9". >JC> > >JC> >MMmmmm... how could I forget. >JC> >JC> Some of us only count in octal. > > So how come there was a PDP8 ? Smarty pants. Two possibilities come to mind: 1. It was our first off-by-one bug. 2. It was the first lesson in converting customers to octal. /BAH Subtract a hundred and four for e-mail. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10,alt.os.multics Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <6ud04b.lah.ln@escape.shannon.net> From: Morten Reistad X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: mrr@acer.reistad.priv.no (Morten Reistad) Message-ID: Lines: 104 Date: Tue, 4 Mar 2003 03:51:00 +0100 NNTP-Posting-Host: 80.111.42.204 X-Complaints-To: abuse@chello.no X-Trace: amstwist00 1046754002 80.111.42.204 (Tue, 04 Mar 2003 06:00:02 MET) NNTP-Posting-Date: Tue, 04 Mar 2003 06:00:02 MET Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news2.euro.net!amsnews01.chello.com!amstwist00.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131052 alt.sys.pdp10:12701 According to Charles Shannon Hendrix : >In article , >nospam wrote: > [snip] >> > It would be an underground fallout shelter with Multics machines, >> > PDP-11s, S/370s. >> Not sure 11's would be high on my list, certainly a PDP-10 would be. I did >> once tell the wife that a concrete bunker would suite for the new >> home/office security and aircon over views etc. She never liked that idea >> and I'm sure too never live it down ;) > >My list wasn't complete of course. I left out the Crays, the Cyber 205 >that would be a must, and a few others. > >Of course, I'd probably stick to stuff I wanted to play with. Toys are >no fun just sitting in the closet. You would need some power source for this; like a hydropower station or something. Also a good supply of useable spares. You may just as well find yourself a tropical island somwhere where there is rock, not just coral, and where the geological instability has waned. Pitcairn comes to mind. They will welcome you with open arms. >> > Wealthy people are among the most dull and boring people on Earth. >> Well I kind of like sailing - so I'd not consider people that do are either >> dull or boring. > >No, I said 12-meter yacht. I didn't say sailing was boring... :) > >With that kind of money, I'd have something more on the order of 100 >meters, with a jolly roger on top. A 12-meter is about as much as one person can handle all by him/herself without the need for fancy servos etc. that are never allowed in regattas anyway. This is a serious bit of fun that you should really try. It definatly distinguishes the sailing folks from the wannabees. (What I am thinking of something well rigged; with an 18m masthead or 7/8 rig, and > 60 m2 main+genoa sail area. Spinnaker of around 100m2. ) When you go for a 100 meter yacht you'll need a crew of around 10 just to move the thing out of port. You are not even allowed to move anything beyond ~82 feet all by yourself, even under engine power. There is amusingly little actual codified law regarding sail anywhere, so that is left to your judgement. There is an amusing twist to public regulations. This is a practice governed by a lot of "Best Current Practice", like the Gentlemans's club the Internet was before 1997. It is a common body of law for all seafaring powers, called Admirality Law. It is based on a long history of case law, with amazingly little that is less than a century old. Most of the "Best Current Practice" for movement under sail is actually written by captains Nelson, Cook, Bligh and their collegues. Nelson has written the standard works on rigging and how to improvise a broken rig. Cook has some pearls when it comes to moving in uncharted waters. The sailing records set by them and the clipper captains 50-80 years later was not beaten before around 1985. I wonder if someone from a similar distance in the future (mid 23rd century) will look back to multics, unix, As to the jolly roger, there are several administrations that would be only too happy to supply people in the affluence bracket of mr Gates and mr Ellison with a flag for a suitable amount of money. It is called "flag of convenience". A partial list of countries to explore follows : Antigua and Berbuda Germany Greece Hong Kong Liberia Netherlands Antilles Norway Panama Portugal's offshore islands Singapore Uruguay United Kingdom's crown posessions (Man, Channel Islands etc.) You will notice that around 95% of the world's international, commercial fleet is registered in one of these registries. >> Its Larry from oracle that sails, I don't think Mr Gates is really into >> sailing. I would argue that BG is the one TRUE Geek ;) and that geeks are >> dull and boring ;) > >I have never thought of Bill Gates as a geek. His geeksploits are >greatly exaggerated. This looks very much like a propaganda front. Play a geek for the geeks etc. > >His partner... he's a geek. No, not the dancing monkey. -- mrr ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers,alt.sys.pdp10,alt.os.multics Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 4 Mar 2003 11:41:24 -0000 Organization: Titanic Enterprises Unlimited Lines: 45 Message-ID: <1046778084.707238@saucer.planet.gong> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D4816.C974A441@yahoo.com> <6ud04b.lah.ln@escape.shannon.net> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1046778085 16741 80.177.7.220 (4 Mar 2003 11:41:25 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Tue, 4 Mar 2003 11:41:25 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Cache: nntpcache 3.0.1 (see http://www.nntpcache.org/) X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Cache-Post-Path: saucer.planet.gong!unknown@voodoo.planet.gong Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-fra1.dfn.de!news0.de.colt.net!peernews3.colt.net!colt.net!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131031 alt.sys.pdp10:12698 "Charles Shannon Hendrix" wrote in message news:6ud04b.lah.ln@escape.shannon.net... [SNIP] > My list wasn't complete of course. I left out the Crays, the Cyber 205 > that would be a must, and a few others. ETA-10P or ETA-10Q would be a nifty addition to that. Perhaps some Convex boxes too... There were a few going in the mid-90s. If I was a, hmm, it would have to be billionaire, I'd have a crack at creating something that does the job of the INMOS T9000. Up to 16 IEEE1355 style LVDS links, 64bit, a modest chunk of SRAM on board and a simple MMU. No floating point, but support for making it quick and easy to implement. Direct blinkenlight driving would be provided of course. The comms links would support remote booting, debugging, a block-random-access storage protocol and self-test functionality (even JTAG if possible). The whole thing would have ECC up to the wazoo. It would have something like Intel's speed-step to stop it from getting too toasty warm. In today's money I think I'd be aiming at .15u, 256Kwords, 500-800MHz. Nice and conservative to give a chance of decent yields. When that's all done & fabbed I'd have a crack at building a gi-normous box stuffed with these things with lots of blinkenlights too. Looking at around 4096 nodes (1Gw of on-board RAM total), with 1Tw worth of DRAM store backing it up. After doing all that I'd ring up weather centres and say "I've got considerably more SPECints than you". ;) Cheers, Rupert ###### Date: Tue, 4 Mar 2003 21:15:55 +0100 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <20030304211555.4731fa87.steveo@eircom.net> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <0kd04b.lsg.ln@escape.shannon.net> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 24 Organization: EuroNet Internet NNTP-Posting-Date: 04 Mar 2003 20:37:14 GMT NNTP-Posting-Host: i2023.vwr.wanadoo.nl X-Trace: 1046810234 maya.euronet.nl 23904 194.134.215.239:1080 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!newspeer.lavaseals.co.uk!diablo.theplanet.net!newsfeed.icl.net!newsfeed.fjserv.net!skynet.be!skynet.be!213.51.129.3.MISMATCH!newshub1.home.nl!home.nl!news2.euro.net!postnews1.euro.net!maya.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131058 alt.sys.pdp10:12702 On Tue, 04 Mar 03 09:24:57 GMT jmfbahciv@aol.com wrote: JC> In article <0kd04b.lsg.ln@escape.shannon.net>, JC> shannon@daydream.shannon.net (Charles Shannon Hendrix) wrote: JC> >In article , JC> >Peter da Silva wrote: JC> >> In article , JC> >> Charles Shannon Hendrix wrote: JC> >> >NetBSD is the last true UNIX... :) JC> >> JC> >> You misspelled "Plan 9". JC> > JC> >MMmmmm... how could I forget. JC> JC> Some of us only count in octal. So how come there was a PDP8 ? -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### Message-ID: <3E656055.6C436167@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org><0kd04b.lsg.ln@escape.shannon.net> <20030304211555.4731fa87.steveo@eircom.net> <994.193T61T8173778@kltpzyxm.invalid> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 27 Date: Wed, 05 Mar 2003 02:41:26 GMT NNTP-Posting-Host: 12.90.174.205 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc05-news.ops.worldnet.att.net 1046832086 12.90.174.205 (Wed, 05 Mar 2003 02:41:26 GMT) NNTP-Posting-Date: Wed, 05 Mar 2003 02:41:26 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!border3.nntp.aus1.giganews.com!nntp.giganews.com!wn12feed!worldnet.att.net!bgtnsc05-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131099 alt.sys.pdp10:12704 Charlie Gibbs wrote: > (Steve O'Hara-Smith) writes: > > jmfbahciv@aol.com wrote: > >> shannon@daydream.shannon.net (Charles Shannon Hendrix) wrote: > >>> Peter da Silva wrote: > >>>> Charles Shannon Hendrix wrote: > >>>> > >>>>> NetBSD is the last true UNIX... :) > >>>> > >>>> You misspelled "Plan 9". > >>> > >>> MMmmmm... how could I forget. > >> > >> Some of us only count in octal. > > > > So how come there was a PDP8 ? > > Dunno, but it thinks it's a PDP-10. Could this be the reason the C90 standard accepted 8 and 9 in octal fields :-) -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### From: Peter da Silva Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 05 Mar 2003 00:41:14 -0600 Organization: ABB Inc., Network Management Division Lines: 25 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <0kd04b.lsg.ln@escape.shannon.net> <20030304211555.4731fa87.steveo@eircom.net> NNTP-Posting-Host: switchblade.dhcp.abb.abbnm.com Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-Trace: jeeves.eng.abbnm.com 1046846588 17855 10.127.143.26 (5 Mar 2003 06:43:08 GMT) X-Complaints-To: usenet@jeeves.eng.abbnm.com NNTP-Posting-Date: Wed, 5 Mar 2003 06:43:08 +0000 (UTC) User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.1) Gecko/20020826 X-Accept-Language: en-us, en, ja Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!headwall.stanford.edu!newsfeed.stanford.edu!news-proxy.abbnm.com!newsboy.eng.abbnm.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131044 alt.sys.pdp10:12700 Steve O'Hara-Smith wrote: > On Tue, 04 Mar 03 09:24:57 GMT > jmfbahciv@aol.com wrote: > JC> In article <0kd04b.lsg.ln@escape.shannon.net>, > JC> shannon@daydream.shannon.net (Charles Shannon Hendrix) wrote: > JC> >In article , > JC> >Peter da Silva wrote: > JC> >> In article , > JC> >> Charles Shannon Hendrix wrote: > JC> >> >NetBSD is the last true UNIX... :) > JC> >> You misspelled "Plan 9". > JC> >MMmmmm... how could I forget. > JC> Some of us only count in octal. > So how come there was a PDP8 ? Well, you know, there are 10 kinds of people... -- `-_-' Ar rug tú barróg ar do mhactíre inniu? 'U` Peter da Silva ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E60AEEB.4FBB6FD2@yahoo.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 19 Date: Wed, 5 Mar 2003 16:30:45 -0500 NNTP-Posting-Host: 206.246.249.146 X-Trace: sydney.visi.net 1046903174 206.246.249.146 (Wed, 05 Mar 2003 17:26:14 EST) NNTP-Posting-Date: Wed, 05 Mar 2003 17:26:14 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131113 In article <3E60AEEB.4FBB6FD2@yahoo.com>, Peter Flass wrote: > Back in my application programming days we were programming a new > accounting system. My wife was taking evening courses in grad school. > I'd drive her to class and drag along a COBOL pad and the program > specs. By the end of an hour I'd have a program that usually only > needed another hour or so of test and cleanup to work correctly -- say > maybe 500-750 lines. OTOH I've done parsers that took weeks for the > same number of LOC. Of course the 10 LOC *is* an average. I still write code on paper. Makes you think. Not always the right thing, but works out great for the top-down design, and some bottom-up low level stuff. 10 LOC as an average still seems far too low. Perhaps if you are including the days in which you don't code at all, this makes more sense. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1281.186T405T7773222@kltpzyxm.invalid> <3E5D4816.C974A441@yahoo.com> <1fr473e.19tmher19fx3diN%lars@bearnip.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 17 Date: Wed, 5 Mar 2003 18:43:38 -0500 NNTP-Posting-Host: 209.96.185.148 X-Trace: sydney.visi.net 1046910384 209.96.185.148 (Wed, 05 Mar 2003 19:26:24 EST) NNTP-Posting-Date: Wed, 05 Mar 2003 19:26:24 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!nntp.abs.net!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131224 In article <1fr473e.19tmher19fx3diN%lars@bearnip.com>, Lars Duening wrote: > Charles Shannon Hendrix wrote: > >> This is how you know that people like Bill Gates are not true geeks, >> computer scientists, etc. If I were a billionaire, it wouldn't >> be wasted on a 12-meter yacht (maybe a tallship though) or a >> huge mansion. >> >> It would be an underground fallout shelter with Multics machines, >> PDP-11s, S/370s. > > Would an old missile silo complex suffice? That's one of the things in consideration for my bright future... :) ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <3E57B900.4030601@jetnet.ab.ca> <3E5B170F.90501@beagle-ears.com> <3E5ECCE1.7040709@bellatlantic.net> <1fr47ad.ujrbc01gjh24lN%lars@bearnip.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 20 Date: Wed, 5 Mar 2003 19:06:14 -0500 NNTP-Posting-Host: 209.96.185.148 X-Trace: sydney.visi.net 1046913970 209.96.185.148 (Wed, 05 Mar 2003 20:26:10 EST) NNTP-Posting-Date: Wed, 05 Mar 2003 20:26:10 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!news.gtei.net!bloom-beacon.mit.edu!newsswitch.lcs.mit.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131218 In article <1fr47ad.ujrbc01gjh24lN%lars@bearnip.com>, Lars Duening wrote: > Charles Shannon Hendrix wrote: > >> In article , jmfbahciv@aol.com wrote: >> >> >>Richie could always always find a way to bum code. >> >>I think that was what we called it. >> > >> > Yup. That's what we called it, too. >> >> I remember a few people in the 6502 era calling it "page sixing" your >> code, which was just a reference to cramming as much as you could into >> page 6 memory. > > But page 6 was part of the video memory?! No. Page six was 600-6FF, protected from OS usage. ###### Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Reply-To: Anne & Lynn Wheeler From: Anne & Lynn Wheeler Message-ID: Organization: Wheeler&Wheeler User-Agent: Gnus/5.090015 (Oort Gnus v0.15) Emacs/21.2 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Cancel-Lock: sha1:wmb6tUFNnq+wfbjrMR+MyvK2KdE= MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Lines: 25 Date: Thu, 06 Mar 2003 13:35:57 GMT NNTP-Posting-Host: 209.245.9.95 X-Complaints-To: abuse@earthlink.net X-Trace: newsread2.prod.itd.earthlink.net 1046957757 209.245.9.95 (Thu, 06 Mar 2003 05:35:57 PST) NNTP-Posting-Date: Thu, 06 Mar 2003 05:35:57 PST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.icl.net!newsfeed.fjserv.net!newsfeed.news2me.com!newsfeed2.earthlink.net!newsfeed1.earthlink.net!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!newsread2.prod.itd.earthlink.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131157 Pete Fenelon writes: > We're breeding a generation of applications programmers- and very > few systems/tools guys to actually provide the bits. (Which, as a > systems/tools sort of guy, reassures me, because it means those > sort of skills will be increasingly at a premium as a generation > raised on handholding environments teaches another generation.... > ;)) the other way of looking at it ... is that was critical path to deliverables. in the '90s companies were paying obscene amounts of money to people to get applications into the market as quickly as possible ... and frequently the quickest way to demonstrate a deliverable was with scripting and lots of re-use. it hardly matter if it couldn't scale ... somebody would just pay obscene amounts of money for hardware, and if it got hacked ... well, getting hacked was intrinsic characteristic of computers ... nobody could do anything about it. if one was so inclined, one might make an analogy to later periods in egypt ... it was easier to get building materials by scavenging off the pyramids. -- Anne & Lynn Wheeler | http://www.garlic.com/~lynn/ Internet trivia 20th anv http://www.garlic.com/~lynn/rfcietff.htm ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 06 Mar 03 09:36:51 GMT Organization: UltraNet Communications, Inc. Lines: 25 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <0kd04b.lsg.ln@escape.shannon.net> X-Trace: UmFuZG9tSVZCXyA+V8tXEWwSL3JNH8y5g/wUmNlOz1yjJ4kQ1E9Q/Igr1dPk1V/P X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 6 Mar 2003 10:21:39 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!newscore.univie.ac.at!feed.news.nacamar.de!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-81 Xref: chonsp.franklin.ch alt.folklore.computers:131132 In article , Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: >> In article <0kd04b.lsg.ln@escape.shannon.net>, >> shannon@daydream.shannon.net (Charles Shannon Hendrix) wrote: >>>In article , >>>Peter da Silva wrote: >>>> In article , >>>> Charles Shannon Hendrix wrote: >>>> >NetBSD is the last true UNIX... :) >>>> >>>> You misspelled "Plan 9". >>> >>>MMmmmm... how could I forget. >> >> Some of us only count in octal. > >Yes, but if you are enlightened with Hex, you discover there is so much >more afer 7... :) > Yea, bells but no whistles. ;-) /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 06 Mar 03 09:38:37 GMT Organization: UltraNet Communications, Inc. Lines: 29 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <0kd04b.lsg.ln@escape.shannon.net> <3E66E163.7B6442C2@ev1.net> <3E66E45E.BA908739@ev1.net> X-Trace: UmFuZG9tSVbdO2a7juaWYhGmd21eTq3vm7ndrJ8mtRbRz4JzYpv60fdjVI0ogWXb X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 6 Mar 2003 10:23:25 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!feed.news.nacamar.de!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-81 Xref: chonsp.franklin.ch alt.folklore.computers:131134 alt.sys.pdp10:12706 In article <3E66E45E.BA908739@ev1.net>, Charles Richmond wrote: >Charles Richmond wrote: >> >> jmfbahciv@aol.com wrote: >> > >> > In article <0kd04b.lsg.ln@escape.shannon.net>, >> > shannon@daydream.shannon.net (Charles Shannon Hendrix) wrote: >> > >In article , >> > >Peter da Silva wrote: >> > >> In article , >> > >> Charles Shannon Hendrix wrote: >> > >> >NetBSD is the last true UNIX... :) >> > >> >> > >> You misspelled "Plan 9". >> > > >> > >MMmmmm... how could I forget. >> > >> > Some of us only count in octal. >> > >> So that would be "Plan 010"... >> >Err...what I should have said was "Plan 010++"... Inflation? I didn't know there were pluses in octal. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 06 Mar 03 10:42:15 GMT Organization: UltraNet Communications, Inc. Lines: 31 Message-ID: References: <20030306065440.30f12471.steveo@eircom.net> X-Trace: UmFuZG9tSVam1Jt/Q0CEh0sE6yxNs+FplMHkiPrCwlsbCm11GSVyekB1RbQ8p4Lf X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 6 Mar 2003 11:27:02 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-81 Xref: chonsp.franklin.ch alt.folklore.computers:131140 In article <20030306065440.30f12471.steveo@eircom.net>, Steve O'Hara-Smith wrote: >On Wed, 5 Mar 2003 16:30:45 -0500 >Charles Shannon Hendrix wrote: > >CSH> 10 LOC as an average still seems far too low. >CSH> >CSH> Perhaps if you are including the days in which you don't code at all, >CSH> this makes more sense. > > Last time I heard this kind of figure being bandied around it >was indeed averaged over some long period and large number of people to >include all the things that consumed time but did not generate code. >The earliest incarnation I heard for an industry average was three lines >of fully debugged and documented COBOL per day. > > It's a crazy metric though - I've lost count of the number of >times I've made a system do more and better by reducing the total amount >of code in it. > Meetings, field test, non-programming jobs (this include support staff such as secretaries, pencil getters, waste basket emptiers, project management, product prevention). During field test we hoped lots of days when we didn't write 10 lines of code/day. The goal was 0 lines of code per day. Our field tests lasted a year. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 06 Mar 03 10:43:52 GMT Organization: UltraNet Communications, Inc. Lines: 11 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> X-Trace: UmFuZG9tSVYmOc4d3Yc3GdxMKASaz0gUnA0eM3r67nYYsKaTVkdI7IehIykevVbV X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 6 Mar 2003 11:28:40 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.stanford.edu!logbridge.uoregon.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-81 Xref: chonsp.franklin.ch alt.folklore.computers:131144 In article <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com>, Brian Inglis wrote: >640K ought to be enough for any bootstrap! ;^> And an OS. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 06 Mar 03 10:48:24 GMT Organization: UltraNet Communications, Inc. Lines: 48 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> X-Trace: UmFuZG9tSVaMXBv97Agw7YCfQbLH66zoVKvtFZm9QAwfD3PrF1UOOnOs6Vrfah9W X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 6 Mar 2003 11:33:12 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-81 Xref: chonsp.franklin.ch alt.folklore.computers:131155 In article , Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: > >> I watched my nephew play with his computer (about five years ago). >> He had a very strange way of thinking about the thingies we call >> files. I didn't have time and was too sick to clarify his >> concept. But it smelled a lot like the approach JMF told me >> about...huh...now I can't recall the program..we were going >> to buy us one and study how the software arranged its data. >[snip] I think the program was called Hypercard (but that still doesn't sound quite right). > >I've found it very hard to communicate with people at work before, >until I finally realize they have no idea what a file is, what a >disk is, etc. Isn't that positively scary? Our basic life support systems are based on that flavor of knowledge. > >> So, I don't think that kids brought up on Windows thinks of >> file and data organizations the way we do. Ours is based on >> how we "view" the lay of the bits and theirs is based on >> how they "view" the lay of the window. And there's not much in >> common with the two. > >Agreed. > >I find their knowledge incredibly shallow, even when they are somehow >able to crank out working programs. I don't know if I'd call it shallow...I'd have to learn more about how they think. > >I don't like systems which hide things. I like easy, but I don't >like dumb. We used to make dangerous things difficult, which isn't quite the same thing as hidden. /BAH Subtract a hundred and four for e-mail. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 43 Date: Wed, 5 Mar 2003 18:22:51 -0500 NNTP-Posting-Host: 209.96.185.148 X-Trace: sydney.visi.net 1046910371 209.96.185.148 (Wed, 05 Mar 2003 19:26:11 EST) NNTP-Posting-Date: Wed, 05 Mar 2003 19:26:11 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131222 In article , Ben Hutchings wrote: > In article , > Charles Shannon Hendrix wrote: > >> A PC BIOS can only boot an image up to 2.88MB, the size of the highest >> density 3.5" floppy. This is true even for a CD boot. The boot image >> on the CD is also limited to 2.88MB. > > > A PC BIOS only reads the first sector off the boot disk. It's up to > the code squeezed into those 512 bytes (446 in the case of a hard > disk, as 66 bytes are used for the partition table) to tell it to read > additional sectors. There is a limit on the extent of a disk > accessible through the BIOS routines, but it's much higher than > 2.88MB. I don't think so. It is true the BIOS only knows how to load the 512-byte boot code. This code is loaded to a specific location and is executed in real mode. Due to the small size, bare-metal programming to get out of real mode is nearly impossible. This code needs to use BIOS functions to jump to the next stage, and these functions have limitations in placement and size of the next stage. The exact limitations aren't even the same across all BIOS, which makes the problem even more "interesting". I was pretty sure this second stage could only be 2.88MB, and I know that's the case on a CD-ROM. I couldn't find a reference to the limit of this state anywhere. What I don't understand is why a modern PC BIOS doesn't have a "No DOS" flag for when you don't need the 640K split and other DOS/Windows hooks, and why they don't have a 32-bit native loader after all these years. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <0kd04b.lsg.ln@escape.shannon.net> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 18 Date: Wed, 5 Mar 2003 18:25:07 -0500 NNTP-Posting-Host: 209.96.185.148 X-Trace: sydney.visi.net 1046910373 209.96.185.148 (Wed, 05 Mar 2003 19:26:13 EST) NNTP-Posting-Date: Wed, 05 Mar 2003 19:26:13 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!rcn!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131227 alt.sys.pdp10:12712 In article , jmfbahciv@aol.com wrote: > In article <0kd04b.lsg.ln@escape.shannon.net>, > shannon@daydream.shannon.net (Charles Shannon Hendrix) wrote: >>In article , >>Peter da Silva wrote: >>> In article , >>> Charles Shannon Hendrix wrote: >>> >NetBSD is the last true UNIX... :) >>> >>> You misspelled "Plan 9". >> >>MMmmmm... how could I forget. > > Some of us only count in octal. Yes, but if you are enlightened with Hex, you discover there is so much more afer 7... :) ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10,alt.os.multics From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D4816.C974A441@yahoo.com> <6ud04b.lah.ln@escape.shannon.net> <1046778084.707238@saucer.planet.gong> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Followup-To: alt.folklore.computers Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 34 Date: Wed, 5 Mar 2003 18:47:10 -0500 NNTP-Posting-Host: 209.96.185.148 X-Trace: sydney.visi.net 1046910407 209.96.185.148 (Wed, 05 Mar 2003 19:26:47 EST) NNTP-Posting-Date: Wed, 05 Mar 2003 19:26:47 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131220 alt.sys.pdp10:12711 In article <1046778084.707238@saucer.planet.gong>, Rupert Pigott wrote: > ETA-10P or ETA-10Q would be a nifty addition to that. Yeah. I used to work in the building at NASA LaRC where those machines were, along with Star and others. > Perhaps some Convex boxes too... There were a few going > in the mid-90s. At the last NASA LaRC auction I went to, they had a Convex sitting there. Seems so sad for it to be sitting there, and it had bee roughed up a bit too. > If I was a, hmm, it would have to be billionaire, I'd > have a crack at creating something that does the job > of the INMOS T9000. Up to 16 IEEE1355 style LVDS links, > 64bit, a modest chunk of SRAM on board and a simple MMU. > No floating point, but support for making it quick and > easy to implement. Direct blinkenlight driving would > be provided of course. Speaking of blinkenlights, you'd have to have a Thinking Machines system in there. The one with the arrays of lights and the thousands of CPUs. > After doing all that I'd ring up weather centres > and say "I've got considerably more SPECints than > you". ;) Yep. Really, these people waste their money so foolishly... ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.3 (Linux) Lines: 28 Date: Wed, 5 Mar 2003 18:49:04 -0500 NNTP-Posting-Host: 209.96.185.148 X-Trace: sydney.visi.net 1046910413 209.96.185.148 (Wed, 05 Mar 2003 19:26:53 EST) NNTP-Posting-Date: Wed, 05 Mar 2003 19:26:53 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!rcn!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131217 In article , jmfbahciv@aol.com wrote: > I watched my nephew play with his computer (about five years ago). > He had a very strange way of thinking about the thingies we call > files. I didn't have time and was too sick to clarify his > concept. But it smelled a lot like the approach JMF told me > about...huh...now I can't recall the program..we were going > to buy us one and study how the software arranged its data. [snip] I've found it very hard to communicate with people at work before, until I finally realize they have no idea what a file is, what a disk is, etc. > So, I don't think that kids brought up on Windows thinks of > file and data organizations the way we do. Ours is based on > how we "view" the lay of the bits and theirs is based on > how they "view" the lay of the window. And there's not much in > common with the two. Agreed. I find their knowledge incredibly shallow, even when they are somehow able to crank out working programs. I don't like systems which hide things. I like easy, but I don't like dumb. ###### From: bfranchuk@jetnet.ab.ca Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 05 Mar 2003 18:14:57 -0700 Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3E66A111.9060306@jetnet.ab.ca> User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.0.1) Gecko/20021005 X-Accept-Language: en-us, en MIME-Version: 1.0 References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 15 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!sn-xit-03!sn-xit-01!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131124 Charles Shannon Hendrix wrote: > What I don't understand is why a modern PC BIOS doesn't have a "No DOS" > flag for when you don't need the 640K split and other DOS/Windows hooks, > and why they don't have a 32-bit native loader after all these years. Because they still have the stupid XT memory map! Also modern I/O is very software bloated to set up. My gripe is that the bios never set the timer chip to something useful like 50 or 60 HZ. But wait the XT was to slow for that. Ben. ###### From: Ben Hutchings Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 6 Mar 2003 01:27:58 +0000 Lines: 87 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Reply-To: Ben Hutchings NNTP-Posting-Host: pc5-cmbg1-6-cust149.cmbg.cable.ntl.com (80.4.127.149) X-Trace: fu-berlin.de 1046914206 64054546 80.4.127.149 (16 [70929]) X-Orig-Path: 192.168.1.1!nobody User-Agent: slrn/0.9.7.4 (Linux) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.tpinternet.pl!newsfeed01.sul.t-online.de!t-online.de!fu-berlin.de!uni-berlin.de!pc5-cmbg1-6-cust149.cmbg.cable.ntl.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131221 In article , Charles Shannon Hendrix wrote: > In article , > Ben Hutchings wrote: >> In article , >> Charles Shannon Hendrix wrote: >> >>> A PC BIOS can only boot an image up to 2.88MB, the size of the highest >>> density 3.5" floppy. This is true even for a CD boot. The boot image >>> on the CD is also limited to 2.88MB. >> >> >> A PC BIOS only reads the first sector off the boot disk. It's up to >> the code squeezed into those 512 bytes (446 in the case of a hard >> disk, as 66 bytes are used for the partition table) to tell it to read >> additional sectors. There is a limit on the extent of a disk >> accessible through the BIOS routines, but it's much higher than >> 2.88MB. > > I don't think so. > > It is true the BIOS only knows how to load the 512-byte boot code. > > This code is loaded to a specific location and is executed in real mode. > Due to the small size, bare-metal programming to get out of real mode > is nearly impossible. > > This code needs to use BIOS functions to jump to the next stage, and > these functions have limitations in placement and size of the next > stage. > > The exact limitations aren't even the same across all BIOS, which makes > the problem even more "interesting". There are two different limits on drive geometry: the original BIOS disk API is limited to 1024 cylinders, 255 heads (!) and 63 sectors; the original AT HD controller (which ATA/IDE disks remain compatible with) is limited to 16383 cylinders, 16 heads and 63 sectors. Combine these limitations and you get just under 1024 * 16 * 63 sectors ~= 500 MB. Some versions of DOS and Windows have a limit of 2 GB. Later BIOSes invent a new fictitious geometry for the disk that suits the disk API, and translate addresses. This raises the limit to 1024 * 255 * 63 sectors ~= 8 GB. IDE disks now support 28-bit sector addresses, with none of the pretence about the host knowing the disk geometry. This is called LBA (logical block addressing). There is also an extension to the BIOS disk API that allows the use of 32-bit sector addresses. These result in limits of 128 GB and 2 TB respectively. Unfortunately IDE disks are already pushing the lower of these. > I was pretty sure this second stage could only be 2.88MB, and I know > that's the case on a CD-ROM. The BIOS disk API was designed for floppy and hard disks only. There's no place in the configuration table in low memory for the BIOS to describe a CD-ROM drive. When you boot a PC from CD-ROM, the BIOS maps a file on the CD-ROM as either FD 0 (A: in DOS) or HD 0 (C: in DOS). Bootable CDs are mostly used for installing operating systems onto a hard drive, so hiding hard HD 0 is not a good idea - it doesn't prevent access the HD using a proper driver but it does make it impossible to tell what geometry the BIOS presents for the HD, which must be taken into account when installing the boot loader. Therefore it is normal to use a floppy drive image instead. The largest standard size of floppy drive is 2.88 MB. > I couldn't find a reference to the limit of this state anywhere. > > What I don't understand is why a modern PC BIOS doesn't have a "No DOS" > flag for when you don't need the 640K split and other DOS/Windows hooks, > and why they don't have a 32-bit native loader after all these years. Intel came up with an Enhanced Firmware Interface (EFI) for use in PCs with their 64-bit processors and are encouraging its use on x86 PCs now as well. Alternatively you can put Linux in Flash and use that as your boot loader: . -- Ben Hutchings | personal web site: http://womble.decadentplace.org.uk/ Any smoothly functioning technology is indistinguishable from a rigged demo. ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 6 Mar 2003 02:54:57 -0000 Organization: Titanic Enterprises Unlimited Lines: 22 Message-ID: <1046919297.984334@saucer.planet.gong> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D4816.C974A441@yahoo.com> <6ud04b.lah.ln@escape.shannon.net> <1046778084.707238@saucer.planet.gong> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1046919298 8873 80.177.7.220 (6 Mar 2003 02:54:58 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Thu, 6 Mar 2003 02:54:58 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Cache: nntpcache 3.0.1 (see http://www.nntpcache.org/) X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Cache-Post-Path: saucer.planet.gong!unknown@voodoo.planet.gong Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!nntp.theplanet.net!inewsm1.nntp.theplanet.net!195.40.4.120.MISMATCH!easynet-quince!easynet.net!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131174 "Charles Shannon Hendrix" wrote in message news:slrnb6d33u.n0s.shannon@news.widomaker.com... [SNIP] > Speaking of blinkenlights, you'd have to have a Thinking Machines system > in there. The one with the arrays of lights and the thousands of CPUs. Hmm, probably mis-remembering the model designations but : CM-2, CM-200, TM-5 ? The only positive thing I've ever heard about the TM-5 was it's impressive blinkenlights. Everything else I've heard has been along the lines of how much it sucked. I looked at the specs of it (network layout etc) way back when and drew the same conclusion myself. Anyone actually got direct experience of the '5 ? Cheers, Rupert ###### Message-ID: <3E66E45E.BA908739@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <0kd04b.lsg.ln@escape.shannon.net> <3E66E163.7B6442C2@ev1.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 27 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc51.ops.asp.att.net 1046923430 12.241.15.59 (Thu, 06 Mar 2003 04:03:50 GMT) NNTP-Posting-Date: Thu, 06 Mar 2003 04:03:50 GMT Date: Thu, 06 Mar 2003 04:03:50 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!nntp-out.monmouth.com!newspeer.monmouth.com!news.maxwell.syr.edu!wn14feed!wn13feed!wn12feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!rwcrnsc51.ops.asp.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131209 alt.sys.pdp10:12709 Charles Richmond wrote: > > jmfbahciv@aol.com wrote: > > > > In article <0kd04b.lsg.ln@escape.shannon.net>, > > shannon@daydream.shannon.net (Charles Shannon Hendrix) wrote: > > >In article , > > >Peter da Silva wrote: > > >> In article , > > >> Charles Shannon Hendrix wrote: > > >> >NetBSD is the last true UNIX... :) > > >> > > >> You misspelled "Plan 9". > > > > > >MMmmmm... how could I forget. > > > > Some of us only count in octal. > > > So that would be "Plan 010"... > Err...what I should have said was "Plan 010++"... -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### Message-ID: <3E66E5A7.17F82A8A@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org><0kd04b.lsg.ln@escape.shannon.net> <20030304211555.4731fa87.steveo@eircom.net> <994.193T61T8173778@kltpzyxm.invalid> <3E656055.6C436167@yahoo.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 32 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc54 1046923759 12.241.15.59 (Thu, 06 Mar 2003 04:09:19 GMT) NNTP-Posting-Date: Thu, 06 Mar 2003 04:09:19 GMT Date: Thu, 06 Mar 2003 04:09:19 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!solnet.ch!solnet.ch!newsfeed.freenet.de!213.253.16.105.MISMATCH!mephistopheles.news.clara.net!news.clara.net!nntp.theplanet.net!inewsm1.nntp.theplanet.net!195.40.4.120.MISMATCH!easynet-quince!easynet.net!cox.net!cyclone1.gnilink.net!wn14feed!wn13feed!wn12feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!rwcrnsc54.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131212 alt.sys.pdp10:12710 CBFalconer wrote: > > Charlie Gibbs wrote: > > (Steve O'Hara-Smith) writes: > > > jmfbahciv@aol.com wrote: > > >> shannon@daydream.shannon.net (Charles Shannon Hendrix) wrote: > > >>> Peter da Silva wrote: > > >>>> Charles Shannon Hendrix wrote: > > >>>> > > >>>>> NetBSD is the last true UNIX... :) > > >>>> > > >>>> You misspelled "Plan 9". > > >>> > > >>> MMmmmm... how could I forget. > > >> > > >> Some of us only count in octal. > > > > > > So how come there was a PDP8 ? > > > > Dunno, but it thinks it's a PDP-10. > > Could this be the reason the C90 standard accepted 8 and 9 in > octal fields :-) > The older K&R C would accept 8 and 9 in octal numbers...but I do *not* believe that the C89 standard accepted 8 and 9 as octal digits. -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### Date: Thu, 6 Mar 2003 06:42:52 +0100 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <20030306064252.64caca77.steveo@eircom.net> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 17 Organization: EuroNet Internet NNTP-Posting-Date: 06 Mar 2003 06:40:30 GMT NNTP-Posting-Host: 62.234.211.234 X-Trace: 1046932830 maya.euronet.nl 23897 62.234.211.234:1188 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-lei1.dfn.de!newsfeed00.sul.t-online.de!newsfeed01.sul.t-online.de!t-online.de!newspeer1-gui.server.ntli.net!ntli.net!news2.euro.net!postnews1.euro.net!maya.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131188 On Thu, 6 Mar 2003 01:27:58 +0000 Ben Hutchings wrote: BH> in limits of 128 GB and 2 TB respectively. Unfortunately IDE disks BH> are already pushing the lower of these. Pushing ? They went through it some time ago, my favorite PC bits sypplier doesn't list anything smaller than 60GB. There are three at 200GB and one at 180GB (all ATA). The biggest drive (I am amazed to discover) though is apparently a 500GB LaCie firewire drive. I'd hate to back that up on 9-track :) -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### Date: Thu, 6 Mar 2003 06:46:39 +0100 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <20030306064639.1fe42513.steveo@eircom.net> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <0kd04b.lsg.ln@escape.shannon.net> <3E66E163.7B6442C2@ev1.net> <3E66E45E.BA908739@ev1.net> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 17 Organization: EuroNet Internet NNTP-Posting-Date: 06 Mar 2003 06:40:31 GMT NNTP-Posting-Host: 62.234.211.234 X-Trace: 1046932831 maya.euronet.nl 23897 62.234.211.234:1188 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!diablo.theplanet.net!newspeer1-gui.server.ntli.net!ntli.net!news2.euro.net!postnews1.euro.net!maya.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131183 alt.sys.pdp10:12707 On Thu, 06 Mar 2003 04:03:50 GMT Charles Richmond wrote: CR> Charles Richmond wrote: CR> > CR> > So that would be "Plan 010"... CR> > CR> Err...what I should have said was "Plan 010++"... Oh no! The Stroustrop (sp ?) mistake again - make it ++010 and return the incremented version for use :) -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### Date: Thu, 6 Mar 2003 06:54:40 +0100 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <20030306065440.30f12471.steveo@eircom.net> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E60AEEB.4FBB6FD2@yahoo.com> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 23 Organization: EuroNet Internet NNTP-Posting-Date: 06 Mar 2003 06:40:32 GMT NNTP-Posting-Host: 62.234.211.234 X-Trace: 1046932832 maya.euronet.nl 23897 62.234.211.234:1188 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-lei1.dfn.de!newsfeed00.sul.t-online.de!newsfeed01.sul.t-online.de!t-online.de!newspeer1-gui.server.ntli.net!ntli.net!news2.euro.net!postnews1.euro.net!maya.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131180 On Wed, 5 Mar 2003 16:30:45 -0500 Charles Shannon Hendrix wrote: CSH> 10 LOC as an average still seems far too low. CSH> CSH> Perhaps if you are including the days in which you don't code at all, CSH> this makes more sense. Last time I heard this kind of figure being bandied around it was indeed averaged over some long period and large number of people to include all the things that consumed time but did not generate code. The earliest incarnation I heard for an industry average was three lines of fully debugged and documented COBOL per day. It's a crazy metric though - I've lost count of the number of times I've made a system do more and better by reducing the total amount of code in it. -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 53 Date: Thu, 06 Mar 2003 06:49:47 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news2.calgary.shaw.ca 1046933387 24.71.223.147 (Wed, 05 Mar 2003 23:49:47 MST) NNTP-Posting-Date: Wed, 05 Mar 2003 23:49:47 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!c03.atl99!sjc70.webusenet.com!news.webusenet.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news2.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131203 On Wed, 5 Mar 2003 18:22:51 -0500 in alt.folklore.computers, Charles Shannon Hendrix wrote: >In article , Ben Hutchings wrote: >> In article , >> Charles Shannon Hendrix wrote: >> >>> A PC BIOS can only boot an image up to 2.88MB, the size of the highest >>> density 3.5" floppy. This is true even for a CD boot. The boot image >>> on the CD is also limited to 2.88MB. >> >> >> A PC BIOS only reads the first sector off the boot disk. It's up to >> the code squeezed into those 512 bytes (446 in the case of a hard >> disk, as 66 bytes are used for the partition table) to tell it to read >> additional sectors. There is a limit on the extent of a disk >> accessible through the BIOS routines, but it's much higher than >> 2.88MB. > >I don't think so. > >It is true the BIOS only knows how to load the 512-byte boot code. > >This code is loaded to a specific location and is executed in real mode. >Due to the small size, bare-metal programming to get out of real mode >is nearly impossible. > >This code needs to use BIOS functions to jump to the next stage, and >these functions have limitations in placement and size of the next >stage. > >The exact limitations aren't even the same across all BIOS, which makes >the problem even more "interesting". > >I was pretty sure this second stage could only be 2.88MB, and I know >that's the case on a CD-ROM. > >I couldn't find a reference to the limit of this state anywhere. > >What I don't understand is why a modern PC BIOS doesn't have a "No DOS" >flag for when you don't need the 640K split and other DOS/Windows hooks, >and why they don't have a 32-bit native loader after all these years. 640K ought to be enough for any bootstrap! ;^> Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### Message-ID: <3E670FD7.1483B3AC@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org><0kd04b.lsg.ln@escape.shannon.net> <20030304211555.4731fa87.steveo@eircom.net> <994.193T61T8173778@kltpzyxm.invalid> <3E656055.6C436167@yahoo.com> <3E66E5A7.17F82A8A@ev1.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 37 Date: Thu, 06 Mar 2003 09:42:05 GMT NNTP-Posting-Host: 12.90.167.248 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc05-news.ops.worldnet.att.net 1046943725 12.90.167.248 (Thu, 06 Mar 2003 09:42:05 GMT) NNTP-Posting-Date: Thu, 06 Mar 2003 09:42:05 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!wn14feed!worldnet.att.net!bgtnsc05-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131117 alt.sys.pdp10:12705 Charles Richmond wrote: > CBFalconer wrote: > > Charlie Gibbs wrote: > > > (Steve O'Hara-Smith) writes: > > > > jmfbahciv@aol.com wrote: > > > >> shannon@daydream.shannon.net (Charles Shannon Hendrix) wrote: > > > >>> Peter da Silva wrote: > > > >>>> Charles Shannon Hendrix wrote: > > > >>>> > > > >>>>> NetBSD is the last true UNIX... :) > > > >>>> > > > >>>> You misspelled "Plan 9". > > > >>> > > > >>> MMmmmm... how could I forget. > > > >> > > > >> Some of us only count in octal. > > > > > > > > So how come there was a PDP8 ? > > > > > > Dunno, but it thinks it's a PDP-10. > > > > Could this be the reason the C90 standard accepted 8 and 9 in > > octal fields :-) > > > The older K&R C would accept 8 and 9 in octal numbers...but > I do *not* believe that the C89 standard accepted 8 and 9 as > octal digits. Woops - you are right. I just checked it. I have been spreading urban legends. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### From: Pete Fenelon Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 06 Mar 2003 12:40:47 -0000 Organization: Posted via Supernews, http://www.supernews.com Message-ID: Sender: Pete Fenelon References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> User-Agent: tin/1.5.16-20021229 ("Spiders") (UNIX) (FreeBSD/4.7-STABLE (i386)) X-Complaints-To: abuse@supernews.com Lines: 43 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!news.tele.dk!news.tele.dk!small.news.tele.dk!newsfeed.icl.net!newsfeed.fjserv.net!news.maxwell.syr.edu!sn-xit-03!sn-xit-06!sn-xit-01!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131189 jmfbahciv@aol.com wrote: >>> to buy us one and study how the software arranged its data. >>[snip] > > I think the program was called Hypercard (but that still doesn't > sound quite right). > There was a Mac scripting/development environment called Hypercard a while back - you could well be right. >>I find their knowledge incredibly shallow, even when they are somehow >>able to crank out working programs. > > I don't know if I'd call it shallow...I'd have to learn more about > how they think. > I've found that interviewing recent graduates reveals that their knowledge of what goes on inside complex systems (OS, device driver, compiler, debugger, filesystem) is often weak to nonexistent - but they're a lot happier scripting, reusing off the shelf components, and doing user-interface work than "oldtimers". Then again, they're the generation who've grown up with the ubiquitous computer - most of my contemporaries (I'm in my mid 30s) got into computing in the late 70s/early 80s and we've all done our share of hacking on baby micros before starting on "proper computers" - the majority of the guys in their early 20s grew up with point'n'drool environments, slick applications and games - almost no hacking required, and no real incentive to look under the hood. We're breeding a generation of applications programmers- and very few systems/tools guys to actually provide the bits. (Which, as a systems/tools sort of guy, reassures me, because it means those sort of skills will be increasingly at a premium as a generation raised on handholding environments teaches another generation.... ;)) pete -- pete@fenelon.com "there's no room for enigmas in built-up areas" HMHB ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 6 Mar 2003 14:32:17 GMT Organization: TSS Inc. Lines: 13 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org><994.193T61T8173778@kltpzyxm.invalid> <3E656055.6C436167@yahoo.com> <3E66E5A7.17F82A8A@ev1.net> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1046961137 37194 10.0.0.43 (6 Mar 2003 14:32:17 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 6 Mar 2003 14:32:17 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!feed.news.nacamar.de!newsfeed.icl.net!newsfeed.fjserv.net!newsfeed.icl.net!newsfeed.fjserv.net!proxad.net!proxad.net!news.kjsl.com!news.usenet2.org!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131186 alt.sys.pdp10:12708 In article <3E66E5A7.17F82A8A@ev1.net>, Charles Richmond wrote: >The older K&R C would accept 8 and 9 in octal numbers...but >I do *not* believe that the C89 standard accepted 8 and 9 as >octal digits. So should it be C0131 or C0111? -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: Ben Hutchings Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 6 Mar 2003 21:18:55 +0000 Lines: 21 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Reply-To: Ben Hutchings NNTP-Posting-Host: pc5-cmbg1-6-cust149.cmbg.cable.ntl.com (80.4.127.149) X-Trace: fu-berlin.de 1046986218 63703008 80.4.127.149 (16 [70929]) X-Orig-Path: 192.168.1.1!nobody User-Agent: slrn/0.9.7.4 (Linux) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!pc5-cmbg1-6-cust149.cmbg.cable.ntl.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131232 In article , I wrote: > Some versions of DOS and Windows have a limit of 2 GB. I remembered what this limit is. It's the limit on file-system size in FAT16 - the maximum block size is 32K and block numbers are 16-bit. This limit applies to the original Windows 95, NT 4 and all earlier MS operating systems using FAT. > Later BIOSes invent a new fictitious geometry for the disk that suits > the disk API, and translate addresses. This raises the limit to > 1024 * 255 * 63 sectors ~= 8 GB. This is assuming they get translation right. Apparently some buggy BIOSes have a limit somewhere between 500 MB and 8 GB. -- Ben Hutchings | personal web site: http://womble.decadentplace.org.uk/ 73.46% of all statistics are made up. ###### From: Ben Hutchings Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 6 Mar 2003 21:31:30 +0000 Lines: 15 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <20030306064252.64caca77.steveo@eircom.net> Reply-To: Ben Hutchings NNTP-Posting-Host: pc5-cmbg1-6-cust149.cmbg.cable.ntl.com (80.4.127.149) X-Trace: fu-berlin.de 1046987107 63551874 80.4.127.149 (16 [70929]) X-Orig-Path: 192.168.1.1!nobody User-Agent: slrn/0.9.7.4 (Linux) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsmi-eu.news.garr.it!NewsITBone-GARR!fu-berlin.de!uni-berlin.de!pc5-cmbg1-6-cust149.cmbg.cable.ntl.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131231 In article <20030306064252.64caca77.steveo@eircom.net>, Steve O'Hara-Smith wrote: > On Thu, 6 Mar 2003 01:27:58 +0000 > Ben Hutchings wrote: > > BH> in limits of 128 GB and 2 TB respectively. Unfortunately IDE disks > BH> are already pushing the lower of these. > > Pushing ? They went through it some time ago, my favorite PC > bits sypplier doesn't list anything smaller than 60GB. There are three > at 200GB and one at 180GB (all ATA). Yes, you're right. ATA-6 extended LBA up to something like 48 bits, so we can worry about the BIOS limit instead now. ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Thu, 6 Mar 2003 19:54:08 -0700 Lines: 24 Message-ID: <1frentm.c01c851nzju4gN%lars@bearnip.com> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <3E57B900.4030601@jetnet.ab.ca> <3E5B170F.90501@beagle-ears.com> <3E5ECCE1.7040709@bellatlantic.net> <1fr47ad.ujrbc01gjh24lN%lars@bearnip.com> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1047005650 64432400 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131409 Charles Shannon Hendrix wrote: > In article <1fr47ad.ujrbc01gjh24lN%lars@bearnip.com>, Lars Duening wrote: > > Charles Shannon Hendrix wrote: > > > >> In article , jmfbahciv@aol.com wrote: > >> > >> >>Richie could always always find a way to bum code. > >> >>I think that was what we called it. > >> > > >> > Yup. That's what we called it, too. > >> > >> I remember a few people in the 6502 era calling it "page sixing" your > >> code, which was just a reference to cramming as much as you could into > >> page 6 memory. > > > > But page 6 was part of the video memory?! > > No. > > Page six was 600-6FF, protected from OS usage. Not on my other machine - there it was non-existent altogether. But then again that machine also sported a 6802, it doesn't really count. ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Fri, 7 Mar 2003 08:48:47 -0700 Lines: 10 Message-ID: <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> References: <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1047052127 65275155 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131414 Charles Shannon Hendrix wrote: > There is nothing wrong with scripting and re-use. In fact, I saw very > little re-use in the majority of "do it yesterday" 90s projects. > > Just for example, C++ and Java often resulted in less re-use and much > more complicated software than the languages and projects they replaced. But that is surely not a problem of the languages, is it? > or how to fix them if they fail. All of the engineers died long ago. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC References: <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 52 Date: Fri, 7 Mar 2003 15:30:03 -0500 NNTP-Posting-Host: 209.96.189.100 X-Trace: sydney.visi.net 1047072365 209.96.189.100 (Fri, 07 Mar 2003 16:26:05 EST) NNTP-Posting-Date: Fri, 07 Mar 2003 16:26:05 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!news.webusenet.com!newsfeed-east.nntpserver.com!nntpserver.com!news-out.visi.com!petbe.visi.com!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131469 In article <1frfh0t.nm8oe5dijljmN%lars@bearnip.com>, Lars Duening wrote: > Charles Shannon Hendrix wrote: > >> There is nothing wrong with scripting and re-use. In fact, I saw very >> little re-use in the majority of "do it yesterday" 90s projects. >> >> Just for example, C++ and Java often resulted in less re-use and much >> more complicated software than the languages and projects they replaced. > > But that is surely not a problem of the languages, is it? I believe that it is partially the problem of the languages, yes. C++ has a lot of good points, but as an OO language, I think its contrived and confusing. Objective C is much cleaner to me. I don't like what happens with C++ libraries. It seems that even small programs end up large because they want to use a class or method which implies a lot of baggage. C++ linkers seem to be really inefficient as well, and the larger the program, the worse it gets. I suppose this could be a compiler problem, but it seems they are all this way. Of course, I've only used DEC, Sun, and GNU compilers. Java... it sucks on every platform except Sun, Windows, and Apple. Write once, run anywhere is mythology. The language itself looks interesting, and I know some people that have used a real Java compiler without the libraries on embedded stuff, and enjoyed it. But I think Jamie Zawinkski's rant on Java is pretty accurate, and he LIKES it: http://www.jwz.org/doc/java.html I don't like the library myself--too big and slow--but I think he is saying here what a lot of other people think. I thought a major point of OO code was to let you shrink the programs and simplify them, but it seems to me that C++ and Java projects balloon instead. So... no, don't put all the blame on the languages, but I'm not sure you can excuse them completely. I resent that some of my C projects were replace with Java code, which was bigger, slower, harder to maintain, and not as portable. Maybe that drives my opinion a little too much. At the same time, I think the idea that all projects are expressed best in OO fashion is ridiculous. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC References: <3E05FE9D.B3A12286@vax6k.openecs.org> <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <3E57B900.4030601@jetnet.ab.ca> <3E5B170F.90501@beagle-ears.com> <3E5ECCE1.7040709@bellatlantic.net> <1fr47ad.ujrbc01gjh24lN%lars@bearnip.com> <1frentm.c01c851nzju4gN%lars@bearnip.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 25 Date: Fri, 7 Mar 2003 15:33:16 -0500 NNTP-Posting-Host: 209.96.189.100 X-Trace: sydney.visi.net 1047072368 209.96.189.100 (Fri, 07 Mar 2003 16:26:08 EST) NNTP-Posting-Date: Fri, 07 Mar 2003 16:26:08 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!rcn!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131472 In article <1frentm.c01c851nzju4gN%lars@bearnip.com>, Lars Duening wrote: >> > But page 6 was part of the video memory?! >> >> No. >> >> Page six was 600-6FF, protected from OS usage. > > Not on my other machine - there it was non-existent altogether. But then > again that machine also sported a 6802, it doesn't really count. Sorry, I thought I had said I was talking about an Atari, but I didn't. Page-sixing code--evidently I'm the only one that knows this slang--was mostly from trying to cram multitasking stuff into page six. You could do DLIs, background tasks, all sorts of things in there. TextPro, a very nice 8-bit word processor, used this area for its background tasks. Aside: I was amazed to see what happened with the 8-bit Atari in the 90s. I saw some German graphics demos that I thought were fakes done on a new machine until I ran them myself. Then I found a couple of multitasking OSs that used extended memory, and a TCP/IP stack. ###### Message-ID: <3E691AC7.78612561@yahoo.com> From: Peter Flass X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 18 Date: Fri, 07 Mar 2003 22:26:53 GMT NNTP-Posting-Host: 24.194.50.82 X-Complaints-To: abuse@rr.com X-Trace: twister.nyroc.rr.com 1047076013 24.194.50.82 (Fri, 07 Mar 2003 17:26:53 EST) NNTP-Posting-Date: Fri, 07 Mar 2003 17:26:53 EST Organization: Road Runner Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!feedme.news.mediaways.net!news-fra1.dfn.de!news0.de.colt.net!peernews3.colt.net!colt.net!news.maxwell.syr.edu!newsfeed-east.nntpserver.com!nntpserver.com!news-west.rr.com!cyclone.nyroc.rr.com!cyclone-out.nyroc.rr.com!twister.nyroc.rr.com.POSTED!53ab2750!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131489 Lars Duening wrote: > > Charles Shannon Hendrix wrote: > > > There is nothing wrong with scripting and re-use. In fact, I saw very > > little re-use in the majority of "do it yesterday" 90s projects. > > > > Just for example, C++ and Java often resulted in less re-use and much > > more complicated software than the languages and projects they replaced. > > But that is surely not a problem of the languages, is it? > > or how to fix them if they fail. All of the engineers died long ago. No, just poor design. I have an interesting book "Programming the IBM 709", and they talk about including (reusable) library code by (ready for this?) appending the cards for the library routine to your program deck, and lots of that stuff was written in machine language, forget assembler. ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Sat, 8 Mar 2003 05:31:21 -0700 Lines: 25 Message-ID: <1frgw7f.xb8cjp8vo1s2N%lars@bearnip.com> References: <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <3E68E636.E638D524@ev1.net> <1539.196T2652T7445477@kltpzyxm.invalid> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1047126682 66006690 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed00.sul.t-online.de!newsfeed01.sul.t-online.de!t-online.de!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131406 Charlie Gibbs wrote: > In article > stanb45@dial.pipex.com (Stan Barr) writes: > > > On Fri, 07 Mar 2003 16:36:17 GMT, Charles Richmond > > wrote: > > > >> So you will like Unix/Linux/FreeBSD. They have commands that allow > >> you to set up your jobs to run "batch". > > > > And many Mac applications are script-aware, allowing them to be run > >from Applescript programs. > > But... but... all _modern_ systems are INTERACTIVE! Batch programming > is a throwback to the dinosaurs! Nobody who wants to be With the > Program [tm] would think of using that archaic stuff! You know, Apple has this IDE called 'Project Builder - pointy-clicky development to your heart's content. It also has a plain program 'pbxbuild' to build your PB projects from the shell. Heresy, no doubt about it. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 07 Mar 03 11:03:48 GMT Organization: UltraNet Communications, Inc. Lines: 33 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> X-Trace: UmFuZG9tSVZUQc8YAnR1ON0XXiJC4hZWMGi1Mlevf0Wdb4rsOjI58uGKcYsjXt4V X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 7 Mar 2003 11:48:43 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-33 Xref: chonsp.franklin.ch alt.folklore.computers:131338 In article , Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: >> In article <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com>, >> Brian Inglis wrote: >> >> >>>640K ought to be enough for any bootstrap! ;^> >> >> And an OS. > >Only if you want a limited one. > >Some filesystems are almost that big, like SGI's xfs. You don't need the file system in core to have one. > >Not all code is big because it is poorly written: some of it just does >more work. > >640K isn't even enough to hold the status information and buffers for >the processes on most heavily loaded servers. I have yet to meet a CPU that can deal with more than one thing at a time (that's why we have multiple-CPU systems). The reason all of that info has to be in core is because disk access is so slow. If you speed that up you don't need all of that data (that is there for JICs--just-in-case) in core. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 07 Mar 03 11:11:05 GMT Organization: UltraNet Communications, Inc. Lines: 48 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> X-Trace: UmFuZG9tSVYgUFba+2NBbiM2j6aWp0/CAc90nzjhN+e7v3yzbVa9ZGpUKfuIQizY X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 7 Mar 2003 11:56:00 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-33 Xref: chonsp.franklin.ch alt.folklore.computers:131327 In article , Charles Shannon Hendrix wrote: >I was using a downloading program on Windows the other day, and almost >pounded a keyboard into dust. The damned program put up a dialog box >FOR EACH @#$%@# DOWNLOAD START! RIGHT!!! The fucking point of having computers was so you could start something up then go away and sleep or something. Having to stay and babysit an automatic process is just a side effect of that UDFH. > >I couldn't do ANYTHING else while I waited for the downloads. It did >not help that the files were coming from one of those stupid web sites >which do not give you a real URL. At least you didn't get hijacked in the middle of them. > >I marvel at the plethora of dialog boxes on modern software. Where do >people get this idea that a program must annoy the absolute shit out of >the user to function or make it "easy"? > >Recently seen: > > +----------------+ > | | > | No errors! | > | | > | +------+ | > | | OK | | > | +------+ | > +----------------+ > >About 99% of program->user communication can be done in the UI without >any kind of dialog, and that includes errors. Whoa!!! Did you just push a hot button. Why do I get the option of saying OK to no errors but, if there is an error, the information that I need gets flashered at me then the screen is drawn blank. /BAH Subtract a hundred and four for e-mail. ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 07 Mar 03 12:24:53 -0800 Organization: http://extra.newsguy.com Lines: 40 Message-ID: <1539.196T2652T7445477@kltpzyxm.invalid> References: <3E68E636.E638D524@ev1.net> NNTP-Posting-Host: p-181.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.he.net!cyclone-sf.pbi.net!129.250.175.17!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews3 Xref: chonsp.franklin.ch alt.folklore.computers:131384 In article stanb45@dial.pipex.com (Stan Barr) writes: > On Fri, 07 Mar 2003 16:36:17 GMT, Charles Richmond > wrote: > >> jmfbahciv@aol.com wrote: >> >>> In article , >>> Charles Shannon Hendrix wrote: >>> >>> >>>> I was using a downloading program on Windows the other day, and >>>> almost pounded a keyboard into dust. The damned program put up >>>> a dialog box FOR EACH @#$%@# DOWNLOAD START! >>> >>> RIGHT!!! The fucking point of having computers was so you could >>> start something up then go away and sleep or something. Having to >>> stay and babysit an automatic process is just a side effect of that >>> UDFH. >> >> So you will like Unix/Linux/FreeBSD. They have commands that allow >> you to set up your jobs to run "batch". > > And many Mac applications are script-aware, allowing them to be run >from Applescript programs. But... but... all _modern_ systems are INTERACTIVE! Batch programming is a throwback to the dinosaurs! Nobody who wants to be With the Program [tm] would think of using that archaic stuff! These silly people. They don't realize that all too often, "interactive" is synonymous with "manual". -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 07 Mar 03 13:03:49 -0800 Organization: http://extra.newsguy.com Lines: 54 Message-ID: <2231.196T2462T7836007@kltpzyxm.invalid> References: <3E05FE9D.B3A12286@vax6k.openecs.org><3E5A710D.6030403@jetnet.ab.ca><3E60AEEB.4FBB6FD2@yahoo.com><20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> NNTP-Posting-Host: p-183.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.he.net!cyclone-sf.pbi.net!129.250.175.17!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews3 Xref: chonsp.franklin.ch alt.folklore.computers:131383 In article <20030307073602.4cd64ef0.steveo@eircom.net> steveo@eircom.net (Steve O'Hara-Smith) writes: >On Thu, 6 Mar 2003 18:13:46 -0500 >Charles Shannon Hendrix wrote: > >> In article <20030306065440.30f12471.steveo@eircom.net>, Steve >> O'Hara-Smith wrote: >> >>> Last time I heard this kind of figure being bandied around >>> it was indeed averaged over some long period and large number of >>> people to include all the things that consumed time but did not >>> generate code. The earliest incarnation I heard for an industry >>> average was three lines of fully debugged and documented COBOL per >>> day. >> >> I still don't know. >> >> I think I wrote 12K lines of code in 1999, mostly C and Perl, and >> never found any serious bugs. > > Sure - I've done similar things (Chris Hedley intercepted the >after the fact rumours). I think the metric is calculated at the end >of a project by counting the number of lines in the final system and >dividing by the number of programmer days it took to make it. Oh, just lovely. That sounds like a lovely way to penalize the poor programmer for all the delays incurred by being forced to sit in endless meetings and chasing unstable specs. Sounds like a PHB's dream come true. >>> It's a crazy metric though - I've lost count of the number >>> of times I've made a system do more and better by reducing the total >>> amount of code in it. >> >> ...sometimes drastically I'll bet. Back when I was taking over programs on a more regular basis, I averaged a 30% reduction in line count - up to 50% when the original programmer was sufficiently clueless. > Usually the more drastic the code reduction the greater the >improvement IME :) There is a persistent feeling that by the time >all the code is eliminated the system will be perfect :) "Perfection comes, not when there is nothing left to add, but when there is nothing left to take away." -- Antoine de Saint-Exupery -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 08 Mar 03 12:16:12 GMT Organization: UltraNet Communications, Inc. Lines: 26 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E68E636.E638D524@ev1.net> X-Trace: UmFuZG9tSVanQo1HyTkA2SXbxpOUgZbskon7ckYRbAYpFCZb7WSLbMHkpQSW+qr8 X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 8 Mar 2003 13:01:18 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news.tele.dk!news.tele.dk!small.news.tele.dk!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-211 Xref: chonsp.franklin.ch alt.folklore.computers:131352 In article <3E68E636.E638D524@ev1.net>, Charles Richmond wrote: >jmfbahciv@aol.com wrote: >> >> In article , >> Charles Shannon Hendrix wrote: >> >> >> >I was using a downloading program on Windows the other day, and almost >> >pounded a keyboard into dust. The damned program put up a dialog box >> >FOR EACH @#$%@# DOWNLOAD START! >> >> RIGHT!!! The fucking point of having computers was so you could >> start something up then go away and sleep or something. Having to >> stay and babysit an automatic process is just a side effect of that >> UDFH. >> >So you will like Unix/Linux/FreeBSD. They have commands that allow >you to set up your jobs to run "batch". I know. I did take Unix internals once. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 08 Mar 03 12:23:01 GMT Organization: UltraNet Communications, Inc. Lines: 52 Message-ID: References: X-Trace: UmFuZG9tSVYtcLDpqPSfjKgnmXzVngEZFdqXp2/S4uDVFQjYoRsfws46YbplpmqC X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 8 Mar 2003 13:08:08 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-211 Xref: chonsp.franklin.ch alt.folklore.computers:131354 In article , Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: > >>>I was using a downloading program on Windows the other day, and almost >>>pounded a keyboard into dust. The damned program put up a dialog box >>>FOR EACH @#$%@# DOWNLOAD START! >> >> RIGHT!!! The fucking point of having computers was so you could >> start something up then go away and sleep or something. Having to >> stay and babysit an automatic process is just a side effect of that >> UDFH. > >What drives me nuts is that in some programs, I can specify a download >directory. The program puts up a dialog for every download. OK, I can >see that the user might want to rename the file. However, why not an >option to avoid this query for when you just want to queue up a bunch >of downloads? Why not have full wildcard support or indirect files. > >Being on a modem link, I'm a patient downloader, and don't sit there >drooling while the progress bar moves to 100%. I hate babysitting little >tasks like this. Especially when they lie. "operation will take ~5 minutes" and the error^Wprogress? bar's advancement predicts 5 hours. > >The other day I was playing a map in a game called "Railroad Tycoon 2" >(delusions of grandeur can be healthy). The guy who made the map was >very clever, but he also figured out how to have the game advertise his >web site now and then while you play. > >>>About 99% of program->user communication can be done in the UI without >>>any kind of dialog, and that includes errors. >> >> Whoa!!! Did you just push a hot button. Why do I get the option >> of saying OK to no errors but, if there is an error, the information >> that I need gets flashered at me then the screen is drawn blank. > >Silly user, put down that mouse, and let us take care of everything... > >:) ROTFLMAO. Why bother flashing me? It's rather like the time my mother was cutting my hair when she said, "Ooops." /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 08 Mar 03 12:35:41 GMT Organization: UltraNet Communications, Inc. Lines: 80 Message-ID: References: <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> X-Trace: UmFuZG9tSVaw/R/T47OGXByWy+ou6Uudbbw1M/AbAH0Y9vWhg4lm04q9f3jMfbop X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 8 Mar 2003 13:20:49 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-211 Xref: chonsp.franklin.ch alt.folklore.computers:131323 In article , Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: > >>>Only if you want a limited one. >>> >>>Some filesystems are almost that big, like SGI's xfs. >> >> You don't need the file system in core to have one. > >Yes you do. There is no way to do it otherwise. > >Filesystems in any non-trivial system are never idle. If you tried to >keep the code out of core, you'd never get any work done. I didn't say "code"; I said "file system". You don't need the data for handling a device until somebody pushes the on-line button. > >>>Not all code is big because it is poorly written: some of it just does >>>more work. >>> >>>640K isn't even enough to hold the status information and buffers for >>>the processes on most heavily loaded servers. >> >> I have yet to meet a CPU that can deal with more than one >> thing at a time (that's why we have multiple-CPU systems). > >????? CPUs are synchronous machines. The art of timesharing OSes is make them appear to be able to do everything at once. > >> The reason all of that info has to be in core is because disk >> access is so slow. If you speed that up you don't need all >> of that data (that is there for JICs--just-in-case) in core. > >Even if a disk were as fast as RAM, I don't see why you would >want to keep the code on disk. You either need it or you >don't. Scenario: /BAH starts up a task that requires the disk structure BARB to be spinning. So the pack is mounted, on-line procedures are completed, and /BAH goes back to office to massage a few bits. A priority interrupt is generated by a PHB or a bit god or a next door neighbor who just did an Ooops!. Until I do a DIRECT BARB:[10,2162]FOO.*, the data needed to access the bits doesn't have to be in core. This is a case where service on demand is a more efficient use of core. > >Pushing things in and out of core is usually not good unless it >is very infrequently used. My point is that no OS will know when or how much data is going to be used. > >Most OS routines are going to be in use almost 24/7, so I don't see much >benefit in making them swappable. No, they're not. If you've got a compute-bound job that does no I/O, the device driver code will NEVER, I mean never, be executed. > >Now, your USB drivers... they might only get used once in awhile. >Lot's of things fall into that catagory. But the bulk of a UNIX OS and >its drivers are going to be in use all the time. There just won't be >a time of non-use long enough to justify moving it out. This tells me that you've never dealt with a real timesharing OS at its instruction level. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 08 Mar 03 14:51:15 GMT Organization: UltraNet Communications, Inc. Lines: 35 Message-ID: References: <3E68E636.E638D524@ev1.net> <3E6A0951.10371A99@ev1.net> X-Trace: UmFuZG9tSVYIqtjgKhuSiPCRipTZVCyVPkgFgJpcOsvVRupqrbdLYejqyPjhkShM X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 8 Mar 2003 15:36:20 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!209-122-234-233 Xref: chonsp.franklin.ch alt.folklore.computers:131319 In article <3E6A0951.10371A99@ev1.net>, Charles Richmond wrote: >jmfbahciv@aol.com wrote: >> I know. I did take Unix internals once. >> >Well, then, you will be right at home with Linux...it's >pretty much all the same. You might need 64 meg of memory >and at least 2 or 3 gigs of disk space. All of that immediately pops out as tons. :-) I seem to not have excised all subroutines of old iron calculations. > ... A Pentium or >Pentium II processor at least would be wonderful... If you >do *not* have this system, go look at some of the local >thrift stores for a used system. The Micky-Suck people are >throwing away hardware of this description. They are? Already? > >But carry a *lot* of cash!!! The system might cost you as >much as twenty or thirty dollars US... ;-) I hate shopping. First I'll have to find the store...JMF was very useful for this; I could always send him out to go hunting. Once he spent two entire afternoons hunting for 15-pound fishing line and he knew nothing about fishing. /BAH Subtract a hundred and four for e-mail. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E60AEEB.4FBB6FD2@yahoo.com> <20030306065440.30f12471.steveo@eircom.net> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 30 Date: Thu, 6 Mar 2003 18:13:46 -0500 NNTP-Posting-Host: 206.246.249.197 X-Trace: sydney.visi.net 1046996768 206.246.249.197 (Thu, 06 Mar 2003 19:26:08 EST) NNTP-Posting-Date: Thu, 06 Mar 2003 19:26:08 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!nntp.abs.net!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131250 In article <20030306065440.30f12471.steveo@eircom.net>, Steve O'Hara-Smith wrote: > Last time I heard this kind of figure being bandied around it > was indeed averaged over some long period and large number of people to > include all the things that consumed time but did not generate code. > The earliest incarnation I heard for an industry average was three lines > of fully debugged and documented COBOL per day. I still don't know. I think I wrote 12K lines of code in 1999, mostly C and Perl, and never found any serious bugs. Of course, it was strictly data processing, and I did some C libraries that handled the pointer nasties, so I only had one chance to make the errors. A lot of people don't like doing C like that, but it sure does cut down on the errors. > It's a crazy metric though - I've lost count of the number of > times I've made a system do more and better by reducing the total amount > of code in it. ...sometimes drastically I'll bet. ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 6 Mar 2003 23:21:48 -0000 Organization: Titanic Enterprises Unlimited Lines: 28 Message-ID: <1046992908.865935@saucer.planet.gong> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D4816.C974A441@yahoo.com> <6ud04b.lah.ln@escape.shannon.net> <1046778084.707238@saucer.planet.gong> <1046919297.984334@saucer.planet.gong> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1046992909 19454 80.177.7.220 (6 Mar 2003 23:21:49 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Thu, 6 Mar 2003 23:21:49 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Cache: nntpcache 3.0.1 (see http://www.nntpcache.org/) X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Cache-Post-Path: saucer.planet.gong!unknown@voodoo.planet.gong Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-lei1.dfn.de!newsfeed00.sul.t-online.de!t-online.de!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131377 "Charles Shannon Hendrix" wrote in message news:slrnb6fefs.og2.shannon@news.widomaker.com... > In article <1046919297.984334@saucer.planet.gong>, Rupert Pigott wrote: > > "Charles Shannon Hendrix" wrote in message > > news:slrnb6d33u.n0s.shannon@news.widomaker.com... > > [SNIP] > > > >> Speaking of blinkenlights, you'd have to have a Thinking Machines system > >> in there. The one with the arrays of lights and the thousands of CPUs. > > > > Hmm, probably mis-remembering the model designations but : > > CM-2, CM-200, TM-5 ? > > CM-2: CM-5 had more blinkenlights. Assuming you could justify a huge amount of money on something that had the same relevance as the Dodo. I'm being mean again... ;) CM-200 was the floating point version of the CM-2 and it had far more promising performance IIRC. Cheers, Ruppert ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <3E66A111.9060306@jetnet.ab.ca> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 18 Date: Thu, 6 Mar 2003 18:56:31 -0500 NNTP-Posting-Host: 206.246.249.197 X-Trace: sydney.visi.net 1046996772 206.246.249.197 (Thu, 06 Mar 2003 19:26:12 EST) NNTP-Posting-Date: Thu, 06 Mar 2003 19:26:12 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!rcn!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131488 In article <3E66A111.9060306@jetnet.ab.ca>, bfranchuk@jetnet.ab.ca wrote: > Charles Shannon Hendrix wrote: > >> What I don't understand is why a modern PC BIOS doesn't have a "No DOS" >> flag for when you don't need the 640K split and other DOS/Windows hooks, >> and why they don't have a 32-bit native loader after all these years. > > > Because they still have the stupid XT memory map! Also modern I/O is > very software bloated to set up. My gripe is that the bios never set the > timer chip to something useful like 50 or 60 HZ. But wait the XT was > to slow for that. Ben. Yes, but I said above why not without the DOS support, or make it optional. At that point, the XT memory map becmes irrelevant. Almost no hardware or software needs it any more. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 19 Date: Thu, 6 Mar 2003 18:57:46 -0500 NNTP-Posting-Host: 206.246.249.197 X-Trace: sydney.visi.net 1046996773 206.246.249.197 (Thu, 06 Mar 2003 19:26:13 EST) NNTP-Posting-Date: Thu, 06 Mar 2003 19:26:13 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!iad-peer.news.verio.net!news.verio.net!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131494 In article , jmfbahciv@aol.com wrote: > In article <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com>, > Brian Inglis wrote: > > >>640K ought to be enough for any bootstrap! ;^> > > And an OS. Only if you want a limited one. Some filesystems are almost that big, like SGI's xfs. Not all code is big because it is poorly written: some of it just does more work. 640K isn't even enough to hold the status information and buffers for the processes on most heavily loaded servers. ###### From: Ben Hutchings Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 7 Mar 2003 02:28:43 +0000 Lines: 43 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Reply-To: Ben Hutchings NNTP-Posting-Host: pc5-cmbg1-6-cust149.cmbg.cable.ntl.com (80.4.127.149) X-Trace: fu-berlin.de 1047004206 64740767 80.4.127.149 (16 [70929]) X-Orig-Path: 192.168.1.1!nobody User-Agent: slrn/0.9.7.4 (Linux) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.belwue.de!fu-berlin.de!uni-berlin.de!pc5-cmbg1-6-cust149.cmbg.cable.ntl.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131257 In article , Charles Shannon Hendrix wrote: > In article >, Ben > Hutchings wrote: > >>> The exact limitations aren't even the same across all BIOS, which makes >>> the problem even more "interesting". >> >> There are two different limits on drive geometry: the original BIOS >> disk API is limited to 1024 cylinders, 255 heads (!) and 63 sectors; >> the original AT HD controller (which ATA/IDE disks remain compatible >> with) is limited to 16383 cylinders, 16 heads and 63 sectors. Combine >> these limitations and you get just under 1024 * 16 * 63 sectors ~= >> 500 MB. > > As I said, I'm not talking about drive geometry. > > I'm talking about the size of the image that the BIOS function your > 446/512 byte loader calls. > > I thought there was a limit to that. Except in the case of a floppy disk or a virtual floppy disk on a CD-ROM, the size limitation is that of the BIOS disk API. > If not, then why do kernels have to be small, They don't. > and/or why do other OS put so much effort into an interrim stage > which loads the kernels? The boot loader needs to work with any hard disk, so it has to use BIOS calls to load, so it has to run in real mode, so it's limited to accessing the first 1 MB of memory. Well, actually it could switch back and forth between real and protected mode, but that would be a pain. -- Ben Hutchings | personal web site: http://womble.decadentplace.org.uk/ 73.46% of all statistics are made up. ###### From: Giles Todd Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 07 Mar 2003 05:53:09 +0100 Organization: The action of organizing or putting into systematic form; the arranging and co-ordinating of parts into a systematic whole. Reply-To: gt@at-dot.org Message-ID: Cancel-Lock: sha1:mGvg9reLgWpMsk8zdBqVGKK5338= References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> X-Newsreader: Forte Agent 1.92/32.572 X-No-Archive: yes MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-NFilter: 1.2.0 Lines: 12 NNTP-Posting-Date: 07 Mar 2003 05:53:23 CET NNTP-Posting-Host: 80.127.198.243 X-Trace: 1047012803 news.xs4all.nl 49101 80.127.198.243:2956 X-Complaints-To: abuse@xs4all.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!solnet.ch!solnet.ch!newsfeed.freenet.de!newsfeed.icl.net!newsfeed.fjserv.net!lnewspeer00.lnd.ops.eu.uu.net!emea.uu.net!news2.euro.net!transit.news.xs4all.nl!newsfeed.xs4all.nl!xs4all!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131451 On Thu, 6 Mar 2003 16:38:31 -0500, Charles Shannon Hendrix wrote in message : > There is also little variety now in the home and educational market, > and the focus is on learning very basic skills instead of learning > how the machine works. Maybe the focus there is on using the machines for what they can do rather than studying what they are. Giles. ###### Date: Fri, 7 Mar 2003 07:36:02 +0100 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <20030307073602.4cd64ef0.steveo@eircom.net> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E60AEEB.4FBB6FD2@yahoo.com> <20030306065440.30f12471.steveo@eircom.net> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 38 Organization: EuroNet Internet NNTP-Posting-Date: 07 Mar 2003 10:28:55 GMT NNTP-Posting-Host: p3014.vwr.wanadoo.nl X-Trace: 1047032935 maya.euronet.nl 23912 194.134.219.207:1318 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-lei1.dfn.de!newsfeed00.sul.t-online.de!newsfeed01.sul.t-online.de!t-online.de!newspeer1-gui.server.ntli.net!ntli.net!news2.euro.net!postnews1.euro.net!maya.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131392 On Thu, 6 Mar 2003 18:13:46 -0500 Charles Shannon Hendrix wrote: CSH> In article <20030306065440.30f12471.steveo@eircom.net>, Steve CSH> O'Hara-Smith wrote: CSH> CSH> > Last time I heard this kind of figure being bandied around it CSH> > was indeed averaged over some long period and large number of CSH> > people to include all the things that consumed time but did not CSH> > generate code. The earliest incarnation I heard for an industry CSH> > average was three lines of fully debugged and documented COBOL per CSH> > day. CSH> CSH> I still don't know. CSH> CSH> I think I wrote 12K lines of code in 1999, mostly C and Perl, and CSH> never found any serious bugs. Sure - I've done similar things (Chris Hedley intercepted the after the fact rumours). I think the metric is calculated at the end of a project by counting the number of lines in the final system and dividing by the number of programmer days it took to make it. CSH> > It's a crazy metric though - I've lost count of the number of CSH> > times I've made a system do more and better by reducing the total CSH> > amount of code in it. CSH> CSH> ...sometimes drastically I'll bet. Usually the more drastic the code reduction the greater the improvement IME :) There is a persistent feeling that by the time all the code is eliminated the system will be perfect :) -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### From: "Peter Ibbotson" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 7 Mar 2003 09:12:48 -0000 Lines: 24 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> NNTP-Posting-Host: mailgate.lakeview.co.uk X-Trace: news.demon.co.uk 1047028368 2079 62.49.243.90 (7 Mar 2003 09:12:48 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Fri, 7 Mar 2003 09:12:48 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-fra1.dfn.de!news0.de.colt.net!peernews3.colt.net!colt.net!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131386 "Charles Shannon Hendrix" wrote in message news:slrnb6ffum.og2.shannon@news.widomaker.com... > About 99% of program->user communication can be done in the UI without > any kind of dialog, and that includes errors. > > And yet, most GUI software seems to use dialogs for every little trivial > thing, thus causing the system UI to interrupt you constantly. I suspect thats because: a) It's quick to write. b) For Windows systems it stops the message queue and returns back to the same place in the code it was called from. Most of the other solutions require paint messages to be processed, which often also allows menu messages to be processed which can cause reenterancy problems. Item b is probably more important, but quite how many programmers realise why is a good question (back to point & drool again). -- Work peteri@lakeview.co.uk.plugh.org | remove magic word .org to reply Home peter@ibbotson.co.uk.plugh.org | I own the domain but theres no MX ###### From: Sami S. Sihvonen Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 07 Mar 2003 14:54:37 +0200 Organization: Janiika Networks Corporation Lines: 13 Sender: root@janiika.com Message-ID: <7rff6v85rmumqchtcau0159cntapekompb@sex.bar> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> Reply-To: Sami S. Sihvonen NNTP-Posting-Host: p-d51cbe96.easy.inet.fi (213.28.190.150) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: fu-berlin.de 1047041511 65415030 213.28.190.150 (16 [158501]) X-Newsreader: GNU Emacs v21.3.50/Gnus v5.9.0 (Debian GNU/Linux) X-PGPKey: 1024/A99F9CBD 1995/07/14 (available at public keyservers) X-Pages: http://www.sunpoint.net/~noco/ (finnish/english) X-Other-News: Forte Agent 1.8/16.553 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!p-d51cbe96.easy.inet.FI!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131502 In article <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com>, Brian Inglis wrote: > 640K ought to be enough for any bootstrap! ;^> GNU/Linux does not care about that silly 640K limit... (1252k kernel code, 412k reserved, 444k data, 60k init) -- Sami Sihvonen, Senior Unix Administrator, Janiika Networks Corporation US-Mail: PO Box 1490, Radio City Station, New York, N.Y. 10019, U.S.A. EU-Mail: PO Box 42, Kauppalankatu, FIN-45101 Kouvola, Finland, Europe. ###### Message-ID: <3E68E636.E638D524@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 22 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc52.ops.asp.att.net 1047054977 12.241.15.59 (Fri, 07 Mar 2003 16:36:17 GMT) NNTP-Posting-Date: Fri, 07 Mar 2003 16:36:17 GMT Date: Fri, 07 Mar 2003 16:36:17 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!24.215.0.32.MISMATCH!newsfeed.mountaincable.net!news-out.nuthinbutnews.com!propagator2-sterling!news-in-sterling.newsfeed.com!news-in.nuthinbutnews.com!cyclone1.gnilink.net!wn14feed!wn13feed!wn12feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!rwcrnsc52.ops.asp.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131436 jmfbahciv@aol.com wrote: > > In article , > Charles Shannon Hendrix wrote: > > > >I was using a downloading program on Windows the other day, and almost > >pounded a keyboard into dust. The damned program put up a dialog box > >FOR EACH @#$%@# DOWNLOAD START! > > RIGHT!!! The fucking point of having computers was so you could > start something up then go away and sleep or something. Having to > stay and babysit an automatic process is just a side effect of that > UDFH. > So you will like Unix/Linux/FreeBSD. They have commands that allow you to set up your jobs to run "batch". -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### From: Greg Finn Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 07 Mar 2003 09:32:55 -0800 Organization: USC Information Sciences Institute Lines: 27 Sender: finn@pbs.isi.edu Message-ID: <6w65qvozeg.fsf@pbs.isi.edu> References: <3E05FE9D.B3A12286@vax6k.openecs.org> NNTP-Posting-Host: pbs.isi.edu Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii User-Agent: Gnus/5.0808 (Gnus v5.8.8) XEmacs/21.4 (Honest Recruiter) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!colt.net!nycmny1-snf1.gtei.net!nycmny1-snh1.gtei.net!news.gtei.net!newsfeed.mathworks.com!news-out.newsfeeds.com!propagator2-maxim!news-in.spamkiller.net!usc.edu!news.isi.edu!usenet Xref: chonsp.franklin.ch alt.folklore.computers:131314 alt.sys.pdp10:12714 Mark Crispin writes: > On Sun, 23 Feb 2003, John R. Levine wrote: > > Considering all the microcode bugs > > that showed up in the Vax, I get the impression that they designed the > > instruction set to match what they thought programmers or compiler > > writers would want without a whole lot of thought about how you would > > implement it other than in a naive low-performance version like the > > 730. > This always seemed obvious to me. More specifically, the VAX instruction > set always looked to me like a PDP-11 assembly language programmer's idea > of what a compiler writer would want. I've heard that much of the VAX > instruction set was never used by compilers. > The PDP-10 instruction set, on the other hand, was an assembly language > programmer's idea of what an assembly language programmer would want! ;-) I would echo that, but to belabor the obvious, the machine was also specifically designed to make higher-level languages easy to compile. Design input for the PDP-6/-10 came from Marvin Minsky, John McCarthy and Peter Sampson. The curious can read: The Evolution of the DECsystem-10 Bell, Kotok, Hastings & Hill Comm. ACM, Jan. 1978. ###### From: stanb45@dial.pipex.com (Stan Barr) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E68E636.E638D524@ev1.net> Organization: Metropolis Grafix Reply-To: stanb45@dial.pipex.com Message-ID: X-Newsreader: slrn (0.9.5.2 UNIX) Date: 07 Mar 2003 19:10:29 GMT Lines: 28 NNTP-Posting-Host: 62-241-188-23.dsl.pipex.com X-Trace: 1047064229 news.dial.pipex.com 5366 62.241.188.23 X-Complaints-To: abuse@uk.uu.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!eusc.inter.net!newsfeed01.sul.t-online.de!t-online.de!uucp.muenster.de!195.129.110.29!dnewsfeed02.dtm.ops.eu.uu.net!dnewsifeed01.dtm.ops.eu.uu.net!lnewsifeed02.lnd.ops.eu.uu.net!lnewspost00.lnd.ops.eu.uu.net!emea.uu.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131328 On Fri, 07 Mar 2003 16:36:17 GMT, Charles Richmond wrote: >jmfbahciv@aol.com wrote: >> >> In article , >> Charles Shannon Hendrix wrote: >> >> >> >I was using a downloading program on Windows the other day, and almost >> >pounded a keyboard into dust. The damned program put up a dialog box >> >FOR EACH @#$%@# DOWNLOAD START! >> >> RIGHT!!! The fucking point of having computers was so you could >> start something up then go away and sleep or something. Having to >> stay and babysit an automatic process is just a side effect of that >> UDFH. >> >So you will like Unix/Linux/FreeBSD. They have commands that allow >you to set up your jobs to run "batch". And many Mac applications are script-aware, allowing them to be run from Applescript programs. -- Cheers, Stan Barr stanb45@dial.pipex.com **Remove the digits from email address** The future was never like this! ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D4816.C974A441@yahoo.com> <6ud04b.lah.ln@escape.shannon.net> <1046778084.707238@saucer.planet.gong> <1046919297.984334@saucer.planet.gong> <1046992908.865935@saucer.planet.gong> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 16 Date: Fri, 7 Mar 2003 15:34:47 -0500 NNTP-Posting-Host: 209.96.189.100 X-Trace: sydney.visi.net 1047072369 209.96.189.100 (Fri, 07 Mar 2003 16:26:09 EST) NNTP-Posting-Date: Fri, 07 Mar 2003 16:26:09 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131462 In article <1046992908.865935@saucer.planet.gong>, Rupert Pigott wrote: > CM-5 had more blinkenlights. Assuming you could justify > a huge amount of money on something that had the same > relevance as the Dodo. I'm being mean again... ;) Whatever. I saw things done on them that nothing else I could get access too could do. At that particular time, what other massively parallel machines were better? ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 24 Date: Fri, 7 Mar 2003 15:36:10 -0500 NNTP-Posting-Host: 209.96.189.100 X-Trace: sydney.visi.net 1047072370 209.96.189.100 (Fri, 07 Mar 2003 16:26:10 EST) NNTP-Posting-Date: Fri, 07 Mar 2003 16:26:10 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131477 In article , Giles Todd wrote: > On Thu, 6 Mar 2003 16:38:31 -0500, Charles Shannon Hendrix > wrote in message > : > >> There is also little variety now in the home and educational market, >> and the focus is on learning very basic skills instead of learning >> how the machine works. > > Maybe the focus there is on using the machines for what they can do > rather than studying what they are. Why not both? Also, I don't think you can know what they can do without knowing a little about what they are. Finally, why teach the basic skills largely at the expense of any others? At least make it available for those who do want to learn more. Microsoft wants to turn them into appliances you can't learn anything from. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 34 Date: Fri, 7 Mar 2003 15:39:43 -0500 NNTP-Posting-Host: 209.96.189.100 X-Trace: sydney.visi.net 1047072371 209.96.189.100 (Fri, 07 Mar 2003 16:26:11 EST) NNTP-Posting-Date: Fri, 07 Mar 2003 16:26:11 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!nntp-out.monmouth.com!newspeer.monmouth.com!petbe.visi.com!news-out.visi.com!hermes.visi.com!out.nntp.be!propagator-sanjose!news-in-sanjose!cyclone-sf.pbi.net!216.218.192.242!news.he.net!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131474 In article , Peter Ibbotson wrote: > I suspect thats because: > a) It's quick to write. It's no quicker than a defined routine for doing it another way... :) > b) For Windows systems it stops the message queue and returns back to the > same place in the code it was called from. Most of the other solutions > require paint messages to be processed, which often also allows menu > messages to be processed which can cause reenterancy problems. Hmmm... post a message with a routine that puts the message in the UI without a dialog. The routine also sets a time to live so the message goes away. Since there are good programs which do this, it must be possible in Windows. In fact, I'm not a Windows programmer, but I thought there was a message window class which would do exactly this. > Item b is probably more important, but quite how many programmers realise > why is a good question (back to point & drool again). On a text program, I handled this as I described above. It was a little work of course. Well, it took me a week to get it right, but that's because I've a head full of rocks sometimes. But it displayed a message in a message area, and a cleanup routine swept it away N seconds later. Sometimes it beeped, but I rarely did anything which interrupted work or slowed you down. I hate programs which notify me of an error, and then make me wait several seconds before I can move on. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 38 Date: Fri, 7 Mar 2003 15:56:41 -0500 NNTP-Posting-Host: 209.96.189.100 X-Trace: sydney.visi.net 1047072373 209.96.189.100 (Fri, 07 Mar 2003 16:26:13 EST) NNTP-Posting-Date: Fri, 07 Mar 2003 16:26:13 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131468 In article , jmfbahciv@aol.com wrote: >>I was using a downloading program on Windows the other day, and almost >>pounded a keyboard into dust. The damned program put up a dialog box >>FOR EACH @#$%@# DOWNLOAD START! > > RIGHT!!! The fucking point of having computers was so you could > start something up then go away and sleep or something. Having to > stay and babysit an automatic process is just a side effect of that > UDFH. What drives me nuts is that in some programs, I can specify a download directory. The program puts up a dialog for every download. OK, I can see that the user might want to rename the file. However, why not an option to avoid this query for when you just want to queue up a bunch of downloads? Being on a modem link, I'm a patient downloader, and don't sit there drooling while the progress bar moves to 100%. I hate babysitting little tasks like this. The other day I was playing a map in a game called "Railroad Tycoon 2" (delusions of grandeur can be healthy). The guy who made the map was very clever, but he also figured out how to have the game advertise his web site now and then while you play. >>About 99% of program->user communication can be done in the UI without >>any kind of dialog, and that includes errors. > > Whoa!!! Did you just push a hot button. Why do I get the option > of saying OK to no errors but, if there is an error, the information > that I need gets flashered at me then the screen is drawn blank. Silly user, put down that mouse, and let us take care of everything... :) ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 45 Date: Fri, 7 Mar 2003 16:03:21 -0500 NNTP-Posting-Host: 209.96.189.100 X-Trace: sydney.visi.net 1047075965 209.96.189.100 (Fri, 07 Mar 2003 17:26:05 EST) NNTP-Posting-Date: Fri, 07 Mar 2003 17:26:05 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131466 In article , jmfbahciv@aol.com wrote: >>Only if you want a limited one. >> >>Some filesystems are almost that big, like SGI's xfs. > > You don't need the file system in core to have one. Yes you do. There is no way to do it otherwise. Filesystems in any non-trivial system are never idle. If you tried to keep the code out of core, you'd never get any work done. >>Not all code is big because it is poorly written: some of it just does >>more work. >> >>640K isn't even enough to hold the status information and buffers for >>the processes on most heavily loaded servers. > > I have yet to meet a CPU that can deal with more than one > thing at a time (that's why we have multiple-CPU systems). ????? > The reason all of that info has to be in core is because disk > access is so slow. If you speed that up you don't need all > of that data (that is there for JICs--just-in-case) in core. Even if a disk were as fast as RAM, I don't see why you would want to keep the code on disk. You either need it or you don't. Pushing things in and out of core is usually not good unless it is very infrequently used. Most OS routines are going to be in use almost 24/7, so I don't see much benefit in making them swappable. Now, your USB drivers... they might only get used once in awhile. Lot's of things fall into that catagory. But the bulk of a UNIX OS and its drivers are going to be in use all the time. There just won't be a time of non-use long enough to justify moving it out. ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 7 Mar 2003 22:10:41 -0000 Organization: Titanic Enterprises Unlimited Lines: 38 Message-ID: <1047075041.277393@saucer.planet.gong> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D4816.C974A441@yahoo.com> <6ud04b.lah.ln@escape.shannon.net> <1046778084.707238@saucer.planet.gong> <1046919297.984334@saucer.planet.gong> <1046992908.865935@saucer.planet.gong> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1047075041 25648 80.177.7.220 (7 Mar 2003 22:10:41 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Fri, 7 Mar 2003 22:10:41 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Cache: nntpcache 3.0.1 (see http://www.nntpcache.org/) X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Cache-Post-Path: saucer.planet.gong!unknown@voodoo.planet.gong Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newspump.monmouth.com!newspeer.monmouth.com!news.maxwell.syr.edu!newsfeed.icl.net!newsfeed.fjserv.net!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131369 "Charles Shannon Hendrix" wrote in message news:slrnb6i0j7.12h.shannon@news.widomaker.com... > In article <1046992908.865935@saucer.planet.gong>, Rupert Pigott wrote: > > > CM-5 had more blinkenlights. Assuming you could justify > > a huge amount of money on something that had the same > > relevance as the Dodo. I'm being mean again... ;) > > Whatever. > > I saw things done on them that nothing else I could get access too > could do. > > At that particular time, what other massively parallel machines were > better? True it was early(ish). On paper there were a number of things which stood out as being poor. Firstly the communications bandwidth & the topology. Secondly the vector units bolted onto the side of the SPARC chips seemed a bit, err, tricky to use. CM-5s looked as if they were designed for a specific application. I have no doubt that you could get some apps to fly on those boxes. Of course my assessment was heavily biased because I was around as the T9000 was being designed (too early in the game to see it's shortcomings). I'd like to learn more about CM-5s though, they looked interesting. :) Cheers, Rupert ###### Lines: 33 X-Admin: news@aol.com From: rsclient@aol.com (Rsclient) Newsgroups: alt.folklore.computers Date: 08 Mar 2003 00:41:46 GMT References: Organization: AOL http://www.aol.com X-Newsreader: Session Scheduler (Queue Name: gng-cu) Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <20030307194146.19751.00003178@mb-cu.aol.com> Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news.ip-plus.net!news.tesion.net!news.belwue.de!newsfeed01.sul.t-online.de!t-online.de!newspeer1-gui.server.ntli.net!ntli.net!cox.net!cyclone1.gnilink.net!ngpeer.news.aol.com!audrey-m1.news.aol.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131364 In article , Pete Fenelon writes: >I've found that interviewing recent graduates reveals that their >knowledge of what goes on inside complex systems (OS, device driver, >compiler, debugger, filesystem) is often weak to nonexistent - but >they're a lot happier scripting, reusing off the shelf components, >and doing user-interface work than "oldtimers". > I suppose all professions feel the same way: Accounting: "these new accountants just come in and say, 'use double entry bookkeeping'. In my day, we had to design our own accounting methods, and really understood what we were doing." Shipbuilding: "these new crew just come in and expect to just pull on lines and sheets. In my day, we had to make our own rope, and really understood what how a lines and sheets really worked" Soldiers: "these new soldiers expect to just come waltzing in here and use a chariot. In my day, we had to build your own chariot, and really understood what using a chariot was all about" I suspect that computing is just getting mature. Why aren't there new language paradigms? Because we've tried a bunch by now, and some clearly work (procedural languages) and some just don't (graphical languages). Peter "Young fart since 1963" ###### Message-ID: <3E697D08.36BD1F8E@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E68E636.E638D524@ev1.net> <1539.196T2652T7445477@kltpzyxm.invalid> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 21 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: sccrnsc03 1047093586 12.241.15.59 (Sat, 08 Mar 2003 03:19:46 GMT) NNTP-Posting-Date: Sat, 08 Mar 2003 03:19:46 GMT Date: Sat, 08 Mar 2003 03:19:46 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!wn14feed!wn13feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!sccrnsc03.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131444 Charlie Gibbs wrote: > > [snip...] [snip...] [snip...] > > But... but... all _modern_ systems are INTERACTIVE! Batch programming > is a throwback to the dinosaurs! Nobody who wants to be With the > Program [tm] would think of using that archaic stuff! > > These silly people. They don't realize that all too often, > "interactive" is synonymous with "manual". > IMHO interactive is wonderful for software *development*. Also, an interactive enviroment is fine for writing or editing of text. In a production environment, where programs are run on huge amounts of data to create large output files, batch (unattended mode) is what works best... -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### From: Kenneth Seefried Newsgroups: alt.folklore.computers,alt.sys.pdp10,alt.os.multics Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D4816.C974A441@yahoo.com> <6ud04b.lah.ln@escape.shannon.net> <1046778084.707238@saucer.planet.gong> X-Newsreader: Forte Agent 1.91/32.564 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 28 NNTP-Posting-Host: 66.56.42.205 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc53 1047096748 66.56.42.205 (Sat, 08 Mar 2003 04:12:28 GMT) NNTP-Posting-Date: Sat, 08 Mar 2003 04:12:28 GMT Organization: AT&T Broadband Date: Sat, 08 Mar 2003 04:12:28 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!wn14feed!wn13feed!wn12feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!rwcrnsc53.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131443 alt.sys.pdp10:12724 On Tue, 4 Mar 2003 11:41:24 -0000, "Rupert Pigott" wrote: >"Charles Shannon Hendrix" wrote in message >news:6ud04b.lah.ln@escape.shannon.net... > >> My list wasn't complete of course. I left out the Crays, the Cyber 205 >> that would be a must, and a few others. > >ETA-10P or ETA-10Q would be a nifty addition to that. >Perhaps some Convex boxes too... Crays yes, especially a 1 or a 2, before they went off the rails with the XMP & YMP...but Cyber 205s & ETA? I dunno...they were interesting and fun, but I think a Multiflow Trace would be more apropos. I haven't seen what long term impact the former had, but the latter still lives in the compiler technology for the Itanium, among other VLIW innovations. FWIW...for all of the otherwise wonderful facilities, Georgia Tech never had a decent supercomputer while I was there. Well...we had a Cyber 180/990, which was a vector machine in the colostomy bag sense of the phrase. I actually had to go to our bitter collegiate rival UGA to use a Cyber 205 (and to Pittsburg Supercomputing Center for a YMP). Ken ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D4816.C974A441@yahoo.com> <6ud04b.lah.ln@escape.shannon.net> <1046778084.707238@saucer.planet.gong> <1046919297.984334@saucer.planet.gong> <1046992908.865935@saucer.planet.gong> <1047075041.277393@saucer.planet.gong> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 23 Date: Sat, 8 Mar 2003 01:11:22 -0500 NNTP-Posting-Host: 209.96.179.178 X-Trace: sydney.visi.net 1047108362 209.96.179.178 (Sat, 08 Mar 2003 02:26:02 EST) NNTP-Posting-Date: Sat, 08 Mar 2003 02:26:02 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!nntp.abs.net!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131490 In article <1047075041.277393@saucer.planet.gong>, Rupert Pigott wrote: >> At that particular time, what other massively parallel machines were >> better? > > True it was early(ish). > > On paper there were a number of things which stood out > as being poor. Firstly the communications bandwidth & > the topology. Secondly the vector units bolted onto the > side of the SPARC chips seemed a bit, err, tricky to > use. I don't doubt it had problems, but at the time there just were not a lot of those systems around. Nothing that looks that damned cool can be useless... :) > I'd like to learn more about CM-5s though, they looked > interesting. :) Mostly all I've seen is CGI movies made on them. ###### Message-ID: <3E6A0951.10371A99@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E68E636.E638D524@ev1.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 34 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc51.ops.asp.att.net 1047129498 12.241.15.59 (Sat, 08 Mar 2003 13:18:18 GMT) NNTP-Posting-Date: Sat, 08 Mar 2003 13:18:18 GMT Date: Sat, 08 Mar 2003 13:18:18 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!wn14feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!rwcrnsc51.ops.asp.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131430 jmfbahciv@aol.com wrote: > > In article <3E68E636.E638D524@ev1.net>, > Charles Richmond wrote: > >jmfbahciv@aol.com wrote: > >> > >> [snip...] [snip...] [snip...] > >> > >> RIGHT!!! The fucking point of having computers was so you could > >> start something up then go away and sleep or something. Having to > >> stay and babysit an automatic process is just a side effect of that > >> UDFH. > >> > > So you will like Unix/Linux/FreeBSD. They have commands that allow > > you to set up your jobs to run "batch". > > I know. I did take Unix internals once. > Well, then, you will be right at home with Linux...it's pretty much all the same. You might need 64 meg of memory and at least 2 or 3 gigs of disk space. A Pentium or Pentium II processor at least would be wonderful... If you do *not* have this system, go look at some of the local thrift stores for a used system. The Micky-Suck people are throwing away hardware of this description. But carry a *lot* of cash!!! The system might cost you as much as twenty or thirty dollars US... ;-) -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 8 Mar 2003 13:22:48 -0000 Organization: Titanic Enterprises Unlimited Lines: 40 Message-ID: <1047129768.313584@saucer.planet.gong> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5D4816.C974A441@yahoo.com> <6ud04b.lah.ln@escape.shannon.net> <1046778084.707238@saucer.planet.gong> <1046919297.984334@saucer.planet.gong> <1046992908.865935@saucer.planet.gong> <1047075041.277393@saucer.planet.gong> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1047129768 25717 80.177.7.220 (8 Mar 2003 13:22:48 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Sat, 8 Mar 2003 13:22:48 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Cache: nntpcache 3.0.1 (see http://www.nntpcache.org/) X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Cache-Post-Path: saucer.planet.gong!unknown@voodoo.planet.gong Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!solnet.ch!solnet.ch!newsfeed.freenet.de!newsfeed00.sul.t-online.de!t-online.de!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131374 "Charles Shannon Hendrix" wrote in message news:slrnb6j2ca.4qp.shannon@news.widomaker.com... > In article <1047075041.277393@saucer.planet.gong>, Rupert Pigott wrote: > > >> At that particular time, what other massively parallel machines were > >> better? > > > > True it was early(ish). > > > > On paper there were a number of things which stood out > > as being poor. Firstly the communications bandwidth & > > the topology. Secondly the vector units bolted onto the > > side of the SPARC chips seemed a bit, err, tricky to > > use. > > I don't doubt it had problems, but at the time there just were not a > lot of those systems around. > > Nothing that looks that damned cool can be useless... :) > > > I'd like to learn more about CM-5s though, they looked > > interesting. :) > > Mostly all I've seen is CGI movies made on them. That figures, not wanting to detract from the CM-5, but that kind of app is usually trivial to parallelise. It probably wouldn't stretch it's interconnect too hard. Pixar were using T8's for a while before the CM-5 turned up. The last I heard was that the T8s were mostly used for the networking, not computation. Certainly by the time the CM-5 came along T8 FP performance was not cutting edge anymore. ;) Cheers, Rupert ###### Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> Organization: Honest Chris' Sysadmin Emporium X-Newsreader: trn 4.0-test76 (Apr 2, 2001) From: cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) Originator: cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) Date: Sat, 8 Mar 2003 14:57:04 +0000 Message-ID: <0c0d4b.tr1.ln@teabag.cbhnet> Lines: 24 NNTP-Posting-Date: 08 Mar 2003 15:11:01 GMT NNTP-Posting-Host: 62.254.134.35 X-Trace: 1047136261 news.gradwell.net 66675 cbh/62.254.134.35 X-Complaints-To: news-abuse@gradwell.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed00.sul.t-online.de!newsfeed01.sul.t-online.de!t-online.de!newspeer1-gui.server.ntli.net!ntli.net!nntp.theplanet.net!inewsm1.nntp.theplanet.net!195.40.4.120.MISMATCH!easynet-quince!easynet.net!news-peer.gradwell.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131296 According to Steve O'Hara-Smith : > Sure - I've done similar things (Chris Hedley intercepted the > after the fact rumours). I think the metric is calculated at the end > of a project by counting the number of lines in the final system and > dividing by the number of programmer days it took to make it. I did indeed, amazing how little time had taken for certain bits to become legends in their own right. :) Shame that none of it will ever see the light of day, some of the guys there wrote some lovely bits of code. I also recall quite a lot of code being rather less lovely, but very voluminous. I can't throw too much mud as I was still a fairly inex- perienced programmer myself back then and caused the odd faux pas, but certain internal coders had been writing reams of COBOL for decades and had clearly been quite hastily trained... I remember one of the senior developers getting distinctly unhappy with someone's dislike of subroutines and any other similar labour- and space-saving methodology! Chris. -- "If the world was an orange it would be like much too small, y'know?" Neil, '84 Currently playing: Sing Sing - "The Joy Of Sing Sing" (again) http://www.chrishedley.com My stuff, including genealogy, other things, etc ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 8 Mar 2003 17:05:14 -0000 Organization: Titanic Enterprises Unlimited Lines: 30 Message-ID: <1047143114.973510@saucer.planet.gong> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> <3E68E636.E638D524@ev1.net> <3E6A0951.10371A99@ev1.net> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1047143115 23612 80.177.7.220 (8 Mar 2003 17:05:15 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Sat, 8 Mar 2003 17:05:15 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Cache: nntpcache 3.0.1 (see http://www.nntpcache.org/) X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Cache-Post-Path: saucer.planet.gong!unknown@voodoo.planet.gong Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!nntp.theplanet.net!inewsm1.nntp.theplanet.net!195.40.4.120.MISMATCH!easynet-quince!easynet.net!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131368 "Charles Richmond" wrote in message news:3E6A0951.10371A99@ev1.net... > jmfbahciv@aol.com wrote: > > > > In article <3E68E636.E638D524@ev1.net>, > > Charles Richmond wrote: > > >jmfbahciv@aol.com wrote: > > >> > > >> [snip...] [snip...] [snip...] > > >> > > >> RIGHT!!! The fucking point of having computers was so you could > > >> start something up then go away and sleep or something. Having to > > >> stay and babysit an automatic process is just a side effect of that > > >> UDFH. > > >> > > > So you will like Unix/Linux/FreeBSD. They have commands that allow > > > you to set up your jobs to run "batch". > > > > I know. I did take Unix internals once. > > > Well, then, you will be right at home with Linux...it's > pretty much all the same. You might need 64 meg of memory There is one drawback with Linux. It's less appropriate for a.f.c. than a BSD. :) Cheers, Rupert ###### From: gorilla@elaine.furryape.com (Alan Barclay) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 8 Mar 2003 18:40:00 GMT Organization: Furryape Lines: 17 Message-ID: <1047148800.630756@elaine.furryape.com> References: <3E6A0951.10371A99@ev1.net> NNTP-Posting-Host: p-851.newsdawg.com X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: gorilla@elaine.furryape.com (Alan Barclay) Cache-Post-Path: elaine.furryape.com!unknown@elaine.furryape.com X-Cache: nntpcache 3.0.1 (see http://www.nntpcache.org/) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!ecngs!feeder.ecngs.de!zen.net.uk!diablo.theplanet.net!news.maxwell.syr.edu!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131510 In article , wrote: >In article <3E6A0951.10371A99@ev1.net>, > Charles Richmond wrote: >> ... A Pentium or >>Pentium II processor at least would be wonderful... If you >>do *not* have this system, go look at some of the local >>thrift stores for a used system. The Micky-Suck people are >>throwing away hardware of this description. > >They are? Already? Anything with less than 1Ghz processor is basically useless for running a 'modern' OS. Which is why when this system died a couple of months ago, I could get a new motherboard and CPU of an 'obselete' design for virtually nothing. ###### From: Ben Hutchings Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sun, 9 Mar 2003 00:33:51 +0000 Lines: 43 Message-ID: References: <3E6A0951.10371A99@ev1.net> <1047148800.630756@elaine.furryape.com> Reply-To: Ben Hutchings NNTP-Posting-Host: pc5-cmbg1-6-cust149.cmbg.cable.ntl.com (80.4.127.149) X-Trace: fu-berlin.de 1047170704 66550698 80.4.127.149 (16 [70929]) X-Orig-Path: 192.168.1.1!nobody User-Agent: slrn/0.9.7.4 (Linux) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!pc5-cmbg1-6-cust149.cmbg.cable.ntl.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131503 In article <1047148800.630756@elaine.furryape.com>, Alan Barclay wrote: > In article , wrote: >>In article <3E6A0951.10371A99@ev1.net>, >> Charles Richmond wrote: >>> ... A Pentium or >>>Pentium II processor at least would be wonderful... If you >>>do *not* have this system, go look at some of the local >>>thrift stores for a used system. The Micky-Suck people are >>>throwing away hardware of this description. >> >>They are? Already? > > Anything with less than 1Ghz processor is basically useless > for running a 'modern' OS. Actually, Microsoft seems to have stopped requiring much more processor power for each new Windows release, and the free Unixes certainly don't tend to do that. Windows 2000 is fine with a ~300 MHz processor and XP should not require much more than that, provided the new themable GUI is disabled. As for the free Unixes, Debian GNU/Linux 3.0 (woody) is fine on a 200 MHz PC I have here though some recent GUI apps (Mozilla and OpenOffice.org especially) are slow on that. Memory requirements seem to keep on rising though. I wouldn't want to use a recent Windows, or Linux with GNOME or KDE, with less than 128 MB. If you stick with the command-line Linux should be comfortable in 16-32 MB. With less than 16 MB you can probably run but not install most distributions (installation would have to be done with the disk in another machine). > Which is why when this system died a couple of months ago, I could > get a new motherboard and CPU of an 'obselete' design for virtually > nothing. Maybe I should do that to upgrade the above-mentioned PC a bit as my wife still uses that while I have a new one. She isn't complaining about it though... -- Ben Hutchings | personal web site: http://womble.decadentplace.org.uk/ Beware of bugs in the above code; I have only proved it correct, not tried it. - Donald Knuth ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Sun, 9 Mar 2003 11:31:15 -0700 Lines: 36 Message-ID: <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com> References: <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1047234676 67113883 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131569 Charles Shannon Hendrix wrote: > C++ has a lot of good points, but as an OO language, I think its contrived > and confusing. Objective C is much cleaner to me. On the other hand, C++ supports multi-paradigmatic programming because > At the same time, I think the idea that all projects are expressed best > in OO fashion is ridiculous. It's by far not a perfect language (especially the C heritage), but it's one of the lesser evil ones (at least you can shoot your own foot with it!) > I don't like what happens with C++ libraries. It seems that even small > programs end up large because they want to use a class or method which > implies a lot of baggage. It reminds me a bit of the first days of C on small computers: programmers left and right were bitching about the incredible size of 'Hello World' in C, when a 28 Byte assembler program could to the same. The large size of even small executables is annoying, granded, and my gut feeling is that in part it is caused by the C model of separating compilation and linking. > Java... it sucks on every platform except Sun, Windows, and Apple. > > Write once, run anywhere is mythology. Maybe not anywhere, but on enough platforms to make it useful :-) > The language itself looks interesting, It has it's good points; but for my taste it is still too close to C to be a real improvement. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sun, 09 Mar 03 11:36:23 GMT Organization: UltraNet Communications, Inc. Lines: 181 Message-ID: References: <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> X-Trace: UmFuZG9tSVZ4qRfhb6bYaAFjkH3LofJ+IzYdl4wGFkTkTw23avxQuhZ6VNe41R2d X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 9 Mar 2003 12:21:42 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-71 Xref: chonsp.franklin.ch alt.folklore.computers:131557 In article , Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: > >>>>>Some filesystems are almost that big, like SGI's xfs. >>>> >>>> You don't need the file system in core to have one. >>> >>>Yes you do. There is no way to do it otherwise. >>> >>>Filesystems in any non-trivial system are never idle. If you tried to >>>keep the code out of core, you'd never get any work done. >> >> I didn't say "code"; I said "file system". You don't >> need the data for handling a device until somebody pushes the on-line >> button. > >Sorry, I had thought it obvious I was talking about filesystem code, >not the filesystem itself. Aha! I misread your post. > >Filesystem data (metadata) is read as needed of course. Most of the >time you want some of it in core all the time for performance reasons. The only time is if the file contains some form of data base that requires inplace management of the bits. > >In fact, you'll usually have a lot of cached filesystem information. >If memory isn't in use by a program, there is no sense in letting it >sit idle when it can be used to speed I/O operations. Sure. But those I/O ops are the ones happening now. > >>>> I have yet to meet a CPU that can deal with more than one >>>> thing at a time (that's why we have multiple-CPU systems). >>> >>>????? >> >> CPUs are synchronous machines. The art of timesharing OSes >> is make them appear to be able to do everything at once. > >Actually, they aren't all like that. Some of them now have >hyperthreading: more than one instruction stream per CPU. This is getting into a grey area :-). I don't consider prefetching an execution. I just consider that a set up for the computation. Things have really been hazy with microcode involved. I won't talk about this one since FMPOV it gets into religion and semantics. > >The art of a lot of OSs is the same BTW. I had always assumed that this was true until certain releases :-(. > .. Basically any system which >allows the CPU to be interrupted is trying to make it as invisible >as possible. I'd use the word efficient. It's just struck me that "invisible" doesn't mean what we used to mean anymore. > >Timesharing just brought that out to interaction, some with more success >than others. > >> Scenario: /BAH starts up a task that requires the disk structure BARB >> to be spinning. So the pack is mounted, on-line procedures are >> completed, and /BAH goes back to office to massage a few bits. A >> priority interrupt is generated by a PHB or a bit god or a next door >> neighbor who just did an Ooops!. Until I do a DIRECT >> BARB:[10,2162]FOO.*, the data needed to access the bits doesn't have >> to be in core. This is a case where service on demand is a more >> efficient use of core. > >This is not true on a UNIX system, or just about any current system. > >BARB is likely an IDE or SCSI hard drive or removable media. The drivers >and filesystem code are already in memory because other drives and >filesystems, including those needed to boot the system, are already >in use. Yes. I'm not advocating a separate copy of code/job/device. > >Unless BARB happens to be on a drive on a unique type of controller >and filesystem, it's just not very bloody likely that this code has not >already been loaded. > >Perhaps an example or two will help: > >Now, on my system just now, I've not used my SCSI CD-ROM in days. >When I put in a CD and mounted it, several things happened in >the system: > >* the EATA SCSI driver is loaded (main controller is Adaptec) >* the bus is initialized >* the CD-ROM driver is loaded >* the iso9660 filesystem code is loaded >* the volume is mounted at /mnt/cdrom, which includes reading > and checking at least some filesystem metadata > >No further actions will cause any other drivers to be loaded. My system >is not set up to unload drivers not in use. Just my preference. >Note: I now have dual Adaptec SCSI controllers, so the drivers is >already in memory for both cards upon OS boot. OK. I think we may be talking past each again. There are days that I hate medium. > >Now another example. Let's say I mount a new SCSI hard drive. When >I mount this the system will: > >* mount the volume at /, which includes reading and > checking at least some filesystem metadata > >The controller drivers and filesystem code are already in core, because >this code is needed for other mounted systems. > >The first case I listed above is going to be VERY rare. It is? I would think that the last case (new device installed) would be the rare beast. > >> My point is that no OS will know when or how much data is going >> to be used. > >True, which is why there has been so much work on unified memory >management in UNIX over the last ten years. Unified memory management? I don't think I know what you mean by this. > >>>Most OS routines are going to be in use almost 24/7, so I don't see much >>>benefit in making them swappable. >> >> No, they're not. If you've got a compute-bound job that does >> no I/O, the device driver code will NEVER, I mean never, be >> executed. > >This compute bound job... how did it get loaded? With a RUN command 30 days ago. (I'm exaggerating slightly.) > >The system it is running on, how did it boot and start its other jobs >(surely it has other work to do) without loading device drivers, >filesystem code, etc? > >Now for older systems, it might be more common to have this situation, but >most modern systems are going to be pretty busy, and even compute-bound >jobs cause at least enough I/O to get loaded. Sure but they're compute-bound; that, by definition, implies that the only computing serice they require is CPU. They're not going to do I/O by design since that takes away heavy computation time. Just because they had some I/O at the beginning of run, it makes absolutely no sense to keep all of that code and/or data to start the program in core using up address space. >>>Now, your USB drivers... they might only get used once in awhile. >>>Lot's of things fall into that catagory. But the bulk of a UNIX OS and >>>its drivers are going to be in use all the time. There just won't be >>>a time of non-use long enough to justify moving it out. >> >> This tells me that you've never dealt with a real timesharing >> OS at its instruction level. > >Nope, I think something has confused you. It might be my fault, I >don't know. Both our faults :-). We're mixing new gear with old resource management techniques in the conversation. And it probably has to do with the invention of "smart" controllers. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sun, 09 Mar 03 11:42:24 GMT Organization: UltraNet Communications, Inc. Lines: 40 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> X-Trace: UmFuZG9tSVYSgJJWXX2sEbSEOQrs62oVU7t1g4XKj6TRAGW6lFfih9CShkyyW6el X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 9 Mar 2003 12:27:39 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!newsfeed.gol.com!feed1.news.rcn.net!rcn!208-59-181-71 Xref: chonsp.franklin.ch alt.folklore.computers:131559 In article , Charles Shannon Hendrix wrote: >In article <0c0d4b.tr1.ln@teabag.cbhnet>, Chris Hedley wrote: > >> I also recall quite a lot of code being rather less lovely, but very >> voluminous. > >I have written some code like that. It was mostly a matter of not having >the time to make it smaller. > >When my programs start getting big, I get worried. Problem is finding >someone else to see the same problem. My rule of thumb (and I didn't do fancy stuff at all) was, if the programming was getting unwieldy, then my data definitions were wrong. Most of the time, if my data was defined correctly, the program wrote itself. Caveat: I didn't do very much programming. > >Some people though, seem to tons of code and I don't know if it is >lack of training or some method of training they had. > >I often have read the code and cannot figure out what they are doing. >The logic starts getting heavily nested and obtuse, to the point where >I go find them for explanations. Somtimes they really don't know, >but assure me it's OK because the routines always return good results. > >Eeek! One of our comm guys told the story about going to a customer site, making[1] the code work (over a period of many days and nights with no sleep), getting back to the plant and then having to take 3 months to figure out why the code worked. [1]Making as in bit force trauma. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sun, 09 Mar 03 12:51:25 GMT Organization: UltraNet Communications, Inc. Lines: 72 Message-ID: References: X-Trace: UmFuZG9tSVZlgtSfktdnoXDm+cO3WiX2feAK4dgLMasmuoYTJjN1unxCsLIEh2MN X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 9 Mar 2003 13:36:40 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-71 Xref: chonsp.franklin.ch alt.folklore.computers:131537 In article , Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: > >>>What drives me nuts is that in some programs, I can specify a download >>>directory. The program puts up a dialog for every download. OK, I can >>>see that the user might want to rename the file. However, why not an >>>option to avoid this query for when you just want to queue up a bunch >>>of downloads? >> >> Why not have full wildcard support or indirect files. > >The protocols don't support wildcards, Huh? It doesn't have to be protocol, just a functional spec of the requesting software. I'm thinking of RDH's NFT program. Are we talking past each other again? I seem to be doing this a lot in the last couple of days; I did have to snowblow and remove growing webs with a vacuum that sucked. > ..but some programs can do >this by downloading ftp directories, and doing HTML mirroring. > >Indirect files... not sure what you mean here. Do you still not know? Somebody gave an explanation. I guess the -10 also called them CCL files. > >>>Being on a modem link, I'm a patient downloader, and don't sit there >>>drooling while the progress bar moves to 100%. I hate babysitting little >>>tasks like this. >> >> Especially when they lie. "operation will take ~5 minutes" and >> the error^Wprogress? bar's advancement predicts 5 hours. > >What kills me is having to have the stupid progress window in the first >place. It should be optional. Let me call it up IF I want to see it. Yea! You tell 'em. I want the damned CPU time to be spent doing the work not telling me that it's doing the work. WTH ever happened to T anyway? > >Another thing, why can't a browser schedule a download instead of >doing it immediately? > >This would be especially useful for getting that "final URL" for the >nasty redirects so common today. I don't understand this one. That's OK; you can tell I'm an old-fashioned gal ;-). > >>>Silly user, put down that mouse, and let us take care of everything... >>> >>>:) >> >> ROTFLMAO. Why bother flashing me? It's rather like the time >> my mother was cutting my hair when she said, "Ooops." > >Brought to you by the same people who put oil lights in cars which >come on after you run out of oil, and temperature warnings which >come on after you've seen the steam. Exactly. /BAH Subtract a hundred and four for e-mail. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 24 Date: Sat, 8 Mar 2003 23:55:44 -0500 NNTP-Posting-Host: 206.246.249.25 X-Trace: sydney.visi.net 1047187567 206.246.249.25 (Sun, 09 Mar 2003 00:26:07 EST) NNTP-Posting-Date: Sun, 09 Mar 2003 00:26:07 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131584 In article <0c0d4b.tr1.ln@teabag.cbhnet>, Chris Hedley wrote: > I also recall quite a lot of code being rather less lovely, but very > voluminous. I have written some code like that. It was mostly a matter of not having the time to make it smaller. When my programs start getting big, I get worried. Problem is finding someone else to see the same problem. Some people though, seem to tons of code and I don't know if it is lack of training or some method of training they had. I often have read the code and cannot figure out what they are doing. The logic starts getting heavily nested and obtuse, to the point where I go find them for explanations. Somtimes they really don't know, but assure me it's OK because the routines always return good results. Eeek! ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E68E636.E638D524@ev1.net> <3E6A0951.10371A99@ev1.net> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 24 Date: Sat, 8 Mar 2003 23:58:37 -0500 NNTP-Posting-Host: 206.246.249.25 X-Trace: sydney.visi.net 1047187569 206.246.249.25 (Sun, 09 Mar 2003 00:26:09 EST) NNTP-Posting-Date: Sun, 09 Mar 2003 00:26:09 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131587 In article , jmfbahciv@aol.com wrote: > All of that immediately pops out as tons. :-) I seem to > not have excised all subroutines of old iron calculations. If you want full sources, things can get pretty large. The systems aren't usually that big, but there is just so much software out there now, and driver code is huge due to the vast array of hardware out there. >> ... A Pentium or >>Pentium II processor at least would be wonderful... If you >>do *not* have this system, go look at some of the local >>thrift stores for a used system. The Micky-Suck people are >>throwing away hardware of this description. > > They are? Already? Well, if you'd seen how slow Windows XP was on anything <1GHz and 256MB of RAM, you'd know why. Oh, and this is before you start running programs... :) ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 42 Date: Sun, 9 Mar 2003 00:08:18 -0500 NNTP-Posting-Host: 206.246.249.181 X-Trace: sydney.visi.net 1047194790 206.246.249.181 (Sun, 09 Mar 2003 02:26:30 EST) NNTP-Posting-Date: Sun, 09 Mar 2003 02:26:30 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.stanford.edu!bloom-beacon.mit.edu!newsswitch.lcs.mit.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131574 In article , jmfbahciv@aol.com wrote: >>What drives me nuts is that in some programs, I can specify a download >>directory. The program puts up a dialog for every download. OK, I can >>see that the user might want to rename the file. However, why not an >>option to avoid this query for when you just want to queue up a bunch >>of downloads? > > Why not have full wildcard support or indirect files. The protocols don't support wildcards, but some programs can do this by downloading ftp directories, and doing HTML mirroring. Indirect files... not sure what you mean here. >>Being on a modem link, I'm a patient downloader, and don't sit there >>drooling while the progress bar moves to 100%. I hate babysitting little >>tasks like this. > > Especially when they lie. "operation will take ~5 minutes" and > the error^Wprogress? bar's advancement predicts 5 hours. What kills me is having to have the stupid progress window in the first place. It should be optional. Let me call it up IF I want to see it. Another thing, why can't a browser schedule a download instead of doing it immediately? This would be especially useful for getting that "final URL" for the nasty redirects so common today. >>Silly user, put down that mouse, and let us take care of everything... >> >>:) > > ROTFLMAO. Why bother flashing me? It's rather like the time > my mother was cutting my hair when she said, "Ooops." Brought to you by the same people who put oil lights in cars which come on after you run out of oil, and temperature warnings which come on after you've seen the steam. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 127 Date: Sun, 9 Mar 2003 00:52:06 -0500 NNTP-Posting-Host: 206.246.249.181 X-Trace: sydney.visi.net 1047194824 206.246.249.181 (Sun, 09 Mar 2003 02:27:04 EST) NNTP-Posting-Date: Sun, 09 Mar 2003 02:27:04 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!seven.news.surf.net!newsfeed3.funet.fi!newsfeed1.funet.fi!newsfeeds.funet.fi!enews.sgi.com!logbridge.uoregon.edu!feed2.news.rcn.net!rcn!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131579 In article , jmfbahciv@aol.com wrote: >>>>Some filesystems are almost that big, like SGI's xfs. >>> >>> You don't need the file system in core to have one. >> >>Yes you do. There is no way to do it otherwise. >> >>Filesystems in any non-trivial system are never idle. If you tried to >>keep the code out of core, you'd never get any work done. > > I didn't say "code"; I said "file system". You don't > need the data for handling a device until somebody pushes the on-line > button. Sorry, I had thought it obvious I was talking about filesystem code, not the filesystem itself. Filesystem data (metadata) is read as needed of course. Most of the time you want some of it in core all the time for performance reasons. In fact, you'll usually have a lot of cached filesystem information. If memory isn't in use by a program, there is no sense in letting it sit idle when it can be used to speed I/O operations. >>> I have yet to meet a CPU that can deal with more than one >>> thing at a time (that's why we have multiple-CPU systems). >> >>????? > > CPUs are synchronous machines. The art of timesharing OSes > is make them appear to be able to do everything at once. Actually, they aren't all like that. Some of them now have hyperthreading: more than one instruction stream per CPU. The art of a lot of OSs is the same BTW. Basically any system which allows the CPU to be interrupted is trying to make it as invisible as possible. Timesharing just brought that out to interaction, some with more success than others. > Scenario: /BAH starts up a task that requires the disk structure BARB > to be spinning. So the pack is mounted, on-line procedures are > completed, and /BAH goes back to office to massage a few bits. A > priority interrupt is generated by a PHB or a bit god or a next door > neighbor who just did an Ooops!. Until I do a DIRECT > BARB:[10,2162]FOO.*, the data needed to access the bits doesn't have > to be in core. This is a case where service on demand is a more > efficient use of core. This is not true on a UNIX system, or just about any current system. BARB is likely an IDE or SCSI hard drive or removable media. The drivers and filesystem code are already in memory because other drives and filesystems, including those needed to boot the system, are already in use. Unless BARB happens to be on a drive on a unique type of controller and filesystem, it's just not very bloody likely that this code has not already been loaded. Perhaps an example or two will help: Now, on my system just now, I've not used my SCSI CD-ROM in days. When I put in a CD and mounted it, several things happened in the system: * the EATA SCSI driver is loaded (main controller is Adaptec) * the bus is initialized * the CD-ROM driver is loaded * the iso9660 filesystem code is loaded * the volume is mounted at /mnt/cdrom, which includes reading and checking at least some filesystem metadata No further actions will cause any other drivers to be loaded. My system is not set up to unload drivers not in use. Just my preference. Note: I now have dual Adaptec SCSI controllers, so the drivers is already in memory for both cards upon OS boot. Now another example. Let's say I mount a new SCSI hard drive. When I mount this the system will: * mount the volume at /, which includes reading and checking at least some filesystem metadata The controller drivers and filesystem code are already in core, because this code is needed for other mounted systems. The first case I listed above is going to be VERY rare. > My point is that no OS will know when or how much data is going > to be used. True, which is why there has been so much work on unified memory management in UNIX over the last ten years. >>Most OS routines are going to be in use almost 24/7, so I don't see much >>benefit in making them swappable. > > No, they're not. If you've got a compute-bound job that does > no I/O, the device driver code will NEVER, I mean never, be > executed. This compute bound job... how did it get loaded? The system it is running on, how did it boot and start its other jobs (surely it has other work to do) without loading device drivers, filesystem code, etc? Now for older systems, it might be more common to have this situation, but most modern systems are going to be pretty busy, and even compute-bound jobs cause at least enough I/O to get loaded. >>Now, your USB drivers... they might only get used once in awhile. >>Lot's of things fall into that catagory. But the bulk of a UNIX OS and >>its drivers are going to be in use all the time. There just won't be >>a time of non-use long enough to justify moving it out. > > This tells me that you've never dealt with a real timesharing > OS at its instruction level. Nope, I think something has confused you. It might be my fault, I don't know. ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 30 Date: Sun, 09 Mar 2003 09:05:18 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news2.calgary.shaw.ca 1047200718 24.71.223.147 (Sun, 09 Mar 2003 02:05:18 MST) NNTP-Posting-Date: Sun, 09 Mar 2003 02:05:18 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.icl.net!newsfeed.fjserv.net!newsfeed.icl.net!newsfeed.fjserv.net!out.nntp.be!propagator2-SanJose!in.nntp.be!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news2.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131568 On Sun, 9 Mar 2003 00:08:18 -0500 in alt.folklore.computers, Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: > >>>What drives me nuts is that in some programs, I can specify a download >>>directory. The program puts up a dialog for every download. OK, I can >>>see that the user might want to rename the file. However, why not an >>>option to avoid this query for when you just want to queue up a bunch >>>of downloads? >> >> Why not have full wildcard support or indirect files. > >The protocols don't support wildcards, but some programs can do >this by downloading ftp directories, and doing HTML mirroring. > >Indirect files... not sure what you mean here. response files; files containing lists of filenames, and sometimes other parts of commands: used by command shells without command redirection or restrictive command length limits. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### From: Andreas Eder Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 09 Mar 2003 11:48:48 +0100 Organization: T-Online Lines: 19 Sender: are@elgin.eder.de Message-ID: References: Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Trace: news.t-online.com 1047207603 03 26374 nOdoEGcGSbAZxN 030309 11:00:03 X-Complaints-To: abuse@t-online.com X-ID: ZV92h4Zv8eNJHpBoGHO3F8i0SHA+DLgy9oQzBdgSluAdYCaQ6Fcukr User-Agent: Gnus/5.0808 (Gnus v5.8.8) XEmacs/21.4 (Portable Code) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!newsfeed.arcor-online.net!newsfeed01.sul.t-online.de!newsmm00.sul.t-online.com!t-online.de!news.t-online.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131555 Charles Shannon Hendrix writes: > Another thing, why can't a browser schedule a download instead of > doing it immediately? > > This would be especially useful for getting that "final URL" for the > nasty redirects so common today. You could use wwwoffle. The wwwoffled program is a simple proxy server with special features for use with dial-up internet links. This means that it is possible to browse web pages and read them without having to remain connected. Together with a cron script you could make scheduled downloads. URI:http://www.gedanken.demon.co.uk/wwwoffle/ 'Andreas -- Wherever I lay my .emacs, there´s my $HOME. ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Sun, 9 Mar 2003 13:18:54 -0700 Lines: 13 Message-ID: <1frk0xk.1agc9u4198go3kN%lars@bearnip.com> References: <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <3E60AEEB.4FBB6FD2@yahoo.com> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1047241134 65945348 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131596 wrote: > My rule of thumb (and I didn't do fancy stuff at all) was, if the > programming was getting unwieldy, then my data definitions were > wrong. ... or the problem definition. > Most of the time, if my data was defined correctly, the > program wrote itself. Caveat: I didn't do very much programming. Sometimes code simply is unwieldy no matter what, but it is a good rule of thumb. ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 65 Date: Sun, 09 Mar 2003 22:04:03 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news2.calgary.shaw.ca 1047247443 24.71.223.147 (Sun, 09 Mar 2003 15:04:03 MST) NNTP-Posting-Date: Sun, 09 Mar 2003 15:04:03 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.telusplanet.net!ps01-sjc1.MISMATCH!sjc70.webusenet.com!news.webusenet.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news2.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131597 On Sun, 09 Mar 03 11:36:23 GMT in alt.folklore.computers, jmfbahciv@aol.com wrote: >In article , > Charles Shannon Hendrix wrote: >>In article , jmfbahciv@aol.com wrote: >>Perhaps an example or two will help: >> >>Now, on my system just now, I've not used my SCSI CD-ROM in days. >>When I put in a CD and mounted it, several things happened in >>the system: >> >>* the EATA SCSI driver is loaded (main controller is Adaptec) >>* the bus is initialized >>* the CD-ROM driver is loaded >>* the iso9660 filesystem code is loaded >>* the volume is mounted at /mnt/cdrom, which includes reading >> and checking at least some filesystem metadata >> >>No further actions will cause any other drivers to be loaded. My system >>is not set up to unload drivers not in use. Just my preference. >>Note: I now have dual Adaptec SCSI controllers, so the drivers is >>already in memory for both cards upon OS boot. >>Now another example. Let's say I mount a new SCSI hard drive. When >>I mount this the system will: >> >>* mount the volume at /, which includes reading and >> checking at least some filesystem metadata >> >>The controller drivers and filesystem code are already in core, because >>this code is needed for other mounted systems. >> >>The first case I listed above is going to be VERY rare. > >It is? I would think that the last case (new device installed) >would be the rare beast. Device online but not mounted is pretty rare these days, but the first example requires different drivers from the disks already mounted. >>> My point is that no OS will know when or how much data is going >>> to be used. >> >>True, which is why there has been so much work on unified memory >>management in UNIX over the last ten years. > >Unified memory management? I don't think I know what you mean >by this. Treating all disk (if not other) I/O as paging I/O, including loading programs, reading and writing user data, and managing all of memory as a single page pool, unifying the process space and filesystem buffer cache management. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### From: pechter@shell.monmouth.com (Bill/Carolyn Pechter) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 9 Mar 2003 17:10:53 -0500 Organization: Lakewood MicroSystems Lines: 24 Message-ID: References: NNTP-Posting-Host: shell.monmouth.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!nntp-out.monmouth.com!newspeer.monmouth.com!news.monmouth.com!shell.monmouth.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131593 In article , wrote: > >Yea! You tell 'em. I want the damned CPU time to be spent doing >the work not telling me that it's doing the work. WTH ever happened >to T anyway? > > >/BAH Control-T, eh... $ ssh i4got #136 i4got.pechter.dyndns.org:/home/pechter> ^T load: 0.09 cmd: ksh 11019 [ttyin] 0.01u 0.05s 0% 660k Works on FreeBSD... Bill -- +---------------------------------------------------------------------------+ | Bill and/or Carolyn Pechter | pechter@shell.monmouth.com | | Bill Gates is a Persian cat and a monocle away from being a villain in | | a James Bond movie -- Dennis Miller | +---------------------------------------------------------------------------+ ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Mon, 10 Mar 03 14:39:52 GMT Organization: UltraNet Communications, Inc. Lines: 34 Message-ID: References: <3E60AEEB.4FBB6FD2@yahoo.com> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> <1frk0xk.1agc9u4198go3kN%lars@bearnip.com> X-Trace: UmFuZG9tSVayP09wBc83TCCLn49GiW5eMJzlW0PMt46XH8RcdRXuocvvxga1VM6s X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 10 Mar 2003 15:25:18 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!c03.atl99!news.webusenet.com!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-182-226 Xref: chonsp.franklin.ch alt.folklore.computers:131621 In article <1frk0xk.1agc9u4198go3kN%lars@bearnip.com>, lars@bearnip.com (Lars Duening) wrote: > wrote: > >> My rule of thumb (and I didn't do fancy stuff at all) was, if the >> programming was getting unwieldy, then my data definitions were >> wrong. > >.... or the problem definition. Yup. IME, I always had the problem definitions down pat since I took years to study the problem. > >> Most of the time, if my data was defined correctly, the >> program wrote itself. Caveat: I didn't do very much programming. > >Sometimes code simply is unwieldy no matter what, but it is a good rule >of thumb. I really did mean rule of thumb. :-) I was one of the fortunate? ones to be handed FILCOM maintenance. The rule of thumb for that piece of code was to leave well enough alone even if there was a bug. There still exists a known bug. It was to fix a bug in a never-used option (the reporter was just playing around and noticed the bug). Unfortunately, the fix broke the code that was used all the time. I didn't have any problems pulling the fix with a comment line that said, "Known bug; don't do that." /BAH Subtract a hundred and four for e-mail. ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Sun, 9 Mar 2003 18:01:21 -0700 Lines: 12 Message-ID: <1frk7ub.35pzcy1i9nl34N%lars@bearnip.com> References: <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com> <3E6B8CA5.2050601@jetnet.ab.ca> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1047258080 67305356 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131668 wrote: > Lars Duening wrote: > > > It's by far not a perfect language (especially the C heritage), but it's > > one of the lesser evil ones (at least you can shoot your own foot with > > it!) > > Read this to find out why you can't. > http://burks.brighton.ac.uk/burks/language/shoot.htm So C++ even supports Symmetric Multi Foot Shooting! Excellent! ###### Message-ID: <3E6CF378.24FC015D@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <3E60AEEB.4FBB6FD2@yahoo.com> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> <1frk0xk.1agc9u4198go3kN%lars@bearnip.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 26 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc51.ops.asp.att.net 1047320515 12.241.15.59 (Mon, 10 Mar 2003 18:21:55 GMT) NNTP-Posting-Date: Mon, 10 Mar 2003 18:21:55 GMT Date: Mon, 10 Mar 2003 18:21:55 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-han1.dfn.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!wn14feed!wn13feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!rwcrnsc51.ops.asp.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131677 jmfbahciv@aol.com wrote: > > [snip...] [snip...] [snip...] > > I really did mean rule of thumb. :-) I was one of the > fortunate? ones to be handed FILCOM maintenance. The rule of > thumb for that piece of code was to leave well enough alone > even if there was a bug. There still exists a known bug. It > was to fix a bug in a never-used option (the reporter was just > playing around and noticed the bug). Unfortunately, the fix > broke the code that was used all the time. I didn't have any > problems pulling the fix with a comment line that said, "Known > bug; don't do that." > IOW, you "fixed it in the manual..." If you can't make the code do a certain thing, then put in the manual that the code is *not* supposed to do it. Could you have changed the code to FILCOM so that if someone tried to use this function, it would do *nothing*??? Of course, you would still have to "fix it in the manual". -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Mon, 10 Mar 03 11:46:30 GMT Organization: UltraNet Communications, Inc. Lines: 58 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> X-Trace: UmFuZG9tSVavmVUXAghe0bSQvX1Sjpel2gDWVxrQJQX6wjyPjy8kRGSb/gJ4kSsY X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 10 Mar 2003 12:31:58 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-145 Xref: chonsp.franklin.ch alt.folklore.computers:131630 In article , Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: > >> My rule of thumb (and I didn't do fancy stuff at all) was, if the >> programming was getting unwieldy, then my data definitions were >> wrong. Most of the time, if my data was defined correctly, the >> program wrote itself. Caveat: I didn't do very much programming. > >It's a good rule of thumb. > >Another thing this often means is that you've not normalized/modularized >your code enough. If you reading dozens or more levels of nested >code and having to remember too much state information, you probably >have code complexity issues. I never had to remember states; that's what the computer did very well. > >Of course, there are times when the problem is simply too big to express >simply. Nothing to do but press on. > >> One of our comm guys told the story about going to >> a customer site, making[1] the code work (over a period of many >> days and nights with no sleep), getting back to the plant >> and then having to take 3 months to figure out why the code >> worked. > >Something you don't want to here from staff working out a bug in >production code: > > "Look, just run the damn thing before I forget how I'm holding > my mouth!" That's exactly what the customer was told; his trip to the customer site was a firefight. In firefights all production rules are suspended because the only goal is to get the customer back up and running. The reason he spent three months decrypting the code was to be able to get it into production so that this customer wouldn't have a crisis again. That's how the biz is done.. or was done in our niche. In the real olden days, before my time, each computer installation was hand assembled and installed in Maynard, including the software. Then it was desembled very carefully, packaged up, shipped (the humans were including in the packaging), and reassembled on site. Eventually, DEC figured out how to do big machines using the production line model. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Mon, 10 Mar 03 11:48:26 GMT Organization: UltraNet Communications, Inc. Lines: 27 Message-ID: References: X-Trace: UmFuZG9tSVb9DJQjoeiADwxebyEnRJ5vUQ7Lha92FyBS1CkyHIy8KBZWgjHonICU X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 10 Mar 2003 12:33:53 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!cyclone.bc.net!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-145 Xref: chonsp.franklin.ch alt.folklore.computers:131627 In article , pechter@shell.monmouth.com (Bill/Carolyn Pechter) wrote: >In article , wrote: >> >>Yea! You tell 'em. I want the damned CPU time to be spent doing >>the work not telling me that it's doing the work. WTH ever happened >>to T anyway? >> >> >>/BAH > >Control-T, eh... >$ ssh i4got > >#136 i4got.pechter.dyndns.org:/home/pechter> ^T >load: 0.09 cmd: ksh 11019 [ttyin] 0.01u 0.05s 0% 660k > >Works on FreeBSD... That command truly created superstition. There were days that it sure felt like a ^T was necessary in order for the machine to pay you some attention. Even those who knew the code intimately would ^T to get attention. /BAH Subtract a hundred and four for e-mail. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 29 Date: Sun, 9 Mar 2003 22:35:28 -0500 NNTP-Posting-Host: 209.96.185.190 X-Trace: sydney.visi.net 1047270360 209.96.185.190 (Sun, 09 Mar 2003 23:26:00 EST) NNTP-Posting-Date: Sun, 09 Mar 2003 23:26:00 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!nntp.abs.net!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131684 In article , jmfbahciv@aol.com wrote: > My rule of thumb (and I didn't do fancy stuff at all) was, if the > programming was getting unwieldy, then my data definitions were > wrong. Most of the time, if my data was defined correctly, the > program wrote itself. Caveat: I didn't do very much programming. It's a good rule of thumb. Another thing this often means is that you've not normalized/modularized your code enough. If you reading dozens or more levels of nested code and having to remember too much state information, you probably have code complexity issues. Of course, there are times when the problem is simply too big to express simply. Nothing to do but press on. > One of our comm guys told the story about going to > a customer site, making[1] the code work (over a period of many > days and nights with no sleep), getting back to the plant > and then having to take 3 months to figure out why the code > worked. Something you don't want to here from staff working out a bug in production code: "Look, just run the damn thing before I forget how I'm holding my mouth!" ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 19 Date: Sun, 9 Mar 2003 22:36:51 -0500 NNTP-Posting-Host: 209.96.185.190 X-Trace: sydney.visi.net 1047270361 209.96.185.190 (Sun, 09 Mar 2003 23:26:01 EST) NNTP-Posting-Date: Sun, 09 Mar 2003 23:26:01 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131688 In article , Andreas Eder wrote: > Charles Shannon Hendrix writes: > >> Another thing, why can't a browser schedule a download instead of >> doing it immediately? >> >> This would be especially useful for getting that "final URL" for the >> nasty redirects so common today. > > You could use wwwoffle. Hmmm... didn't see a way for downloads to be automated. Is there some way to recognize file extensions and intercept the downloads? I use squid as a proxy right now, and proxy-off all downloads. I use scripts which call wget for most scheduled downloads. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 116 Date: Sun, 9 Mar 2003 23:07:16 -0500 NNTP-Posting-Host: 209.96.185.190 X-Trace: sydney.visi.net 1047273958 209.96.185.190 (Mon, 10 Mar 2003 00:25:58 EST) NNTP-Posting-Date: Mon, 10 Mar 2003 00:25:58 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131694 In article , jmfbahciv@aol.com wrote: >>Filesystem data (metadata) is read as needed of course. Most of the >>time you want some of it in core all the time for performance reasons. > > The only time is if the file contains some form of data base > that requires inplace management of the bits. No, you have to have metadata in memory most of the time, at least a minimal set. It's almost never going to be a win to flush it all. >>> CPUs are synchronous machines. The art of timesharing OSes >>> is make them appear to be able to do everything at once. >> >>Actually, they aren't all like that. Some of them now have >>hyperthreading: more than one instruction stream per CPU. > > This is getting into a grey area :-). I don't consider prefetching > an execution. I just consider that a set up for the computation. > Things have really been hazy with microcode involved. I won't > talk about this one since FMPOV it gets into religion and semantics. I don't mean prefetching, I mean literally multiple instruction streams. They show up under Linux as multiple processors. The reality is that benchmarking shows only moderate increases in performance. >>No further actions will cause any other drivers to be loaded. My system >>is not set up to unload drivers not in use. Just my preference. >>Note: I now have dual Adaptec SCSI controllers, so the drivers is >>already in memory for both cards upon OS boot. > > OK. I think we may be talking past each again. There are days > that I hate medium. It also takes a lot of time to do this stuff. I'm frequently read and replying with books and web browsers open for research. Only can afford the time once in awhile. I really don't see how some people post like they do, daily, on many different groups. >>The controller drivers and filesystem code are already in core, because >>this code is needed for other mounted systems. >> >>The first case I listed above is going to be VERY rare. > > It is? I would think that the last case (new device installed) > would be the rare beast. Well, the first case was on my secondary drive controller. When I first boot up, nothing in the system needs the devices on that SCSI chain. However, if I were to add a new SCSI drive (let's say I managed to save enough $$$ for a hot-swap bay) I'd likely put that on the main controller, and so it would not have to load any drivers at all, unless that drive needed a filesystem module that had not yet been used and/or loaded. >>> My point is that no OS will know when or how much data is going >>> to be used. >> >>True, which is why there has been so much work on unified memory >>management in UNIX over the last ten years. > > Unified memory management? I don't think I know what you mean > by this. Older UNIX kernels managed various areas of memory seperated. Most now merge at least some of these and managed them as one pool of memory. The idea is to gain speed and also use more of available free memory so it doesn't go wasted. Of course, the above paragraph is like saying anthrax makes you sick, so Google for white papers and stuff like that. Here's a link to information about NetBSD's UVM code: http://www.netbsd.org/Documentation/kernel/uvm.html Not all systems have exactly the same ideas and implementations of course. There may be better generic resources than this one too. >>This compute bound job... how did it get loaded? > > With a RUN command 30 days ago. (I'm exaggerating slightly.) It had to come off of a device somewhere... > Sure but they're compute-bound; that, by definition, implies that > the only computing serice they require is CPU. They're not going > to do I/O by design since that takes away heavy computation time. Yes, but what about the other programs which are running? They might be doing I/O. Also, the system will be writing to logs, updating accounting information, updating filesystem metadata. There is a lot going on besides this one compute-bound job. Also, its taking up pages and there is the possibility it might get some of its unused pages swapped out, and this requires I/O drivers. > Just because they had some I/O at the beginning of run, it makes > absolutely no sense to keep all of that code and/or data to > start the program in core using up address space. If you have a single-tasking batch system, sure. But I can't imagine a UNIX system that would run one compute-bound job and then do no further I/O. An embedded system might, but in that case you may as well load from a ROM and avoid the drive altogether. ###### From: "Peter Ibbotson" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Mon, 10 Mar 2003 10:14:24 -0000 Lines: 54 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> NNTP-Posting-Host: mailgate.lakeview.co.uk X-Trace: news.demon.co.uk 1047291264 14762 62.49.243.90 (10 Mar 2003 10:14:24 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Mon, 10 Mar 2003 10:14:24 +0000 (UTC) X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!newspeer.lavaseals.co.uk!diablo.theplanet.net!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131659 "Charles Shannon Hendrix" wrote in message news:slrnb6i0se.12h.shannon@news.widomaker.com... > In article , Peter Ibbotson wrote: > > > I suspect thats because: > > a) It's quick to write. > > It's no quicker than a defined routine for doing it another way... :) True, but it's always present and for quick 'n' dirty purposes it is easy. > > > b) For Windows systems it stops the message queue and returns back to the > > same place in the code it was called from. Most of the other solutions > > require paint messages to be processed, which often also allows menu > > messages to be processed which can cause reenterancy problems. > > Hmmm... post a message with a routine that puts the message in the UI > without a dialog. The routine also sets a time to live so the > message goes away. > > Since there are good programs which do this, it must be possible in > Windows. In fact, I'm not a Windows programmer, but I thought there > was a message window class which would do exactly this. It's possible, however if you're writing in VB it's relatively tricky unless you're displaying a modal form to show all error messages. > I hate programs which notify me of an error, and then make me wait > several seconds before I can move on. Ah, but you'll look at an error report window, most of my users wouldn't (They'd take the "no message box, so it's all alright" view of the world) Certainly if you're programming in VB (Also applies to MFC but then you could write a generic message handler) doing what you're suggesting is quite tough for simple routines. The problem is that unless you're already showing a modal form it's fairly hard to stop the message processing for your main application which will eventually bring reenterancy problems. To get the screen to repaint you have to allow the message queue to be examined so it's back around to the re-entrancy problems. For long running bits of code, displaying a progress form with a cancel button means that adding an error list is trivial, since you'll already be calling the message queue examination code to make the cancel button work. It's not that it's difficult to implement this stuff, just painful for novices. -- Work peteri@lakeview.co.uk.plugh.org | remove magic word .org to reply Home peter@ibbotson.co.uk.plugh.org | I own the domain but theres no MX ###### Date: Mon, 10 Mar 2003 20:05:41 +0100 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <20030310200541.12ec6fa1.steveo@eircom.net> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 16 Organization: EuroNet Internet NNTP-Posting-Date: 10 Mar 2003 20:18:27 GMT NNTP-Posting-Host: i0798.vwr.wanadoo.nl X-Trace: 1047327507 maya.euronet.nl 23899 194.134.211.34:1383 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!solnet.ch!solnet.ch!newsfeed.freenet.de!newsfeed00.sul.t-online.de!newsfeed01.sul.t-online.de!t-online.de!newspeer1-gui.server.ntli.net!ntli.net!news2.euro.net!postnews1.euro.net!maya.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131702 On Sat, 8 Mar 2003 14:57:04 +0000 cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) wrote: CH> but CH> certain internal coders had been writing reams of COBOL for decades CH> and had clearly been quite hastily trained... I can still remember the brain-fried expressions on their faces as the pace of courses picked up (C, SQL, UNIX, OO design ...). They went from keen to frazzled remarkably quickly :) -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Mon, 10 Mar 2003 21:58:42 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 19 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030304211555.4731fa87.steveo@eircom.net> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1047333522 9804 146.186.61.46 (10 Mar 2003 21:58:42 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Mon, 10 Mar 2003 21:58:42 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131697 alt.sys.pdp10:12730 In article , wrote: >In article <20030304211555.4731fa87.steveo@eircom.net>, > Steve O'Hara-Smith wrote: >Two possibilities come to mind: >1. It was our first off-by-one bug. >2. It was the first lesson in converting customers to octal. Tough company--don't pay your bills, and get converted to a shipping product . . . "No, that's my nose, not a power co--" hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Tue, 11 Mar 03 13:09:37 GMT Organization: UltraNet Communications, Inc. Lines: 59 Message-ID: References: <3E60AEEB.4FBB6FD2@yahoo.com> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> <1frk0xk.1agc9u4198go3kN%lars@bearnip.com> <3E6CF378.24FC015D@ev1.net> X-Trace: UmFuZG9tSVaasGA32HSJgFPlDWqKHsX+7QDHhp/pKyxwQjLGjHj9SHCpiQie0TNm X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 11 Mar 2003 13:55:14 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!headwall.stanford.edu!newsfeed.stanford.edu!logbridge.uoregon.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-30 Xref: chonsp.franklin.ch alt.folklore.computers:131763 In article <3E6CF378.24FC015D@ev1.net>, Charles Richmond wrote: >jmfbahciv@aol.com wrote: >> >> [snip...] [snip...] [snip...] >> >> I really did mean rule of thumb. :-) I was one of the >> fortunate? ones to be handed FILCOM maintenance. The rule of >> thumb for that piece of code was to leave well enough alone >> even if there was a bug. There still exists a known bug. It >> was to fix a bug in a never-used option (the reporter was just >> playing around and noticed the bug). Unfortunately, the fix >> broke the code that was used all the time. I didn't have any >> problems pulling the fix with a comment line that said, "Known >> bug; don't do that." >> >IOW, you "fixed it in the manual..." I wasn't allowed to put it in the manual. I wasn't even allowed to officially announce don't do that. But I did it anyway. > ... If you can't make the code >do a certain thing, then put in the manual that the code is >*not* supposed to do it. Ah, but that was putting a restriction on a published piece of functionality; the fact that no idiot would want to do this particular thing with the option was taboo to code. > >Could you have changed the code to FILCOM so that if someone >tried to use this function, it would do *nothing*??? No. I wouldn't do that. > .. Of >course, you would still have to "fix it in the manual". Not really. However, in our shop, if it changed the manual, all kinds of people would have to "approve" putting it into field test. One of the ways I snuck code out (in other words, included all maintenance fixes on each distribution tape) was to make sure that docs were not impacted. That usually got the product manager's attention and in those days, her goal was product prevention. Anyway, I consider duplicating the code into another routine, adding a test for the rare option and using the clone routine with the fix applied to do the comparing. When I looked at the code, I got afraid to do even that. FILCOM was not straight-forward for naif programmer. That's one of the reasons people didn't want to own it. /BAH Subtract a hundred and four for e-mail. ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Mon, 10 Mar 2003 20:05:06 -0700 Lines: 17 Message-ID: <1frlud0.141htge15djdywN%lars@bearnip.com> References: <50AE869F34558C8B.2523E7CF0B289EF1.50F8E1A90B3FCFEA@lp.airnews.net> <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1047351908 67782872 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131796 Charles Shannon Hendrix wrote: > Just for example, C++ and Java often resulted in less re-use and much > more complicated software than the languages and projects they replaced. Not exactly connected to re-use, but I am currently lurking on a different newsgroups where some actually question the need for hashing when doing data lookups (by a string key, for example) - on the grounds that modern processors are so much faster today. In their defense: they acknowledge that in large systems hashing still has its uses, but for small or medium sized system they don't deem it useful anymore. Maybe processors today are really that much faster, but chosing an O(N) or worse approach over an O(1) just seems ... wrong. Especially if the programs are meant to serve an unknown number of concurrent users. ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <3E5A710D.6030403@jetnet.ab.ca> <1frlud0.141htge15djdywN%lars@bearnip.com> X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 48 Date: Tue, 11 Mar 2003 06:08:27 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news2.calgary.shaw.ca 1047362907 24.71.223.147 (Mon, 10 Mar 2003 23:08:27 MST) NNTP-Posting-Date: Mon, 10 Mar 2003 23:08:27 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!nntp-relay.ihug.net!ihug.co.nz!west.cox.net!cox.net!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news2.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131806 On Mon, 10 Mar 2003 20:05:06 -0700 in alt.folklore.computers, lars@bearnip.com (Lars Duening) wrote: >Charles Shannon Hendrix wrote: > >> Just for example, C++ and Java often resulted in less re-use and much >> more complicated software than the languages and projects they replaced. > >Not exactly connected to re-use, but I am currently lurking on a >different newsgroups where some actually question the need for hashing >when doing data lookups (by a string key, for example) - on the grounds >that modern processors are so much faster today. > >In their defense: they acknowledge that in large systems hashing still >has its uses, but for small or medium sized system they don't deem it >useful anymore. > >Maybe processors today are really that much faster, but chosing an O(N) >or worse approach over an O(1) just seems ... wrong. Especially if the >programs are meant to serve an unknown number of concurrent users. Measurement time! -- too difficult to estimate performance on cached multiple-dispatch machines, other than to say that a really tight brute force machine loop, with code and data in first level cache, and no stride issues, is often faster than a more sophisticated approach up to some number of items N, where N is getting larger every year. If the actual machine can vary from small to large cache, e.g. Celeron to Xeon, and cycle speed, e.g. .2 to 2GHz, YMMV substantially. For instance, I don't see the on chip cache size going up linearly with cycle speed, e.g. AFAIK 2GHz Xeon does not have 10x cache of 200MHz Pentium Pro, so cache hit rates must be getting lower, and fast Celerons with small caches must be doing a lot of cache thrashing IMO. Things were easier to control when you had memory to interleave, and thrashing pages to worry about: now you've another two or three levels of side effects you can't do much about, once you've bought a machine. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 11 Mar 03 13:11:48 GMT Organization: UltraNet Communications, Inc. Lines: 24 Message-ID: References: X-Trace: UmFuZG9tSVaNR7YM7tAfxipipDzXL4GJgy7U7A30tl6aVoc15eJE+Rsdtl2mFJcZ X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 11 Mar 2003 13:57:25 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!headwall.stanford.edu!newsfeed.stanford.edu!logbridge.uoregon.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-30 Xref: chonsp.franklin.ch alt.folklore.computers:131742 In article , Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: > >> That command truly created superstition. There were >> days that it sure felt like a ^T was necessary in order for >> the machine to pay you some attention. Even those who knew >> the code intimately would ^T to get attention. > >Not much different than: > >% sync >% sync >% sync > >....on a UNIX machine, just in case that first sync missed something. It would. Think about getting a car into first gear while still moving. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 11 Mar 03 13:15:11 GMT Organization: UltraNet Communications, Inc. Lines: 19 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> <20030310200541.12ec6fa1.steveo@eircom.net> X-Trace: UmFuZG9tSVa02YbikcKke8QzruEHYAR6ZTbZd7761lgQdEa7udD50sQVsDfhu67j X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 11 Mar 2003 14:00:48 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!msc1.onvoy!onvoy.com!arclight.uoregon.edu!logbridge.uoregon.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-30 Xref: chonsp.franklin.ch alt.folklore.computers:131757 In article <20030310200541.12ec6fa1.steveo@eircom.net>, Steve O'Hara-Smith wrote: >On Sat, 8 Mar 2003 14:57:04 +0000 >cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) wrote: > >CH> but >CH> certain internal coders had been writing reams of COBOL for decades >CH> and had clearly been quite hastily trained... > > I can still remember the brain-fried expressions on their faces >as the pace of courses picked up (C, SQL, UNIX, OO design ...). They >went from keen to frazzled remarkably quickly :) > I managed to escape. Doing machine language is a great HLL avoidance coping strategy. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 11 Mar 03 13:15:48 GMT Organization: UltraNet Communications, Inc. Lines: 22 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> X-Trace: UmFuZG9tSVbNkHy+DHyKZ/uA143wKOL/yl+B6PrQn4EREJS6+bp5HnzNtXBd+bMk X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 11 Mar 2003 14:01:25 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.telusplanet.net!newsfeed.bc.tac.net!cyclone.bc.net!logbridge.uoregon.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-30 Xref: chonsp.franklin.ch alt.folklore.computers:131764 In article , Howard S Shubs wrote: >In article , jmfbahciv@aol.com wrote: > >> The reason he spent three months decrypting the code was to >> be able to get it into production so that this customer >> wouldn't have a crisis again. That's how the biz is done.. >> or was done in our niche. In the real olden days, before >> my time, each computer installation was hand assembled >> and installed in Maynard, including the software. Then it >> was desembled very carefully, packaged up, shipped (the >> humans were including in the packaging), and reassembled >> on site. > >I -do- seem to recall seeing "field engineer" on the manifest list for a >VAX 9000. > ROTFL. I hope they weren't shrink-wrapped. /BAH Subtract a hundred and four for e-mail. ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 11 Mar 03 10:14:28 -0800 Organization: http://extra.newsguy.com Lines: 35 Message-ID: <1400.200T1419T6144712@kltpzyxm.invalid> References: <20030310200541.12ec6fa1.steveo@eircom.net> NNTP-Posting-Host: p-751.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews2 Xref: chonsp.franklin.ch alt.folklore.computers:131776 In article jmfbahciv@aol.com (jmfbahciv) writes: >In article <20030310200541.12ec6fa1.steveo@eircom.net>, >Steve O'Hara-Smith wrote: > >>On Sat, 8 Mar 2003 14:57:04 +0000 >>cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) wrote: >> >>> but certain internal coders had been writing reams of COBOL for >>> decades and had clearly been quite hastily trained... >> >> I can still remember the brain-fried expressions on their faces >>as the pace of courses picked up (C, SQL, UNIX, OO design ...). They >>went from keen to frazzled remarkably quickly :) > >I managed to escape. Doing machine language is a great >HLL avoidance coping strategy. I learned that art in my university days, eventually dropping out for a job in a small service bureau. They _really_ loved HLLs at university - the more obscure the better. (One course hit you with a new HLL every two weeks.) Meanwhile, I had quickly taken to assembly language, having discovered the joys of not having a compiler getting in the way. The only thing that could wean me away from an assembler was C, in which I happily program to this day. -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: Howard S Shubs Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Mon, 10 Mar 2003 23:31:05 -0500 Organization: ='SEQUENTIAL' Lines: 17 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> NNTP-Posting-Host: p-593.newsdawg.com Mail-Copies-To: nobody User-Agent: MT-NewsWatcher/3.2 (PPC Mac OS X) X-Face: "S"r{U%bs].&Ud}Pc~~~0a]M:t5l>>EN\1Faw10M9NK1Xq59wo7-"s0S+[{etQorO /Nf-Ci"i9v'MT!R8)J]N[4|2&x1r^Iq&{SB"6dknr0=+6UFb.>+{zMn_1=rw&/V+"d@* ZS5\LoW_ Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!howard Xref: chonsp.franklin.ch alt.folklore.computers:131794 In article , jmfbahciv@aol.com wrote: > The reason he spent three months decrypting the code was to > be able to get it into production so that this customer > wouldn't have a crisis again. That's how the biz is done.. > or was done in our niche. In the real olden days, before > my time, each computer installation was hand assembled > and installed in Maynard, including the software. Then it > was desembled very carefully, packaged up, shipped (the > humans were including in the packaging), and reassembled > on site. I -do- seem to recall seeing "field engineer" on the manifest list for a VAX 9000. -- Today, on Paper-view: The World Origami Championship ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5A710D.6030403@jetnet.ab.ca> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 38 Date: Tue, 11 Mar 2003 00:28:12 -0500 NNTP-Posting-Host: 209.96.185.251 X-Trace: sydney.visi.net 1047363956 209.96.185.251 (Tue, 11 Mar 2003 01:25:56 EST) NNTP-Posting-Date: Tue, 11 Mar 2003 01:25:56 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!eusc.inter.net!priapus.visi.com!news-out.visi.com!petbe.visi.com!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131710 In article , Peter Ibbotson wrote: >> Since there are good programs which do this, it must be possible in >> Windows. In fact, I'm not a Windows programmer, but I thought there >> was a message window class which would do exactly this. > > It's possible, however if you're writing in VB it's relatively tricky unless > you're displaying a modal form to show all error messages. I don't think you can justify one problem with yet another... :) > Certainly if you're programming in VB (Also applies to MFC but then you > could write a generic message handler) doing what you're suggesting is quite > tough for simple routines. Then the system is broken. But then, we knew that anyway, right? Win32? The irony of having "win" in there is just too much for words. > The problem is that unless you're already showing > a modal form it's fairly hard to stop the message processing for your main > application which will eventually bring reenterancy problems. To get the > screen to repaint you have to allow the message queue to be examined so it's > back around to the re-entrancy problems. Why a reentrancy problem? The event loop can even be used to handle the non-dialog notification. Or is this some problme that exists in Windows? I always thought the whole Win32 was a bag of rocks where a hammer was needed. > It's not that it's difficult to implement this stuff, just painful for > novices. Either way, the UIs stink... ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 15 Date: Tue, 11 Mar 2003 00:35:17 -0500 NNTP-Posting-Host: 209.96.185.251 X-Trace: sydney.visi.net 1047363960 209.96.185.251 (Tue, 11 Mar 2003 01:26:00 EST) NNTP-Posting-Date: Tue, 11 Mar 2003 01:26:00 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!news-out.visi.com!petbe.visi.com!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:131707 In article , jmfbahciv@aol.com wrote: > That command truly created superstition. There were > days that it sure felt like a ^T was necessary in order for > the machine to pay you some attention. Even those who knew > the code intimately would ^T to get attention. Not much different than: % sync % sync % sync ...on a UNIX machine, just in case that first sync missed something. ###### Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <6w65qvozeg.fsf@pbs.isi.edu> Organization: Chez Inwap X-Newsreader: trn 4.0-test76 (Apr 2, 2001) From: inwap@inwap.com (Joe Smith) Originator: inwap@inwap.com (Joe Smith) Lines: 20 Message-ID: Date: Tue, 11 Mar 2003 06:29:28 GMT NNTP-Posting-Host: 130.94.177.175 X-Complaints-To: abuse@verio.net X-Trace: iad-read.news.verio.net 1047364168 130.94.177.175 (Tue, 11 Mar 2003 06:29:28 GMT) NNTP-Posting-Date: Tue, 11 Mar 2003 06:29:28 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed-west.nntpserver.com!hub1.meganetnews.com!nntpserver.com!sjc-peer.news.verio.net!news.verio.net!iad-read.news.verio.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131797 alt.sys.pdp10:12741 In article <6w65qvozeg.fsf@pbs.isi.edu>, Greg Finn wrote: :Mark Crispin Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: X-Newsreader: Forte Agent 1.92/32.572 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 34 Date: Tue, 11 Mar 2003 16:19:49 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news2.calgary.shaw.ca 1047399589 24.71.223.147 (Tue, 11 Mar 2003 09:19:49 MST) NNTP-Posting-Date: Tue, 11 Mar 2003 09:19:49 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!nntp.theplanet.net!inewsm1.nntp.theplanet.net!195.40.4.120.MISMATCH!easynet-quince!easynet.net!cox.net!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news2.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131800 On Tue, 11 Mar 2003 00:35:17 -0500 in alt.folklore.computers, Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: > >> That command truly created superstition. There were >> days that it sure felt like a ^T was necessary in order for >> the machine to pay you some attention. Even those who knew >> the code intimately would ^T to get attention. > >Not much different than: > >% sync >% sync >% sync > >...on a UNIX machine, just in case that first sync missed something. On Suns, we used stop/sync/sync/halt to stop a machine immediately without warning. The second sync is required to wait until the whole buffer cache is flushed asynchronously to disk, before halting. The third sync is pure superstition, or maybe a buggy implementation of sync? The commands are issued to the console boot prom. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### From: "Christopher McNabb" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 11 Mar 2003 11:35:20 -0500 Organization: Virginia Tech Lines: 22 Message-ID: References: NNTP-Posting-Host: hc652a89b.dhcp.vt.edu Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 X-Trace: solaris.cc.vt.edu 1047400520 27790 198.82.168.155 (11 Mar 2003 16:35:20 GMT) X-Complaints-To: abuse@vt.edu NNTP-Posting-Date: Tue, 11 Mar 2003 16:35:20 +0000 (UTC) User-Agent: Pan/0.13.0 (The whole remains beautiful) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!nntp-out.monmouth.com!newspeer.monmouth.com!newsgate.duke.edu!solaris.cc.vt.edu!news.vt.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131780 On Tue, 11 Mar 2003 16:19:49 +0000, Brian Inglis wrote: > On Suns, we used stop/sync/sync/halt to stop a machine > immediately without warning. > The second sync is required to wait until the whole buffer cache > is flushed asynchronously to disk, before halting. > The third sync is pure superstition, or maybe a buggy > implementation of sync? > The commands are issued to the console boot prom. If a sun system is hung to the point where it's console is no longer usable, here is what I do: Stop-A (gives me the OK prompt) sync (Just one is necessary, since it Syncs disks and then forces a PANIC-ZERO) The system should reboot automagically after the OpenBoot sync. If everything is set up correctly (large enough swap space, enough free space in /var/crash) you should have a system dump in /var/crash/ when the system comes up. You can then use adb to determine exactly which process hung the system. ###### Message-ID: <3E6E0E58.E976BA50@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 26 Date: Tue, 11 Mar 2003 16:46:21 GMT NNTP-Posting-Host: 12.90.168.171 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc04-news.ops.worldnet.att.net 1047401181 12.90.168.171 (Tue, 11 Mar 2003 16:46:21 GMT) NNTP-Posting-Date: Tue, 11 Mar 2003 16:46:21 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!nycmny1-snh1.gtei.net!chcgil2-snh1.gtei.net!news.gtei.net!wn12feed!worldnet.att.net!bgtnsc04-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131713 jmfbahciv@aol.com wrote: > Charles Shannon Hendrix wrote: > > ... snip ... > > > >% sync > >% sync > >% sync > > > >....on a UNIX machine, just in case that first sync missed > > something. > > It would. Think about getting a car into first gear while > still moving. When you have lost your clutch cable, it is necessary. You start from a hill, coast, carefully synchronize engine/road speed, shift. All subsequent shifts are multi-phase - to neutral, sync, shift. An acquired skill. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### From: "Christopher McNabb" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 11 Mar 2003 12:01:12 -0500 Organization: Virginia Tech Lines: 20 Message-ID: References: <3E6E0E58.E976BA50@yahoo.com> NNTP-Posting-Host: hc652a89b.dhcp.vt.edu Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 X-Trace: solaris.cc.vt.edu 1047402072 11 198.82.168.155 (11 Mar 2003 17:01:12 GMT) X-Complaints-To: abuse@vt.edu NNTP-Posting-Date: Tue, 11 Mar 2003 17:01:12 +0000 (UTC) User-Agent: Pan/0.13.0 (The whole remains beautiful) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!iad-peer.news.verio.net!news.verio.net!solaris.cc.vt.edu!news.vt.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131778 On Tue, 11 Mar 2003 16:46:21 +0000, CBFalconer wrote: > > When you have lost your clutch cable, it is necessary. You start > from a hill, coast, carefully synchronize engine/road speed, > shift. All subsequent shifts are multi-phase - to neutral, sync, > shift. An acquired skill. I once had a 64 Chevy 3/4 ton pickup that had a clutch that was "locked up". The disk and the plate had literally welded together. I drove it for three months like that. Fortunately, the starter motor was sufficiently strong to get the truck rolling and the engine cranked over if I had the transmission in granny (very low gear - lower than first). Driving was real fun: 1. Put in Granny 2. Start motor 3. Speed shift through gears 4. At red light/stop sign let engine stall 5. Goto 1. ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 11 Mar 2003 20:27:26 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 26 Message-ID: References: <3E68E636.E638D524@ev1.net> <1539.196T2652T7445477@kltpzyxm.invalid> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1047414446 25226 146.186.61.46 (11 Mar 2003 20:27:26 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Tue, 11 Mar 2003 20:27:26 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!headwall.stanford.edu!newsfeed.stanford.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131750 In article <1539.196T2652T7445477@kltpzyxm.invalid>, Charlie Gibbs wrote: >In article >stanb45@dial.pipex.com (Stan Barr) writes: >>> So you will like Unix/Linux/FreeBSD. They have commands that allow >>> you to set up your jobs to run "batch". >> And many Mac applications are script-aware, allowing them to be run >>from Applescript programs. >But... but... all _modern_ systems are INTERACTIVE! Batch programming >is a throwback to the dinosaurs! Nobody who wants to be With the >Program [tm] would think of using that archaic stuff! > These silly people. They don't realize that all too often, >"interactive" is synonymous with "manual". let's see'em write a web server hawk, leering & cackling -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Tue, 11 Mar 2003 14:47:06 -0700 Lines: 28 Message-ID: <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> References: <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1frlud0.141htge15djdywN%lars@bearnip.com> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1047419227 67676911 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131829 Brian Inglis wrote: > On Mon, 10 Mar 2003 20:05:06 -0700 in alt.folklore.computers, > lars@bearnip.com (Lars Duening) wrote: > > >Maybe processors today are really that much faster, but chosing an O(N) > >or worse approach over an O(1) just seems ... wrong. Especially if the > >programs are meant to serve an unknown number of concurrent users. > > Measurement time! -- too difficult to estimate performance on > cached multiple-dispatch machines, other than to say that a > really tight brute force machine loop, with code and data in > first level cache, and no stride issues, is often faster than a > more sophisticated approach up to some number of items N, where N > is getting larger every year. True. Truth is running code, and I have often nixed a nifty algorithm simply because it didn't improve anything. However, one of these guys was arguing using guesstimated instruction counts (to paraphrase: "A comparison takes up 10..20 instruction cycles, so a linear search through a table of 100 entries..."). Hmpf. > Things were easier to control when you had memory to interleave, > and thrashing pages to worry about ...and even easier when every memory access took n cycles, regardless the circumstances :-) ###### NNTP-Posting-Date: Tue, 11 Mar 2003 16:05:31 -0600 From: "Jack Peacock" Newsgroups: alt.folklore.computers References: <3E05FE9D.B3A12286@vax6k.openecs.org><20030306065440.30f12471.steveo@eircom.net><20030307073602.4cd64ef0.steveo@eircom.net><0c0d4b.tr1.ln@teabag.cbhnet> <20030310200541.12ec6fa1.steveo@eircom.net> Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 11 Mar 2003 14:05:31 -0800 X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 Message-ID: Lines: 13 NNTP-Posting-Host: 208.57.2.36 X-Trace: sv3-kUw28wcA+lD3n1b4DmtZI1UeROlFScqD2L2LzhoDdMVcIje2g/4/ui7m3DVOMFEiKegHVPdc5nTygjp!aupsA2Er25ldrg+8y85ylK8dJucnGv1BXKCzV/Gvcceo3DSVNeCZLdMdXmTih0kTlOfVzHNjMAu+!nQ== X-Complaints-To: abuse@mpowercom.net X-DMCA-Complaints-To: abuse@mpowercom.net X-Abuse-and-DMCA-Info: Please be sure to forward a copy of ALL headers X-Abuse-and-DMCA-Info: Otherwise we will be unable to process your complaint properly X-Postfilter: 1.1 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!uni-erlangen.de!newsfeed.arcor-online.net!newsfeed.freenet.de!amsnews01.chello.com!news-hub.cableinet.net!blueyonder!nntp2.aus1.giganews.com!nntp.giganews.com!nntp3.aus1.giganews.com!nntp.mpowercom.net!news.mpowercom.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131828 wrote in message news:b4kq6g$fne$13@bob.news.rcn.net... > In article <20030310200541.12ec6fa1.steveo@eircom.net>, > I managed to escape. Doing machine language is a great > HLL avoidance coping strategy. > Best of all were the looks of awe from COBOLers when you added two hex number in your head. And then converting characters into ASCII hex from memory...but that was pure showing off. What impressed me is were the ones who could multiply in multi-digit hex. Some kind of mutant? Jack Peacock ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 11 Mar 2003 22:34:46 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 32 Message-ID: References: <3E6A0951.10371A99@ev1.net> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1047422086 22192 146.186.61.46 (11 Mar 2003 22:34:46 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Tue, 11 Mar 2003 22:34:46 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131812 In article , wrote: >In article <3E6A0951.10371A99@ev1.net>, > Charles Richmond wrote: >>jmfbahciv@aol.com wrote: >> ... A Pentium or >>Pentium II processor at least would be wonderful... If you >>do *not* have this system, go look at some of the local >>thrift stores for a used system. The Micky-Suck people are >>throwing away hardware of this description. >They are? Already? >>But carry a *lot* of cash!!! The system might cost you as >>much as twenty or thirty dollars US... ;-) >I hate shopping. First I'll have to find the store...JMF >was very useful for this; I could always send him out to go hunting. >Once he spent two entire afternoons hunting for 15-pound >fishing line and he knew nothing about fishing. Did he at least catch one? And did you ever catch the fish? hawk, curious -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 11 Mar 03 18:48:30 -0800 Organization: http://extra.newsguy.com Lines: 24 Message-ID: <892.200T1510T11284819@kltpzyxm.invalid> References: <20030310200541.12ec6fa1.steveo@eircom.net> NNTP-Posting-Host: p-855.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news-hog.berkeley.edu!ucberkeley!enews.sgi.com!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews3 Xref: chonsp.franklin.ch alt.folklore.computers:131883 In article peacock@simconv.com (Jack Peacock) writes: > wrote in message news:b4kq6g$fne$13@bob.news.rcn.net... > >> In article <20030310200541.12ec6fa1.steveo@eircom.net>, >> I managed to escape. Doing machine language is a great >> HLL avoidance coping strategy. > >Best of all were the looks of awe from COBOLers when you added two hex >number in your head. And then converting characters into ASCII hex >from memory...but that was pure showing off. What impressed me is >were the ones who could multiply in multi-digit hex. Some kind of >mutant? I tried long division. Once. -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 12 Mar 03 11:45:52 GMT Organization: UltraNet Communications, Inc. Lines: 24 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org><20030306065440.30f12471.steveo@eircom.net><20030307073602.4cd64ef0.steveo@eircom.net><0c0d4b.tr1.ln@teabag.cbhnet> <20030310200541.12ec6fa1.steveo@eircom.net> X-Trace: UmFuZG9tSVYQJN99Bl5gSsG8l3gfMw7Xo1XtM46aaj7+cUBy35DXTDiC5bwfWgyA X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 12 Mar 2003 12:31:39 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!news.webusenet.com!cyclone1.gnilink.net!wn14feed!wn13feed!wn12feed!worldnet.att.net!199.184.165.233!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-62 Xref: chonsp.franklin.ch alt.folklore.computers:131850 In article , "Jack Peacock" wrote: > wrote in message news:b4kq6g$fne$13@bob.news.rcn.net... >> In article <20030310200541.12ec6fa1.steveo@eircom.net>, > >> I managed to escape. Doing machine language is a great >> HLL avoidance coping strategy. >> >Best of all were the looks of awe from COBOLers when you added two hex >number in your head. And then converting characters into ASCII hex from >memory...but that was pure showing off. [awed emoticon here] I never learned to hex. >What impressed me is were the ones >who could multiply in multi-digit hex. Some kind of mutant? Too many hours spent debugging I bet. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 12 Mar 03 11:54:05 GMT Organization: UltraNet Communications, Inc. Lines: 54 Message-ID: References: <20030310200541.12ec6fa1.steveo@eircom.net> <1400.200T1419T6144712@kltpzyxm.invalid> X-Trace: UmFuZG9tSVbf1Q/Ku4tB5M5Ktee+sCnAvnrRhQHCZIhObCA9dkFSysbrzYfw/hgv X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 12 Mar 2003 12:39:52 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntp01.fccn.pt!news-out.nuthinbutnews.com!propagator2-sterling!news-in-sterling.newsfeed.com!priapus.visi.com!news-out.visi.com!hermes.visi.com!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-62 Xref: chonsp.franklin.ch alt.folklore.computers:131859 In article <1400.200T1419T6144712@kltpzyxm.invalid>, "Charlie Gibbs" wrote: >In article jmfbahciv@aol.com >(jmfbahciv) writes: > >>In article <20030310200541.12ec6fa1.steveo@eircom.net>, >>Steve O'Hara-Smith wrote: >> >>>On Sat, 8 Mar 2003 14:57:04 +0000 >>>cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) wrote: >>> >>>> but certain internal coders had been writing reams of COBOL for >>>> decades and had clearly been quite hastily trained... >>> >>> I can still remember the brain-fried expressions on their faces >>>as the pace of courses picked up (C, SQL, UNIX, OO design ...). They >>>went from keen to frazzled remarkably quickly :) >> >>I managed to escape. Doing machine language is a great >>HLL avoidance coping strategy. > >I learned that art in my university days, eventually dropping out >for a job in a small service bureau. They _really_ loved HLLs at >university - the more obscure the better. (One course hit you with >a new HLL every two weeks.) That's good; it kept kiddies from marrying one. > ... Meanwhile, I had quickly taken to >assembly language, having discovered the joys of not having a >compiler getting in the way. I don't JMF ever made the transition from using MACRO-10 and DDT to BLISS (or was it C?) for just that reason. The compiler kept getting in his way and refused to generate the instructions he needed to get the job done. These were always determined to be compiler bugs. > >The only thing that could wean me away from an assembler was C, >in which I happily program to this day. How long did it take you to wean? hmmm...maybe a better question is how old were you? I think Jim was 45 or 46. He converted and became skilled at it. But I don't think he ever got that thrill you can get when telling a machine directly what to do and it does and you actually accomplish what you wanted to get done. I can't explain it better. But I sure didn't get the same kind of thrill doing COBOL or FORTRAN; there was too much obscurity. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 12 Mar 03 11:56:52 GMT Organization: UltraNet Communications, Inc. Lines: 37 Message-ID: References: X-Trace: UmFuZG9tSVZfW1GnsNI6NuzHDeNrZLJW5MgCzZtfbeZrH9seRCAfFb4Fn3mD1r9l X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 12 Mar 2003 12:42:39 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-62 Xref: chonsp.franklin.ch alt.folklore.computers:131861 In article , Brian Inglis wrote: >On Tue, 11 Mar 2003 00:35:17 -0500 in alt.folklore.computers, >Charles Shannon Hendrix wrote: > >>In article , jmfbahciv@aol.com wrote: >> >>> That command truly created superstition. There were >>> days that it sure felt like a ^T was necessary in order for >>> the machine to pay you some attention. Even those who knew >>> the code intimately would ^T to get attention. >> >>Not much different than: >> >>% sync >>% sync >>% sync >> >>...on a UNIX machine, just in case that first sync missed something. > >On Suns, we used stop/sync/sync/halt to stop a machine >immediately without warning. >The second sync is required to wait until the whole buffer cache >is flushed asynchronously to disk, before halting. >The third sync is pure superstition, or maybe a buggy >implementation of sync? Nope. It's got to be an act of skin preservation. The reason 13 is a baker's dozen is to make sure the count isn't short. Why did we type three ^Cs when 2 (and sometimes one) is sufficient. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 12 Mar 03 12:00:04 GMT Organization: UltraNet Communications, Inc. Lines: 37 Message-ID: References: X-Trace: UmFuZG9tSVZdRrHhCkRdf9WGG4AMHhc062F+Cu4Tz7+/0ey0XknAm3wWZ2St+S3+ X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 12 Mar 2003 12:45:52 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!cyclone.bc.net!logbridge.uoregon.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-62 Xref: chonsp.franklin.ch alt.folklore.computers:131849 In article , hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: >In article , wrote: >>In article , >> Charles Shannon Hendrix wrote: > >>>What drives me nuts is that in some programs, I can specify a download >>>directory. The program puts up a dialog for every download. OK, I can >>>see that the user might want to rename the file. However, why not an >>>option to avoid this query for when you just want to queue up a bunch >>>of downloads? > >>Why not have full wildcard support or indirect files. > >You can't put it into windows unless an idiot can do it without >instruction. This produces a system which any idiot can use, but >there's no gains to experience for the intelligent user. This is just bullshit. Plenty of idiots learned how to talk to a PDP-10. >>>Silly user, put down that mouse, and let us take care of everything... > >>ROTFLMAO. Why bother flashing me? It's rather like the time >>my mother was cutting my hair when she said, "Ooops." > >Ahh, but that beats hearing it at a bris . . . > >:) ROTFL. I don't know; I'd say it's about the same thing with the exception that one them grows back. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 12 Mar 03 12:02:02 GMT Organization: UltraNet Communications, Inc. Lines: 39 Message-ID: References: <3E6A0951.10371A99@ev1.net> X-Trace: UmFuZG9tSVZvpa4u1ODTAlOa+Lzd0m+0ww6M+rjrUwJzQ7HiOwnil/85gE3w6eFT X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 12 Mar 2003 12:47:49 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-62 Xref: chonsp.franklin.ch alt.folklore.computers:131865 In article , hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: >In article , wrote: >>In article <3E6A0951.10371A99@ev1.net>, >> Charles Richmond wrote: >>>jmfbahciv@aol.com wrote: > > >>> ... A Pentium or >>>Pentium II processor at least would be wonderful... If you >>>do *not* have this system, go look at some of the local >>>thrift stores for a used system. The Micky-Suck people are >>>throwing away hardware of this description. > >>They are? Already? > >>>But carry a *lot* of cash!!! The system might cost you as >>>much as twenty or thirty dollars US... ;-) > >>I hate shopping. First I'll have to find the store...JMF >>was very useful for this; I could always send him out to go hunting. >>Once he spent two entire afternoons hunting for 15-pound >>fishing line and he knew nothing about fishing. > >Did he at least catch one? > >And did you ever catch the fish? Oh, it wasn't for fishing. (I could not convince JMF to go salmon fishing when we were in Ireland.) I needed the line to retie a chime that fell off. > >hawk, curious Disappointed? /BAH Subtract a hundred and four for e-mail. ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 12 Mar 03 10:10:09 -0800 Organization: http://extra.newsguy.com Lines: 25 Message-ID: <1003.201T458T6104211@kltpzyxm.invalid> References: NNTP-Posting-Host: p-144.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.he.net!cyclone-sf.pbi.net!129.250.175.17!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews3 Xref: chonsp.franklin.ch alt.folklore.computers:131879 In article hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) writes: >You can't put it into windows unless an idiot can do it without >instruction. This produces a system which any idiot can use, but >there's no gains to experience for the intelligent user. On the other hand, intelligent users are crippled by all the Fisher-Price stuff. But I suppose if you're trying to equalize everyone, it's the way to go. (The latest Analog science fiction magazine has a good story about equalization gone mad.) >>ROTFLMAO. Why bother flashing me? It's rather like the time >>my mother was cutting my hair when she said, "Ooops." > >Ahh, but that beats hearing it at a bris . . . Bill Cosby did a good skit about "oops". -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### Message-ID: <3E6EB195.BDABD66F@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org><20030306065440.30f12471.steveo@eircom.net><20030307073602.4cd64ef0.steveo@eircom.net><0c0d4b.tr1.ln@teabag.cbhnet> <20030310200541.12ec6fa1.steveo@eircom.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 24 Date: Wed, 12 Mar 2003 05:16:35 GMT NNTP-Posting-Host: 12.90.168.90 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc04-news.ops.worldnet.att.net 1047446195 12.90.168.90 (Wed, 12 Mar 2003 05:16:35 GMT) NNTP-Posting-Date: Wed, 12 Mar 2003 05:16:35 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!nntp-relay.ihug.net!ihug.co.nz!west.cox.net!cox.net!border3.nntp.aus1.giganews.com!nntp.giganews.com!wn12feed!worldnet.att.net!bgtnsc04-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131838 Jack Peacock wrote: > wrote in message news: > > > I managed to escape. Doing machine language is a great > > HLL avoidance coping strategy. > > > Best of all were the looks of awe from COBOLers when you added > two hex number in your head. And then converting characters > into ASCII hex from memory...but that was pure showing off. > What impressed me is were the ones who could multiply in > multi-digit hex. Some kind of mutant? 0x100 * 0x100 = 0x10000. What's the problem :-) However all it takes to impress the kiddies is to add any column of numbers. While they are busy misentering digits in a calculator, you look over their shoulder and tell them the sum. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 12 Mar 2003 15:38:48 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 41 Message-ID: References: NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1047483528 35710 146.186.61.46 (12 Mar 2003 15:38:48 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Wed, 12 Mar 2003 15:38:48 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!headwall.stanford.edu!newsfeed.stanford.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131856 In article , wrote: >In article , > hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: >>>Why not have full wildcard support or indirect files. >>You can't put it into windows unless an idiot can do it without >>instruction. This produces a system which any idiot can use, but >>there's no gains to experience for the intelligent user. >This is just bullshit. Plenty of idiots learned how to talk >to a PDP-10. You know that. I know that. Most of the people reading this newsgroup no that. Now if you could kindly get it across to Gates? :) I've never spent any serious amount of time in front of windows. It has always taken me less time to find a mac or a unix, or to write a 20k program, than to deal with it. It's strictly for loading tames on the kids computer, and for opening documents that Star Office can't. >>>>Silly user, put down that mouse, and let us take care of everything... >>>ROTFLMAO. Why bother flashing me? It's rather like the time >>>my mother was cutting my hair when she said, "Ooops." >>Ahh, but that beats hearing it at a bris . . . >>:) >ROTFL. I don't know; I'd say it's about the same thing with >the exception that one them grows back. I'll grant that unexpected hair loss for a woman will be tramatic, but I'm not convinced that it's in the same league . . . hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 12 Mar 03 11:16:20 -0800 Organization: http://extra.newsguy.com Lines: 37 Message-ID: <1675.201T1000T6764369@kltpzyxm.invalid> References: <20030310200541.12ec6fa1.steveo@eircom.net> <1400.200T1419T6144712@kltpzyxm.invalid> NNTP-Posting-Host: p-895.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news-fra1.dfn.de!eusc.inter.net!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.he.net!cyclone-sf.pbi.net!129.250.175.17!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews1 Xref: chonsp.franklin.ch alt.folklore.computers:131914 In article jmfbahciv@aol.com (jmfbahciv) writes: >In article <1400.200T1419T6144712@kltpzyxm.invalid>, >"Charlie Gibbs" wrote: > >>The only thing that could wean me away from an assembler was C, >>in which I happily program to this day. > >How long did it take you to wean? hmmm...maybe a better question >is how old were you? I think Jim was 45 or 46. He converted and >became skilled at it. But I don't think he ever got that thrill >you can get when telling a machine directly what to do and >it does and you actually accomplish what you wanted to get done. It happened when I was in my mid to late 30s, and took several years. I've never seen a mainframe C compiler, so it required the advent of C compilers running on personal computers (and personal computers powerful enough to run them) to complete the transition. (Aiding the process was the fact that x86 assembly language was sufficiently horrible that I _didn't_ want to get involved with it.) >I can't explain it better. But I sure didn't get the same kind >of thrill doing COBOL or FORTRAN; there was too much obscurity. Yeah, porting Dungeon from PDP-11 FORTRAN to something a Sperry->Unisys mainframe could digest was quite an experience. But I found creating and dissecting terminal control strings in my COBOL programs to be less of a pain than using the standard screen-handling packages... -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 12 Mar 03 11:18:36 -0800 Organization: http://extra.newsguy.com Lines: 20 Message-ID: <590.201T1829T6784954@kltpzyxm.invalid> References: NNTP-Posting-Host: p-896.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.belwue.de!newsfeed.arcor-online.net!news.tele.dk!news.tele.dk!small.news.tele.dk!newsfeed.icl.net!newsfeed.fjserv.net!logbridge.uoregon.edu!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews1 Xref: chonsp.franklin.ch alt.folklore.computers:131912 In article jmfbahciv@aol.com (jmfbahciv) writes: >>>% sync >>>% sync >>>% sync >Why did we type three ^Cs when 2 (and sometimes one) is sufficient. "Sync three times on the console if you want me. Close all the pipes if the answer is no..." (With apologies to Tony Orlando and... heck no, he deserves it.) -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 13 Mar 03 12:16:35 GMT Organization: UltraNet Communications, Inc. Lines: 26 Message-ID: References: <20030310200541.12ec6fa1.steveo@eircom.net> <892.200T1510T11284819@kltpzyxm.invalid> X-Trace: UmFuZG9tSVa0+0qAr6sJdm1Aw3XdlV/0OIJFWF4NlB+HE857ZgLmepkDaN7WqBJx X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 13 Mar 2003 13:02:33 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-83 Xref: chonsp.franklin.ch alt.folklore.computers:131983 In article <892.200T1510T11284819@kltpzyxm.invalid>, "Charlie Gibbs" wrote: >In article peacock@simconv.com >(Jack Peacock) writes: > >> wrote in message news:b4kq6g$fne$13@bob.news.rcn.net... >> >>> In article <20030310200541.12ec6fa1.steveo@eircom.net>, > >>> I managed to escape. Doing machine language is a great >>> HLL avoidance coping strategy. >> >>Best of all were the looks of awe from COBOLers when you added two hex >>number in your head. And then converting characters into ASCII hex >>from memory...but that was pure showing off. What impressed me is >>were the ones who could multiply in multi-digit hex. Some kind of >>mutant? > >I tried long division. Once. > Right. That's what computers are there for. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 13 Mar 03 12:18:47 GMT Organization: UltraNet Communications, Inc. Lines: 23 Message-ID: References: <590.201T1829T6784954@kltpzyxm.invalid> X-Trace: UmFuZG9tSVYvMW0/EvrzfRXrVaIFap9/lpnvM+VSja9YFYlwJvAvq9TWr2r9mJES X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 13 Mar 2003 13:04:45 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed00.sul.t-online.de!t-online.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-83 Xref: chonsp.franklin.ch alt.folklore.computers:131992 In article <590.201T1829T6784954@kltpzyxm.invalid>, "Charlie Gibbs" wrote: >In article jmfbahciv@aol.com (jmfbahciv) >writes: > >>>>% sync >>>>% sync >>>>% sync > >>Why did we type three ^Cs when 2 (and sometimes one) is sufficient. > >"Sync three times on the console if you want me. > Close all the pipes if the answer is no..." > >(With apologies to Tony Orlando and... heck no, he deserves it.) > Good one. And it's technically correct. Weren't those comm thingies called pipelines before CPUs got them installed? /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 13 Mar 03 12:21:00 GMT Organization: UltraNet Communications, Inc. Lines: 27 Message-ID: References: X-Trace: UmFuZG9tSVani2JC2uoa7SKMshhu51/zor3WfZCWZwHlATYLqAYSCyKHepjm5S9X X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 13 Mar 2003 13:06:58 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeeds.sol.net!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-181-83 Xref: chonsp.franklin.ch alt.folklore.computers:131997 In article , hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: >In article , wrote: >>In article , >> hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: > >>>>Why not have full wildcard support or indirect files. > >>>You can't put it into windows unless an idiot can do it without >>>instruction. This produces a system which any idiot can use, but >>>there's no gains to experience for the intelligent user. > >>This is just bullshit. Plenty of idiots learned how to talk >>to a PDP-10. > >You know that. I know that. Most of the people reading this newsgroup >no that. Now if you could kindly get it across to Gates? :) That's why he's a fucking idiot. He does know that and chooses to ignore that fact anyway. /BAH Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 13 Mar 03 10:14:09 -0800 Organization: http://extra.newsguy.com Lines: 23 Message-ID: <822.202T450T6143479@kltpzyxm.invalid> References: NNTP-Posting-Host: p-009.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!newsfeed.arcor-online.net!newsfeed.freenet.de!194.168.222.61.MISMATCH!newspeer1-gui.server.ntli.net!ntli.net!newsfeed.news2me.com!arclight.uoregon.edu!logbridge.uoregon.edu!lax2-feed1.news.algx.net!allegiance!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews3 Xref: chonsp.franklin.ch alt.folklore.computers:132025 In article jmfbahciv@aol.com (jmfbahciv) writes: >In article , >hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: > >>You know that. I know that. Most of the people reading this >>newsgroup no that. Now if you could kindly get it across to Gates? >>:) > >That's why he's a fucking idiot. He does know that and chooses >to ignore that fact anyway. He ignores it because it would interfere with his agenda. Bill Gates might be many things, but an idiot (fucking or otherwise) is not one of them. -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 13 Mar 03 11:48:27 -0800 Organization: http://extra.newsguy.com Lines: 21 Message-ID: <761.202T1385T7085194@kltpzyxm.invalid> References: <1003.201T458T6104211@kltpzyxm.invalid> NNTP-Posting-Host: p-010.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews3 Xref: chonsp.franklin.ch alt.folklore.computers:132028 In article hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) writes: >In article <1003.201T458T6104211@kltpzyxm.invalid>, >Charlie Gibbs wrote: > >>(The latest Analog science fiction magazine has a good story about >>equalization gone mad.) > >Did they reprint Vonnegut's "Handicapper General" (or whatever it was >called?). No, this is an original. But with all the talk about Vonnegut lately, it's obvious that I haven't read enough of his stuff. -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 14 Mar 03 11:53:34 GMT Organization: UltraNet Communications, Inc. Lines: 25 Message-ID: References: <822.202T450T6143479@kltpzyxm.invalid> X-Trace: UmFuZG9tSVYLt3igYQZ7ybDbFh5nCbiQejPRqZRuX/+pGcAqo5x2Vxj0kZlzz7RJ X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 14 Mar 2003 12:39:41 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news-fra1.dfn.de!news0.de.colt.net!peernews3.colt.net!colt.net!newsfeed.icl.net!newsfeed.fjserv.net!newshosting.com!news-xfer2.atl.newshosting.com!news-feed01.roc.ny.frontiernet.net!logbridge.uoregon.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-149 Xref: chonsp.franklin.ch alt.folklore.computers:131980 In article <822.202T450T6143479@kltpzyxm.invalid>, "Charlie Gibbs" wrote: >In article jmfbahciv@aol.com >(jmfbahciv) writes: > >>In article , >>hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: >> >>>You know that. I know that. Most of the people reading this >>>newsgroup no that. Now if you could kindly get it across to Gates? >>>:) >> >>That's why he's a fucking idiot. He does know that and chooses >>to ignore that fact anyway. > >He ignores it because it would interfere with his agenda. >Bill Gates might be many things, but an idiot (fucking or >otherwise) is not one of them. > I didn't say he was stupid. Idiot is a special category. Fucking idiot is even more refined. /BAH Subtract a hundred and four for e-mail. ###### Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> <20030310200541.12ec6fa1.steveo@eircom.net> Organization: Honest Chris' Sysadmin Emporium X-Newsreader: trn 4.0-test76 (Apr 2, 2001) From: cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) Originator: cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) Date: Thu, 13 Mar 2003 01:13:45 +0000 Message-ID: <90mo4b.f96.ln@teabag.cbhnet> Lines: 13 NNTP-Posting-Date: 13 Mar 2003 01:13:57 GMT NNTP-Posting-Host: 62.254.134.35 X-Trace: 1047518037 news.gradwell.net 66680 cbh/62.254.134.35 X-Complaints-To: news-abuse@gradwell.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!news.cabal.org.uk!news-peer.gradwell.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131940 According to Steve O'Hara-Smith : > I can still remember the brain-fried expressions on their faces > as the pace of courses picked up (C, SQL, UNIX, OO design ...). They > went from keen to frazzled remarkably quickly :) Some of them stayed frazzled for quite some time afterwards (I'm sure you know who :) Chris. -- "If the world was an orange it would be like much too small, y'know?" Neil, '84 Currently playing: Sing Sing - "The Joy Of Sing Sing" (again) http://www.chrishedley.com My stuff, including genealogy, other things, etc ###### Date: Thu, 13 Mar 2003 08:18:21 +0100 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <20030313081821.6893c1d7.steveo@eircom.net> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> <20030310200541.12ec6fa1.steveo@eircom.net> <90mo4b.f96.ln@teabag.cbhnet> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 27 Organization: EuroNet Internet NNTP-Posting-Date: 13 Mar 2003 15:31:43 GMT NNTP-Posting-Host: i0610.vwr.wanadoo.nl X-Trace: 1047569503 maya.euronet.nl 23914 194.134.210.101:1740 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!seven.news.surf.net!surfnet.nl!diablo.bit.nl!diablo.bit.nl!news2.euro.net!postnews1.euro.net!maya.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132042 On Thu, 13 Mar 2003 01:13:45 +0000 cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) wrote: CH> According to Steve O'Hara-Smith : CH> > I can still remember the brain-fried expressions on their faces CH> > as the pace of courses picked up (C, SQL, UNIX, OO design ...). They CH> > went from keen to frazzled remarkably quickly :) I notice that I forgot to mention, SSADM and Yourdon - the mix with OO design must have been particularly confusing :) CH> Some of them stayed frazzled for quite some time afterwards (I'm sure CH> you know who :) By the time I left Telford for the joys of Basingstoke they had divided into two groups. There were a few whose brains were still functioning and were starting to integrate it all and become useful. The others were given harmless side tasks and treated gently. There were a couple of people pretty much full time looking after them by then. I was later told that these two groups corresponded to those that put themselves on the transfer list and those that were put on it. -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 13 Mar 2003 15:09:49 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 26 Message-ID: References: <590.201T1829T6784954@kltpzyxm.invalid> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1047568189 41820 146.186.61.46 (13 Mar 2003 15:09:49 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Thu, 13 Mar 2003 15:09:49 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!triton.net!smallfeed.triton.net!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131988 In article <590.201T1829T6784954@kltpzyxm.invalid>, Charlie Gibbs wrote: >In article jmfbahciv@aol.com (jmfbahciv) >writes: >>>>% sync >>>>% sync >>>>% sync >>Why did we type three ^Cs when 2 (and sometimes one) is sufficient. >"Sync three times on the console if you want me. > Close all the pipes if the answer is no..." >(With apologies to Tony Orlando and... heck no, he deserves it.) I think that song goes back to at least the late 50's or early 60's. We'd all recognized the name of the country singer, but it's beyond me at the moment. hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 13 Mar 2003 15:13:24 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 43 Message-ID: References: <1003.201T458T6104211@kltpzyxm.invalid> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1047568404 41820 146.186.61.46 (13 Mar 2003 15:13:24 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Thu, 13 Mar 2003 15:13:24 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!iad-peer.news.verio.net!news.verio.net!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:131957 In article <1003.201T458T6104211@kltpzyxm.invalid>, Charlie Gibbs wrote: >In article hawk@slytherin.ds.psu.edu >(Dr. Richard E. Hawkins) writes: >>You can't put it into windows unless an idiot can do it without >>instruction. This produces a system which any idiot can use, but >>there's no gains to experience for the intelligent user. >On the other hand, intelligent users are crippled by all the >Fisher-Price stuff. But I suppose if you're trying to equalize >everyone, it's the way to go. (The latest Analog science fiction >magazine has a good story about equalization gone mad.) Did they reprint Vonnegut's "Handicapper General" (or whatever it was called?). It actually appeared on the op-ed page of the Wall Street Journal a few year ago (I think following a particularly silly application of the ADA, but before California closed down a "Shower Dance" facility in a strip joint as it would be inaccessible to a wheelchair bound stripper . . . hawk > >>>ROTFLMAO. Why bother flashing me? It's rather like the time >>>my mother was cutting my hair when she said, "Ooops." >> >>Ahh, but that beats hearing it at a bris . . . > >Bill Cosby did a good skit about "oops". > >-- >/~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) >\ / I'm really at ac.dekanfrus if you read it the right way. > X Top-posted messages will probably be ignored. See RFC1855. >/ \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! > -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: Pete Fenelon Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 13 Mar 2003 15:38:48 -0000 Organization: Posted via Supernews, http://www.supernews.com Message-ID: Sender: Pete Fenelon References: <1003.201T458T6104211@kltpzyxm.invalid> User-Agent: tin/1.5.16-20021229 ("Spiders") (UNIX) (FreeBSD/4.7-STABLE (i386)) X-Complaints-To: abuse@supernews.com Lines: 13 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newsfeed00.sul.t-online.de!t-online.de!diablo.theplanet.net!news.maxwell.syr.edu!sn-xit-03!sn-xit-01!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132056 Dr. Richard E. Hawkins wrote: > Did they reprint Vonnegut's "Handicapper General" (or whatever it was > called?). "Harrison Bergeron", ISTR? I last dug that one out when Bernie Ecclestone was suggesting weight penalties for successful cars in Formula 1 ;) pete -- pete@fenelon.com "there's no room for enigmas in built-up areas" HMHB ###### From: stremler@rohan.sdsu.edu Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 16 Mar 2003 01:25:35 GMT Organization: San Diego State University Lines: 18 Message-ID: References: NNTP-Posting-Host: rohan.sdsu.edu X-Trace: gondor.sdsu.edu 1047777935 30705 130.191.3.100 (16 Mar 2003 01:25:35 GMT) X-Complaints-To: news@newshub.sdsu.edu NNTP-Posting-Date: 16 Mar 2003 01:25:35 GMT User-Agent: tin/1.4.2-20000123 ("Polish") (UNIX) (SunOS/5.8 (sun4u)) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!psinet-eu-nl!news-out.newsfeeds.com!propagator2-maxim!news-in.spamkiller.net!tethys.csu.net!arclight.uoregon.edu!newshub.sdsu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132272 begin quoting Charles Shannon Hendrix : [snip] > Not much different than: > > % sync > % sync > % sync > > ...on a UNIX machine, just in case that first sync missed something. Heh. When I'm idling trying to remember what I meant to do next, I'll do that. It's apparently the no-op command used by my spinal cord when my brain is swapping in the "to do" list. -- --Stewart Stremler--------------------------------stremler@rohan.sdsu.edu-- Let me tell you about physics babes. On second thought, I'd better not. -- Paul Tomblin (1999) ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 96 Date: Mon, 17 Mar 2003 13:53:04 -0500 NNTP-Posting-Host: 209.96.179.108 X-Trace: sydney.visi.net 1047929130 209.96.179.108 (Mon, 17 Mar 2003 14:25:30 EST) NNTP-Posting-Date: Mon, 17 Mar 2003 14:25:30 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!ecngs!feeder.ecngs.de!213.24.141.54.MISMATCH!newsfeed.gamma.ru!Gamma.RU!feed2.news.rcn.net!rcn!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:132372 In article <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com>, Lars Duening wrote: > Charles Shannon Hendrix wrote: > >> C++ has a lot of good points, but as an OO language, I think its contrived >> and confusing. Objective C is much cleaner to me. > > On the other hand, C++ supports multi-paradigmatic programming because Because what? :) >> At the same time, I think the idea that all projects are expressed best >> in OO fashion is ridiculous. > > It's by far not a perfect language (especially the C heritage), but it's > one of the lesser evil ones (at least you can shoot your own foot with > it!) I think you meant "can't". I'd disagree, as C++ seems to allow all kinds of foot shooting, only now everyone who gets your classes can shoot yours or their foot... :) >> I don't like what happens with C++ libraries. It seems that even small >> programs end up large because they want to use a class or method which >> implies a lot of baggage. > > It reminds me a bit of the first days of C on small computers: > programmers left and right were bitching about the incredible size of > 'Hello World' in C, when a 28 Byte assembler program could to the same. That was the difference between code generation, and the difference has been acceptable. With C++ libraries, I often find a single function needs a huge amount of dependencies. That's not quite the same as a C compiler generating more overhead for an exec than assembler. It's also worth noting that this might not be C++'s fault. Maybe you can create classes such that low level routines don't need the whole class loaded to function, but that's how it seems to work. I wrote a program with a 4MB library. The function I wanted was a few K, but I got the whole chunk. In plain C, I could get only what I needed. Now, maybe there is a way to structure your C++ libraries to avoid this, I don't know. C++ by itself isn't that bad a compiler. It seems to be the libraries and/or the way they are used which causes the problems. > The large size of even small executables is annoying, granded, and my > gut feeling is that in part it is caused by the C model of separating > compilation and linking. I don't really see why. In C I can link in only functions I use. Now, if I am using shared libraries, the whole library will get loaded, but then that's expected. If you don't need the whole thing then you shouldn't be using it. >> Java... it sucks on every platform except Sun, Windows, and Apple. >> >> Write once, run anywhere is mythology. > > Maybe not anywhere, but on enough platforms to make it useful :-) Java to me was based on assumptions that haven't happened, like a quality runtime being available everywhere, finding a way to make it run fast, and people would magically write code that didn't make non-portable assumptions. I think too that there were some assumptions about certain parts of the Java runtime being highly optimized, and there being machines which could coprocess java bytecode. In the years since the Java idea got started, a lot of multi-platform libraries appeared which got much better. There is C/C++ code out there which runs across many more platforms than Java does, and is also fast, native code. So, you can't really say that Java enables something which cannot otherwise be done. The only serious advantage I see is running code without a build cycle, but that's of limited use to me. Most people either know how to compile their own packages, or will simply buy/download them prebuilt. Maybe it just needs more time. Of course, I'll never like the library, but I suppose that isn't changable now. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <3E5680EB.4030807@jetnet.ab.ca> <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1frlud0.141htge15djdywN%lars@bearnip.com> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 14 Date: Mon, 17 Mar 2003 13:55:41 -0500 NNTP-Posting-Host: 209.96.179.108 X-Trace: sydney.visi.net 1047929131 209.96.179.108 (Mon, 17 Mar 2003 14:25:31 EST) NNTP-Posting-Date: Mon, 17 Mar 2003 14:25:31 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.gamma.ru!Gamma.RU!feed2.news.rcn.net!rcn!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:132373 In article <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com>, Lars Duening wrote: > True. Truth is running code, and I have often nixed a nifty algorithm > simply because it didn't improve anything. In fact, you can often choose a better O(...) algorithm only to find it is worse. I remember Dr. Dobbs had an issue one time in the past where they showed the contradictions caused by some CPU architectures. Another one showed how writing a matrix multiply backward made it much faster on one CPU. ###### From: pechter@shell.monmouth.com (Bill/Carolyn Pechter) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 17 Mar 2003 16:42:08 -0500 Organization: Lakewood MicroSystems Lines: 60 Message-ID: References: NNTP-Posting-Host: shell.monmouth.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!nntp-out.monmouth.com!newspeer.monmouth.com!news.monmouth.com!shell.monmouth.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132377 In article , Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: >> In article , >> Charles Shannon Hendrix wrote: >>>In article , jmfbahciv@aol.com wrote: >>> >>>> That command truly created superstition. There were >>>> days that it sure felt like a ^T was necessary in order for >>>> the machine to pay you some attention. Even those who knew >>>> the code intimately would ^T to get attention. >>> >>>Not much different than: >>> >>>% sync >>>% sync >>>% sync >>> >>>....on a UNIX machine, just in case that first sync missed something. >> >> It would. Think about getting a car into first gear while > still >> moving. > >I don't follow you. > >The extra calls to sync don't do anything for you. > HOW UNTRUE... In the old PDP11 days (pretty much applicable here in alt.folklore.computers) the sync command was followed by seeking of RK05/RP04 or other disk heads in an attempt go get all the disk buffers written out to the disk media. You could hear and feel the activity and if you typed sync;sync;sync and halted the machine there was a fair chance of the disk being screwed and needing fsck... The reason is THE DISKS WERE MUCH SLOWER THEN as were PDP11's. The disk buffers could take a couple of seconds to be written out and the sync sync sync on the LA36 terminal at 300 baud gave enough time to flush buffers before you SAFELY hit the halt switch. (the la36 was fairly slow and their were fill characters used quite often with them). sync;sync;sync did not. Bill -- +---------------------------------------------------------------------------+ | Bill and/or Carolyn Pechter | pechter@shell.monmouth.com | | Bill Gates is a Persian cat and a monocle away from being a villain in | | a James Bond movie -- Dennis Miller | +---------------------------------------------------------------------------+ ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: 18 Mar 03 12:35:37 -0800 Organization: http://extra.newsguy.com Lines: 20 Message-ID: <851.207T1872T7555017@kltpzyxm.invalid> References: <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com> <1frznw0.8e1waqbp4y8N%lars@bearnip.com> NNTP-Posting-Host: p-785.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews3 Xref: chonsp.franklin.ch alt.folklore.computers:132517 In article <1frznw0.8e1waqbp4y8N%lars@bearnip.com> lars@bearnip.com (Lars Duening) writes: >Well, Java enables me to write Windows programs without touching a >Windows box except for testing purposes. With a cross-platform C/C++ >environment I would at least have to build the program on Windows - >with all the pain that brings with it. No need for pain - just use makefiles, like everywhere else. I edit my source on my Amiga, ship it over to the Windoze box, type "make" in an MS-DOS window, and it's done. (Then I send it to my SCO and Linux boxes and do the same thing there. Portability rules!) -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Tue, 18 Mar 2003 05:50:48 -0700 Lines: 67 Message-ID: <1frznw0.8e1waqbp4y8N%lars@bearnip.com> References: <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1047991849 73923672 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed00.sul.t-online.de!newsfeed01.sul.t-online.de!t-online.de!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132577 Charles Shannon Hendrix wrote: > In article <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com>, Lars Duening wrote: > > Charles Shannon Hendrix wrote: > > > >> C++ has a lot of good points, but as an OO language, I think its contrived > >> and confusing. Objective C is much cleaner to me. > > > > On the other hand, C++ supports multi-paradigmatic programming because > > Because what? :) Because the implementors have the same opinion as you do: > >> At the same time, I think the idea that all projects are expressed best > >> in OO fashion is ridiculous. > > It's by far not a perfect language (especially the C heritage), but it's > > one of the lesser evil ones (at least you can shoot your own foot with > > it!) > > I think you meant "can't". No, I did mean 'can', as it indicates usefulness for real programming tasks. > >> I don't like what happens with C++ libraries. It seems that even small > >> programs end up large because they want to use a class or method which > >> implies a lot of baggage. > > > > It reminds me a bit of the first days of C on small computers: > > programmers left and right were bitching about the incredible size of > > 'Hello World' in C, when a 28 Byte assembler program could to the same. > > That was the difference between code generation, and the difference > has been acceptable. Code generation and libraries. Most novice C programmers used printf(), and back then printf() tended to pull in all the floating point libraries as well. C++ implementations have the same problem, one magnitude larger, as the standard library is rather complex. > > The large size of even small executables is annoying, granded, and my > > gut feeling is that in part it is caused by the C model of separating > > compilation and linking. > > I don't really see why. In C I can link in only functions I use. Templates, for example. A smart build system would generate the code for required templates once and then use that wherever the same template is required. With the current build model however the compiler has to re-create the template code over and over for the different source files, hoping that the linker throws out the duplicates. Some compilers (like g++) improve over this behaviour, but you have to drop heavy non-portable hints into the source for that. > There is C/C++ code out there which runs across many more platforms than > Java does, and is also fast, native code. So, you can't really say that > Java enables something which cannot otherwise be done. Well, Java enables me to write Windows programs without touching a Windows box except for testing purposes. With a cross-platform C/C++ environment I would at least have to build the program on Windows - with all the pain that brings with it. ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Tue, 18 Mar 2003 05:50:50 -0700 Lines: 8 Message-ID: <1frzp18.nrtco23ml5z4N%lars@bearnip.com> References: <3E5A710D.6030403@jetnet.ab.ca> <1003.201T458T6104211@kltpzyxm.invalid> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1047991850 73923672 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!opentransit.net!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132579 Charles Shannon Hendrix wrote: > Racing used to be so much more interesting before massive ads, and > track designed to fit in a TV camera lens. Especially since on the TV you only get to see the three leaders driving their rounds, and nothing from what is happening in the ranks further back. ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Tue, 18 Mar 2003 15:19:28 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 25 Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1048000768 48804 146.186.61.46 (18 Mar 2003 15:19:28 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Tue, 18 Mar 2003 15:19:28 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132469 In article , Charles Shannon Hendrix wrote: >In article <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com>, Lars Duening wrote: >> True. Truth is running code, and I have often nixed a nifty algorithm >> simply because it didn't improve anything. >In fact, you can often choose a better O(...) algorithm only to find >it is worse. Bubble sorts for small lists! Yay! :) Actually, the most recent model I was working on had a screwball, hand-written sort for groups of 4, as I only needed to distinguish the 1st & 2nd hightest. An earlier version didn't even distinguish that, but just the top two and bottom two. hawk hawk, poking the emb -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> X-Newsreader: Forte Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 34 Date: Wed, 19 Mar 2003 05:58:47 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news1.calgary.shaw.ca 1048053527 24.71.223.147 (Tue, 18 Mar 2003 22:58:47 MST) NNTP-Posting-Date: Tue, 18 Mar 2003 22:58:47 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!nntp-relay.ihug.net!ihug.co.nz!west.cox.net!cox.net!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news1.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132583 On Tue, 18 Mar 2003 15:19:28 +0000 (UTC) in alt.folklore.computers, hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: >In article , >Charles Shannon Hendrix wrote: >>In article <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com>, Lars Duening wrote: > >>> True. Truth is running code, and I have often nixed a nifty algorithm >>> simply because it didn't improve anything. > >>In fact, you can often choose a better O(...) algorithm only to find >>it is worse. > >Bubble sorts for small lists! Yay! :) There's never an excuse other than mental laziness for a bubble sort. *ALL* other sorts are faster! >Actually, the most recent model I was working on had a screwball, >hand-written sort for groups of 4, as I only needed to distinguish the >1st & 2nd hightest. An earlier version didn't even distinguish that, >but just the top two and bottom two. Hopefully, a tailored selection sort. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### From: "Charlie Gibbs" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 17 Mar 03 16:09:31 -0800 Organization: http://extra.newsguy.com Lines: 27 Message-ID: <1145.206T1599T9694896@kltpzyxm.invalid> References: <3E6E0E58.E976BA50@yahoo.com> NNTP-Posting-Host: p-677.newsdawg.com X-Newsreader: THOR 2.5a (Amiga;TCP/IP) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews3 Xref: chonsp.franklin.ch alt.folklore.computers:132530 In article shannon@news.widomaker.com (Charles Shannon Hendrix) writes: > In article <3E6E0E58.E976BA50@yahoo.com>, CBFalconer wrote: > >> When you have lost your clutch cable, it is necessary. You start > > Like in a VW Beetle when the clutch cable goes. Fun, trying to get > all >the way home, and fudging a little at yellow lights and stop > signs... :) > > You can't shift out of the gear, and if in town you just hope you are > in second gear. I once had a motorbike whose transmission linkage would occasionally pop out. The clutch would still work, but the transmission would go to 4th gear and stay there. It took a lot of running to go from a standing start with that little 80cc engine. The decision as to whether to make a roadside repair was based on how close to home I was, and whether the screws had frozen in the aluminum housing. -- /~\ cgibbs@kltpzyxm.invalid (Charlie Gibbs) \ / I'm really at ac.dekanfrus if you read it the right way. X Top-posted messages will probably be ignored. See RFC1855. / \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign! ###### From: pechter@shell.monmouth.com (Bill/Carolyn Pechter) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 17 Mar 2003 16:42:08 -0500 Organization: Lakewood MicroSystems Lines: 60 Message-ID: References: NNTP-Posting-Host: shell.monmouth.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!nntp-out.monmouth.com!newspeer.monmouth.com!news.monmouth.com!shell.monmouth.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132392 In article , Charles Shannon Hendrix wrote: >In article , jmfbahciv@aol.com wrote: >> In article , >> Charles Shannon Hendrix wrote: >>>In article , jmfbahciv@aol.com wrote: >>> >>>> That command truly created superstition. There were >>>> days that it sure felt like a ^T was necessary in order for >>>> the machine to pay you some attention. Even those who knew >>>> the code intimately would ^T to get attention. >>> >>>Not much different than: >>> >>>% sync >>>% sync >>>% sync >>> >>>....on a UNIX machine, just in case that first sync missed something. >> >> It would. Think about getting a car into first gear while > still >> moving. > >I don't follow you. > >The extra calls to sync don't do anything for you. > HOW UNTRUE... In the old PDP11 days (pretty much applicable here in alt.folklore.computers) the sync command was followed by seeking of RK05/RP04 or other disk heads in an attempt go get all the disk buffers written out to the disk media. You could hear and feel the activity and if you typed sync;sync;sync and halted the machine there was a fair chance of the disk being screwed and needing fsck... The reason is THE DISKS WERE MUCH SLOWER THEN as were PDP11's. The disk buffers could take a couple of seconds to be written out and the sync sync sync on the LA36 terminal at 300 baud gave enough time to flush buffers before you SAFELY hit the halt switch. (the la36 was fairly slow and their were fill characters used quite often with them). sync;sync;sync did not. Bill -- +---------------------------------------------------------------------------+ | Bill and/or Carolyn Pechter | pechter@shell.monmouth.com | | Bill Gates is a Persian cat and a monocle away from being a villain in | | a James Bond movie -- Dennis Miller | +---------------------------------------------------------------------------+ ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030306065440.30f12471.steveo@eircom.net> <20030307073602.4cd64ef0.steveo@eircom.net> <0c0d4b.tr1.ln@teabag.cbhnet> <20030310200541.12ec6fa1.steveo@eircom.net> <3E6EB195.BDABD66F@yahoo.com> X-Newsreader: Forte Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 65 Date: Tue, 18 Mar 2003 09:09:36 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news1.calgary.shaw.ca 1047978576 24.71.223.147 (Tue, 18 Mar 2003 02:09:36 MST) NNTP-Posting-Date: Tue, 18 Mar 2003 02:09:36 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!c03.atl99!sjc70.webusenet.com!news.webusenet.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news1.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132601 On Mon, 17 Mar 2003 14:59:09 -0500 in alt.folklore.computers, Charles Shannon Hendrix wrote: >In article <3E6EB195.BDABD66F@yahoo.com>, CBFalconer wrote: > >> 0x100 * 0x100 = 0x10000. What's the problem :-) > >What a cheap problem, ya bum. > >This is more interesting: > >0xF00F * 0x0FF1 == F00F * (1000 - 0010 + 0001) == (10000 - 1000 + 10 - 1) * (1000 - 10 + 1) == (10010 - 1001) * (1001 - 10) == 10010000 + 10010 - 1001000 - 1001 - 100100 + 10010 == 10030020 -1102101 -------- EF2DF1F ======== OR F00F000 - F00F0 ------- EF1EF10 + F00F ------- EF2DF1F ======= OR F00F000 + F00F ------- F01E00F - F00F0 ------- EF2DF1F ======= checked on TI35+ after the fact >> However all it takes to impress the kiddies is to add any column >> of numbers. While they are busy misentering digits in a >> calculator, you look over their shoulder and tell them the sum. > >Do you do this with a stack? > >I always go down the column and remember as little as possible. One >stack for the answer digits, another for the carries. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 18 Mar 2003 15:08:59 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 34 Message-ID: References: NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1048000139 48804 146.186.61.46 (18 Mar 2003 15:08:59 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Tue, 18 Mar 2003 15:08:59 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!iad-peer.news.verio.net!news.verio.net!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132476 In article , wrote: >In article , > hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: >>>I hate shopping. First I'll have to find the store...JMF >>>was very useful for this; I could always send him out to go hunting. >>>Once he spent two entire afternoons hunting for 15-pound >>>fishing line and he knew nothing about fishing. >>Did he at least catch one? >>And did you ever catch the fish? >Oh, it wasn't for fishing. (I could not convince JMF to go >salmon fishing when we were in Ireland.) I needed the line >to retie a chime that fell off. >>hawk, curious >Disappointed? Nah, not as long as he at least bought you a fish :) And to keep a folklore notion, there was indeed product defect litigation over fishing line that lacked warnings about using it to hang pictures. Even though this wasn't the intended use, the fact that it was used that way often enough . . . hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 18 Mar 2003 15:11:40 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 23 Message-ID: References: <3E6E0E58.E976BA50@yahoo.com> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1048000300 48804 146.186.61.46 (18 Mar 2003 15:11:40 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Tue, 18 Mar 2003 15:11:40 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132480 In article , Charles Shannon Hendrix wrote: >In article <3E6E0E58.E976BA50@yahoo.com>, CBFalconer wrote: >> When you have lost your clutch cable, it is necessary. You start >Like in a VW Beetle when the clutch cable goes. Fun, trying to get all >the way home, and fudging a little at yellow lights and stop signs... :) Or when it's too tight. My girlfriend and I drove 80 miles the day after my clutch was replaced, off to meet her parents. Partway there, the engine would start racing if I went to fast. hmm. Her father looked at it, drove 10 feet, and told me what the problem was. It was an interesting drive back, as the maximum speed kept dropping, eventually down to 53 . . . hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### Message-ID: <3E776132.2B2A2AAD@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E6E0E58.E976BA50@yahoo.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 28 Date: Tue, 18 Mar 2003 18:51:52 GMT NNTP-Posting-Host: 12.90.170.15 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc04-news.ops.worldnet.att.net 1048013512 12.90.170.15 (Tue, 18 Mar 2003 18:51:52 GMT) NNTP-Posting-Date: Tue, 18 Mar 2003 18:51:52 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!eusc.inter.net!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!rcn!wn14feed!worldnet.att.net!bgtnsc04-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132407 "Dr. Richard E. Hawkins" wrote: > > In article , > Charles Shannon Hendrix wrote: > >In article <3E6E0E58.E976BA50@yahoo.com>, CBFalconer wrote: > > >> When you have lost your clutch cable, it is necessary. You start > > >Like in a VW Beetle when the clutch cable goes. Fun, trying to get all > >the way home, and fudging a little at yellow lights and stop signs... :) > > Or when it's too tight. My girlfriend and I drove 80 miles the day > after my clutch was replaced, off to meet her parents. Partway there, > the engine would start racing if I went to fast. hmm. Her father > looked at it, drove 10 feet, and told me what the problem was. > > It was an interesting drive back, as the maximum speed kept dropping, > eventually down to 53 . . . You probably needed another new clutch, after all that slipping. That's why you always keep adequate free play in the clutch pedal. You had a very sloppy mechanic. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### Sender: dowe@krikkit.localhost Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E6E0E58.E976BA50@yahoo.com> From: Dowe Keller Date: 18 Mar 2003 22:19:42 -0800 Message-ID: Lines: 24 X-Newsreader: Gnus v5.7/Emacs 20.7 NNTP-Posting-Host: 63-93-78-241.lsan.sti.net X-Original-NNTP-Posting-Host: 63-93-78-241.lsan.sti.net X-Trace: news.sti.net 1048054776 63.93.78.241 (18 Mar 2003 22:19:36 -0800) Organization: news.sti.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed1.cidera.com!Cidera!news.sti.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132553 hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) writes: > In article , > Charles Shannon Hendrix wrote: > >In article <3E6E0E58.E976BA50@yahoo.com>, CBFalconer wrote: > > >> When you have lost your clutch cable, it is necessary. You start > > >Like in a VW Beetle when the clutch cable goes. Fun, trying to get all > >the way home, and fudging a little at yellow lights and stop signs... :) > > Or when it's too tight. My girlfriend and I drove 80 miles the day > after my clutch was replaced, off to meet her parents. Partway there, > the engine would start racing if I went to fast. hmm. Her father > looked at it, drove 10 feet, and told me what the problem was. Eek! Don't do that! If the clutch slips, stop right there and adjust the cable, otherwise, you end up having to get a new one (clutch that is, not the cable. Kind of like that $10 transistor burning out to save the $0.10 fuse. -- dowe@sierratel.com http://www.sierratel.com/dowe ###### NNTP-Posting-Date: Wed, 19 Mar 2003 10:37:28 -0600 Date: Wed, 19 Mar 2003 11:35:18 -0500 From: John Stracke User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.1) Gecko/20020826 X-Accept-Language: en-us, en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Message-ID: Lines: 14 NNTP-Posting-Host: 65.220.90.253 X-Trace: sv3-2VWwmBd6ozmSxszqXYY20dUSKSFAUpt33ODWL1NnEUwRt1d71hPMsz/LbC+7bDTFvYh0DFRAEjBvc88!LXR4BG06Ojing0yOtf9r07FrorBpi0FEV/BRR3Z29AuxXh8FH+2UGjHpOzDSQ1ps4nrT3r2ZbEj2!VQ+AHC8nCtc5BwRLO9/C X-Complaints-To: abuse@speakeasy.net X-DMCA-Complaints-To: abuse@speakeasy.net X-Abuse-and-DMCA-Info: Please be sure to forward a copy of ALL headers X-Abuse-and-DMCA-Info: Otherwise we will be unable to process your complaint properly X-Postfilter: 1.1 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!colt.net!easynet-quince!easynet.net!cox.net!border3.nntp.aus1.giganews.com!nntp.giganews.com!nntp3.aus1.giganews.com!nntp.speakeasy.net!news.speakeasy.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132696 Brian Inglis wrote: > There's never an excuse other than mental laziness for a bubble > sort. They're fun to watch, though. :-) -- /=============================================================\ |John Stracke | http://www.thibault.org |HTML OK | |Francois Thibault |========================================| |East Kingdom |So what's the gene for belief in genetic| |francis@thibault.org|determinism? | \=============================================================/ ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Wed, 19 Mar 2003 17:14:16 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 35 Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1048094056 45986 146.186.61.46 (19 Mar 2003 17:14:16 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Wed, 19 Mar 2003 17:14:16 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!efnet.com!diablo.efnet.com!cyclone.bc.net!logbridge.uoregon.edu!news.cse.psu.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132637 In article , Brian Inglis wrote: >On Tue, 18 Mar 2003 15:19:28 +0000 (UTC) in >alt.folklore.computers, hawk@slytherin.ds.psu.edu (Dr. Richard E. >Hawkins) wrote: >>>In fact, you can often choose a better O(...) algorithm only to find >>>it is worse. >>Bubble sorts for small lists! Yay! :) >There's never an excuse other than mental laziness for a bubble >sort. *ALL* other sorts are faster! All other sorts have a better O(). For small enough lists, the overhead on each action can (sometimes) make a bubble sort faster, iirc. >>Actually, the most recent model I was working on had a screwball, >>hand-written sort for groups of 4, as I only needed to distinguish the >>1st & 2nd hightest. An earlier version didn't even distinguish that, >>but just the top two and bottom two. >Hopefully, a tailored selection sort. It was defintiely tailored :) A couple of other rolkds took up interst, and I think we actually proved that we had the minimum number of comparisons to cover all cases . . . hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Wed, 19 Mar 2003 17:15:46 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 19 Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1048094146 45986 146.186.61.46 (19 Mar 2003 17:15:46 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Wed, 19 Mar 2003 17:15:46 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132636 In article , John Stracke wrote: >Brian Inglis wrote: >> There's never an excuse other than mental laziness for a bubble >> sort. >They're fun to watch, though. :-) Unfortunately, there are people without the mental capacity to understand how they work! What chance have we to explain other algorithms to these folks? :( hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### Message-ID: From: jguthrie@brokersys.com Subject: Re: PDP10 and RISC Newsgroups: alt.folklore.computers References: <1by94bd27o.fsf@cs.nmsu.edu> User-Agent: tin/1.5.17-20030301 ("Bubbles") (UNIX) (Linux/2.4.18-k7+xfs+freeswan (i686)) Date: Wed, 19 Mar 2003 12:09:14 -0600 NNTP-Posting-Host: 216.39.196.197 X-Trace: news3.mylinuxisp.com 1048098303 216.39.196.197 (19 Mar 2003 12:25:03 -0600) Lines: 19 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!diablo.theplanet.net!newsfeed1.cidera.com!Cidera!news3.mylinuxisp.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132706 Dr. Richard E. Hawkins wrote: > In article , > Brian Inglis wrote: >>There's never an excuse other than mental laziness for a bubble >>sort. *ALL* other sorts are faster! > All other sorts have a better O(). For small enough lists, the overhead > on each action can (sometimes) make a bubble sort faster, iirc. As far as I am aware, neither of your sentences are true. It is not necessarily true that more advanced sorts have better O() in the worst case and it is not true that a bubble sort has the smallest K of any possible sorts. I usually use an extraction sort for doing simple things. It's overhead is pretty small. -- Jonathan Guthrie (jguthrie@brokersys.com) Sto pro veritate ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Wed, 19 Mar 2003 12:16:40 -0700 Lines: 21 Message-ID: <1fs2bof.1xrda5x1w2b5luN%lars@bearnip.com> References: <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com> <1frznw0.8e1waqbp4y8N%lars@bearnip.com> <851.207T1872T7555017@kltpzyxm.invalid> NNTP-Posting-Host: dhcp-dsl-31-26-33.interfold.com (198.31.26.33) X-Trace: fu-berlin.de 1048101402 74454425 198.31.26.33 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-33.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132712 Charlie Gibbs wrote: > In article <1frznw0.8e1waqbp4y8N%lars@bearnip.com> lars@bearnip.com > (Lars Duening) writes: > > >Well, Java enables me to write Windows programs without touching a > >Windows box except for testing purposes. With a cross-platform C/C++ > >environment I would at least have to build the program on Windows - > >with all the pain that brings with it. > > No need for pain - just use makefiles, like everywhere else. > I edit my source on my Amiga, ship it over to the Windoze box, > type "make" in an MS-DOS window, and it's done. Which of course assumes that you first have all the development software installed :-) It's a thought I have whenever I install the Cygwin environment on a Windows machine: since I'm basically installing an OS compatibility environment in order to get programs compiled, what is the point of using Windows at all? ###### Message-ID: <3E78C838.15934485@thinkage.ca> From: "Alan T. Bowler" Organization: Thinkage Ltd. X-Mailer: Mozilla 4.5 [en] (WinNT; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 14 Date: Wed, 19 Mar 2003 14:42:48 -0500 NNTP-Posting-Host: 192.102.11.4 X-Trace: nnrp1.uunet.ca 1048102894 192.102.11.4 (Wed, 19 Mar 2003 14:41:34 EST) NNTP-Posting-Date: Wed, 19 Mar 2003 14:41:34 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!news.uunet.ca!nnrp1.uunet.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132626 Brian Inglis wrote: > >Bubble sorts for small lists! Yay! :) > > There's never an excuse other than mental laziness for a bubble > sort. *ALL* other sorts are faster! > Not really. On very short stuff the low setup of the bubble sorts (I'm including "insertion sort" here), will win out. However, the list last to very short. Shellsort is only just a tad more overhead and wins somewhere between 5 and 20 items. In cases, where there is a high probability that the data is already sorted a bubbble sort can be a clear winner even for longer lists. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 19 Mar 03 13:47:58 GMT Organization: UltraNet Communications, Inc. Lines: 41 Message-ID: References: X-Trace: UmFuZG9tSVbglapG3OtMz+9fNJxX4MbWeMtw/J7gOkcb6KiKz0Y7EhXTNgo5dfWF X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 19 Mar 2003 14:35:01 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-104 Xref: chonsp.franklin.ch alt.folklore.computers:132643 In article , hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: >In article , wrote: >>In article , >> hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: > >>>>I hate shopping. First I'll have to find the store...JMF >>>>was very useful for this; I could always send him out to go hunting. >>>>Once he spent two entire afternoons hunting for 15-pound >>>>fishing line and he knew nothing about fishing. > >>>Did he at least catch one? > >>>And did you ever catch the fish? > >>Oh, it wasn't for fishing. (I could not convince JMF to go >>salmon fishing when we were in Ireland.) I needed the line >>to retie a chime that fell off. > >>>hawk, curious > >>Disappointed? > >Nah, not as long as he at least bought you a fish :) Of course he did. I refused to eat anything that came from Stop&Shop. I discovered that Ireland had the best seafood in the world. That's all I ate when we were there (except the lunch where the pub served real shepherds pie). >And to keep a folklore notion, there was indeed product defect >litigation over fishing line that lacked warnings about using it to hang >pictures. Even though this wasn't the intended use, the fact that it >was used that way often enough . . . Huh? Who would be foolish enough to do that? Fishing line is too thin. /BAH Subtract a hundred and four for e-mail. ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 19 Mar 2003 17:09:58 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 41 Message-ID: References: <3E776132.2B2A2AAD@yahoo.com> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1048093798 45986 146.186.61.46 (19 Mar 2003 17:09:58 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Wed, 19 Mar 2003 17:09:58 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132634 In article <3E776132.2B2A2AAD@yahoo.com>, CBFalconer wrote: >"Dr. Richard E. Hawkins" wrote: >> >> When you have lost your clutch cable, it is necessary. You start >> >Like in a VW Beetle when the clutch cable goes. Fun, trying to get all >> >the way home, and fudging a little at yellow lights and stop signs... :) >> Or when it's too tight. My girlfriend and I drove 80 miles the day >> after my clutch was replaced, off to meet her parents. Partway there, >> the engine would start racing if I went to fast. hmm. Her father >> looked at it, drove 10 feet, and told me what the problem was. >> It was an interesting drive back, as the maximum speed kept dropping, >> eventually down to 53 . . . >You probably needed another new clutch, after all that slipping. >That's why you always keep adequate free play in the clutch >pedal. You had a very sloppy mechanic. Yep. I showed up on monday, and said that they needed to redo my clutch. The lady snarled, "how many miles does it have on it?" "400." "Oh." It wasn't until later that I drew the connection between going in for service there and the new problems that arose within a day or two. plugging that loose wire (a factory test wire?) into the fuse block was the real joy. We'd traced the entire electrical system and lifted the upholstery trying to find the short that caused the fuse to blow when the taillights went on. Suddenly inspiration hit: what if I remove this wire, which seems to have no purpose? Amazing how much better things work when you don't connect +12 to GND . . . hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 19 Mar 2003 17:11:16 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 28 Message-ID: References: NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1048093876 45986 146.186.61.46 (19 Mar 2003 17:11:16 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Wed, 19 Mar 2003 17:11:16 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!msc1.onvoy!onvoy.com!hammer.uoregon.edu!logbridge.uoregon.edu!news.cse.psu.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132649 In article , Dowe Keller wrote: >hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) writes: >> Or when it's too tight. My girlfriend and I drove 80 miles the day >> after my clutch was replaced, off to meet her parents. Partway there, >> the engine would start racing if I went to fast. hmm. Her father >> looked at it, drove 10 feet, and told me what the problem was. >Eek! Don't do that! If the clutch slips, stop right there and adjust >the cable, otherwise, you end up having to get a new one (clutch that >is, not the cable. Had I knwon how, I probably would have. But then, I had a full load of tools, and her father almost certainly knew how. With 100 miles left to go, and being a new clutch and the mechanics screwup, I suppose it was wiser not to get in the way and get blamed :) >Kind of like that $10 transistor burning out to save the $0.10 fuse. Ooh, I hate that :) hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 19 Mar 2003 17:18:53 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 36 Message-ID: References: NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1048094333 45986 146.186.61.46 (19 Mar 2003 17:18:53 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Wed, 19 Mar 2003 17:18:53 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132653 In article , wrote: >In article , > hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: >>>>hawk, curious >>>Disappointed? >>Nah, not as long as he at least bought you a fish :) >Of course he did. I refused to eat anything that came from >Stop&Shop. I discovered that Ireland had the best seafood >in the world. That's all I ate when we were there (except >the lunch where the pub served real shepherds pie). Wow. Irish shepherds need a better union if they're *still* getting cooked by the british . . . :) >>And to keep a folklore notion, there was indeed product defect >>litigation over fishing line that lacked warnings about using it to hang >>pictures. Even though this wasn't the intended use, the fact that it >>was used that way often enough . . . >Huh? Who would be foolish enough to do that? Fishing line is too >thin. "There is no lower limit to human intelligence." --me hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: <2qah7vc7dfm559m3hdpe6o0m0kmpve2ajr@4ax.com> References: <3E776132.2B2A2AAD@yahoo.com> X-Newsreader: Forte Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 66 Date: Wed, 19 Mar 2003 17:52:04 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news1.calgary.shaw.ca 1048096324 24.71.223.147 (Wed, 19 Mar 2003 10:52:04 MST) NNTP-Posting-Date: Wed, 19 Mar 2003 10:52:04 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!news.webusenet.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news1.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132715 On Wed, 19 Mar 2003 17:09:58 +0000 (UTC) in alt.folklore.computers, hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: >In article <3E776132.2B2A2AAD@yahoo.com>, >CBFalconer wrote: >>"Dr. Richard E. Hawkins" wrote: > >>> >> When you have lost your clutch cable, it is necessary. You start > >>> >Like in a VW Beetle when the clutch cable goes. Fun, trying to get all >>> >the way home, and fudging a little at yellow lights and stop signs... :) > >>> Or when it's too tight. My girlfriend and I drove 80 miles the day >>> after my clutch was replaced, off to meet her parents. Partway there, >>> the engine would start racing if I went to fast. hmm. Her father >>> looked at it, drove 10 feet, and told me what the problem was. > >>> It was an interesting drive back, as the maximum speed kept dropping, >>> eventually down to 53 . . . > >>You probably needed another new clutch, after all that slipping. >>That's why you always keep adequate free play in the clutch >>pedal. You had a very sloppy mechanic. > >Yep. I showed up on monday, and said that they needed to redo my >clutch. > The lady snarled, "how many miles does it have on it?" > "400." > "Oh." Looked like they expected you back, but not so soon: must have overtightened the cable too much. > It wasn't until later that I drew the connection between going in for >service there and the new problems that arose within a day or two. >plugging that loose wire (a factory test wire?) into the fuse block was >the real joy. We'd traced the entire electrical system and lifted the >upholstery trying to find the short that caused the fuse to blow when >the taillights went on. Suddenly inspiration hit: what if I remove >this wire, which seems to have no purpose? Amazing how much better >things work when you don't connect +12 to GND . . . Looked like they were doing their best to ensure you returned to give them more business. "Never attribute to malice what can be adequately explained by stupidity." A couple of problems like that is a definite indication that one, the other, or both, is occurrring. Hope you never went back there again, and spread the word. Once went into a brake place to make an appointment, heard the boss tell a (presumably new) mechanic to just replace the drums or rotors or whatever, "'coz that's how we do things here". Said I'd be back later and left to spread the word. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: From: Michael T Pins User-Agent: nn/6.6.5 Date: 19 Mar 2003 20:12:33 GMT Lines: 40 Message-ID: <3e78cf31$0$83864$a1866201@newsreader.visi.com> Organization: VISI.com NNTP-Posting-Date: 19 Mar 2003 14:12:33 CST NNTP-Posting-Host: 09d96b73.newsreader.visi.com X-Trace: DXC=4Z6m;R=^Bm[^O=8[8PP:QQOG;lf4n<5gZhD3OXm`l_bXF writes: >In article , jmfbahciv@aol.com wrote: >> In article , >> Charles Shannon Hendrix wrote: >>>In article , jmfbahciv@aol.com wrote: >>> >>>> That command truly created superstition. There were >>>> days that it sure felt like a ^T was necessary in order for >>>> the machine to pay you some attention. Even those who knew >>>> the code intimately would ^T to get attention. >>> >>>Not much different than: >>> >>>% sync >>>% sync >>>% sync >>> >>>....on a UNIX machine, just in case that first sync missed something. >> >> It would. Think about getting a car into first gear while > still >> moving. >I don't follow you. >The extra calls to sync don't do anything for you. Actually, they do. On some OSs, sync will return before it's actually done sync'ing. If you then halt the machine you can have a problem. In theory, the second sync won't return until the first one is finished, but whether this is true depends upon who wrote sync. The real effect of typing it three times is that hopefully the first one will finish before you're done typing the third one. -- ************************************************************************** * Michael T Pins | mtpins@nndev.org * * keeper of the nn sources | mtpins@visi.com * * ftp://ftp.nndev.org/pub | #include * ###### Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E6E0E58.E976BA50@yahoo.com> From: Michael T Pins User-Agent: nn/6.6.5 Date: 19 Mar 2003 20:16:23 GMT Lines: 24 Message-ID: <3e78d017$0$83864$a1866201@newsreader.visi.com> Organization: VISI.com NNTP-Posting-Date: 19 Mar 2003 14:16:23 CST NNTP-Posting-Host: 09d96b73.newsreader.visi.com X-Trace: DXC=EFI:2mn6OT`XCS:4_nQhXeOG;lf4n<5gjhD3OXm`l_bhF writes: >In article <3E6E0E58.E976BA50@yahoo.com>, CBFalconer wrote: >> When you have lost your clutch cable, it is necessary. You start >Like in a VW Beetle when the clutch cable goes. Fun, trying to get all >the way home, and fudging a little at yellow lights and stop signs... :) >You can't shift out of the gear, and if in town you just hope you are >in second gear. Sure you can. The bug had a good syncromesh tranny. As long as the car was moving at a minimum speed you could shift between gears without using the clutch simply by shifting to neutral, using the gas pedal to adjust the engine speed, then shifting into the correct gear. I used to do it just to freak out passengers. The real problem with driving without a clutch is that if you stop you can't get back into gear. -- ************************************************************************** * Michael T Pins | mtpins@nndev.org * * keeper of the nn sources | mtpins@visi.com * * ftp://ftp.nndev.org/pub | #include * ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Thu, 20 Mar 03 12:35:19 GMT Organization: UltraNet Communications, Inc. Lines: 73 Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <3E791A67.541A3D9D@ev1.net> <3E790C5F.9060304@jetnet.ab.ca> <3E7983CB.F8B21B73@ev1.net> <162j7vgpe89602fibjrt2ao7odohimpt8l@4ax.com> X-Trace: UmFuZG9tSVYjWZQSIyO5TfWg+/BRjdwnfoFBJPyvLi2j3wqx5Ho1pzV/KWXWMTFW X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 20 Mar 2003 13:22:34 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!nntp.abs.net!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-209 Xref: chonsp.franklin.ch alt.folklore.computers:132783 In article <162j7vgpe89602fibjrt2ao7odohimpt8l@4ax.com>, Brian Inglis wrote: >On Thu, 20 Mar 2003 07:05:01 GMT in alt.folklore.computers, >Charles Richmond wrote: > >>Brian Inglis wrote: >>> >>> >Charles Richmond wrote: >>> > >>> >> I am *not* advocating the wide use of bubble sort. I am >>> >> only saying that there is a useful place for bubble sort. >>> >>> Only if you can guarantee that you are sorting ten items or less >>> in almost sorted order: otherwise your choice is regrettable, and >>> likely to be regretted by your users at some point. >>> >>I worked on a program where I *could* guarantee that there would >>be a dozen items or fewer. And I *did* use bubble sort. The >>program had to sort products by product code on a bill of lading. >>My users were so technologically ignorant that all they could >>regret or care about...would be the program failing to work all >>together. > >You haven't met the "Bastard Testers from Hell (TM)"! ROTFLMAO....ACMT (and cleaning my terminal). >The guys I worked with were extremely sick puppies and really >*enjoyed* testing everything to destruction and finding subtle >bugs: they *hated* programmers who provided low hanging fruit as >that was too easy and didn't challenge them. >The new corporate email filters only lasted a few hours before >they could tell us what the holes were, as it happened to be >upgraded during a lull in testing. I would hire them in a nanosecond. >Your code would have been fed a gigantic bunch of different >products all with the same product code, and you'd have a defect >report in your incoming mail, regardless of how your code handled >it. > >I kept them working hard by creating pessimally bad sets of data >in the development and unit test environment so I knew my code >would keep working sensibly regardless of the combinations of >junk thrown at it. >I was able to mock them into finding only three minor defects in >my code during three months of testing: they found one each in >the first month. >A number of my warnings were reclassified as fatal errors before >release, as the business did not want processing to continue when >the data was that bad. > >We were all rewarded by a system implementation that went >flawlessly, with zero internal or external support calls or >defect reports (and a few nice end of project gifts), thanks to >similar diligence by others developing and testing the >implementation, business process testing, and internal and >external training. > >The heavy testing is required because a mistake by one user one >day cost their company a megabuck that day and another company >profited at their expense by at least that amount, probably a >multiple, and the customer support staff had to be able to prove >that the system acted correctly on the user's instructions and >fairly within the contract terms in dealing with the resulting >effects. That must have been a fun project with enormous satisfaction. /BAH Subtract a hundred and four for e-mail. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC References: <3E56AB27.162B9193@yahoo.com> <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com> <1frznw0.8e1waqbp4y8N%lars@bearnip.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 76 Date: Wed, 19 Mar 2003 18:23:51 -0500 NNTP-Posting-Host: 209.96.189.99 X-Trace: sydney.visi.net 1048119916 209.96.189.99 (Wed, 19 Mar 2003 19:25:16 EST) NNTP-Posting-Date: Wed, 19 Mar 2003 19:25:16 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!news.webusenet.com!newsfeed-east.nntpserver.com!nntpserver.com!news-out.visi.com!petbe.visi.com!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:132864 In article <1frznw0.8e1waqbp4y8N%lars@bearnip.com>, Lars Duening wrote: >> > On the other hand, C++ supports multi-paradigmatic programming because >> >> Because what? :) > > Because the implementors have the same opinion as you do: Sorry, right over my head... :) >> >> At the same time, I think the idea that all projects are expressed best >> >> in OO fashion is ridiculous. > >> > It's by far not a perfect language (especially the C heritage), but it's >> > one of the lesser evil ones (at least you can shoot your own foot with >> > it!) >> >> I think you meant "can't". > > No, I did mean 'can', as it indicates usefulness for real programming > tasks. Right, I usually hate languages that have foot-safety in mind. >> That was the difference between code generation, and the difference >> has been acceptable. > > Code generation and libraries. Most novice C programmers used printf(), > and back then printf() tended to pull in all the floating point > libraries as well. I am guilty! Actually, I noticed it was expensive, and wrote my own. I had a really rigid format though, which wasn't near as nice. However, it was small and fast. A lot of people rewrite printf() and routines like that as practice when learning C, or at least I've seen it myself. > C++ implementations have the same problem, one magnitude larger, as the > standard library is rather complex. It seems to have a lot more dependency issues, probably because of the code sharing it encourages. I would have felt better if it was a series of independent classes with interfaces between them, so you could use them seperately. > Templates, for example. A smart build system would generate the code for > required templates once and then use that wherever the same template is > required. With the current build model however the compiler has to > re-create the template code over and over for the different source > files, hoping that the linker throws out the duplicates. Ah, I see where you are coming from. It seems to me that this can't happen. Doesn't this method of generating template code cause name mangling? It would seem the linker would see each duplicate as unique and this leave it in. It seems to me the smart build system you describe would be the only way to really do this right. > Some compilers (like g++) improve over this behaviour, but you have to > drop heavy non-portable hints into the source for that. You have to be careful with GNU C. It's tempting to use certain things which will bite you later. > Well, Java enables me to write Windows programs without touching a > Windows box except for testing purposes. With a cross-platform C/C++ > environment I would at least have to build the program on Windows - with > all the pain that brings with it. Well... that's a very good point. ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 29 Date: Wed, 19 Mar 2003 18:31:54 -0500 NNTP-Posting-Host: 209.96.189.99 X-Trace: sydney.visi.net 1048119917 209.96.189.99 (Wed, 19 Mar 2003 19:25:17 EST) NNTP-Posting-Date: Wed, 19 Mar 2003 19:25:17 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!eusc.inter.net!priapus.visi.com!news-out.visi.com!petbe.visi.com!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:132730 In article , Dr. Richard E. Hawkins wrote: > In article , > Charles Shannon Hendrix wrote: >>In article <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com>, Lars Duening wrote: > >>> True. Truth is running code, and I have often nixed a nifty algorithm >>> simply because it didn't improve anything. > >>In fact, you can often choose a better O(...) algorithm only to find >>it is worse. > > Bubble sorts for small lists! Yay! :) Well, it can be more than that. For example, a better O(n) algorithm might be greatly reducing data copies, but might also introduce another kind of code which looks good, O(n) wise. However, on a CPU that is blindlingly fast doing the copies, you might have just slowed the machine down. > Actually, the most recent model I was working on had a screwball, > hand-written sort for groups of 4, as I only needed to distinguish the > 1st & 2nd hightest. An earlier version didn't even distinguish that, > but just the top two and bottom two. I've done stuff like that, and people often don't like finding it. I'll also bubble sort or do no sort at all if its not necessary, but some people really don't like it. ###### Message-ID: <3E791A67.541A3D9D@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 33 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc53 1048116921 12.241.15.59 (Wed, 19 Mar 2003 23:35:21 GMT) NNTP-Posting-Date: Wed, 19 Mar 2003 23:35:21 GMT Date: Wed, 19 Mar 2003 23:35:21 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!news.webusenet.com!cyclone1.gnilink.net!wn14feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!rwcrnsc53.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132841 Brian Inglis wrote: > > On Tue, 18 Mar 2003 15:19:28 +0000 (UTC) in > alt.folklore.computers, hawk@slytherin.ds.psu.edu (Dr. Richard E. > Hawkins) wrote: > > > > > [snip...] [snip...] [snip...] > > > > Bubble sorts for small lists! Yay! :) > > There's never an excuse other than mental laziness for a bubble > sort. *ALL* other sorts are faster! > Au contraire...the bubble sort is faster than a quicksort if you only have to sort 10 items. Also, you have 5000 items that are *almost* sorted already, bubble sort can do better than quicksort. Quicksort has a worst-case performance on "almost sorted" input data. Almost sorted is a real aid for bubble sort. Also, a bubble sort is *so* easy to implement...it is less likely to have a bug in it. I am *not* advocating the wide use of bubble sort. I am only saying that there is a useful place for bubble sort. Additional bonus: bubble sort is a stable sort, quicksort is *not*. -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### From: bfranchuk@jetnet.ab.ca Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Wed, 19 Mar 2003 17:33:35 -0700 Organization: Posted via Supernews, http://www.supernews.com Message-ID: <3E790C5F.9060304@jetnet.ab.ca> User-Agent: Mozilla/5.0 (X11; U; FreeBSD i386; en-US; rv:1.0.1) Gecko/20021005 X-Accept-Language: en-us, en MIME-Version: 1.0 References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <3E791A67.541A3D9D@ev1.net> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Complaints-To: abuse@supernews.com Lines: 14 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!sn-xit-03!sn-xit-06!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132750 Charles Richmond wrote: > I am *not* advocating the wide use of bubble sort. I am > only saying that there is a useful place for bubble sort. > Additional bonus: bubble sort is a stable sort, quicksort > is *not*. You really have to know the problem too at had. I really don't think many people have a Basic version of quick sort just lieing around :) ###### Message-ID: <3E792CCD.EF04D3E2@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <3E78C838.15934485@thinkage.ca> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 31 Date: Thu, 20 Mar 2003 03:44:39 GMT NNTP-Posting-Host: 12.90.167.217 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc04-news.ops.worldnet.att.net 1048131879 12.90.167.217 (Thu, 20 Mar 2003 03:44:39 GMT) NNTP-Posting-Date: Thu, 20 Mar 2003 03:44:39 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!nntp.theplanet.net!inewsm1.nntp.theplanet.net!195.40.4.120.MISMATCH!easynet-quince!easynet.net!cox.net!cyclone1.gnilink.net!wn14feed!wn13feed!worldnet.att.net!bgtnsc04-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132731 "Alan T. Bowler" wrote: > Brian Inglis wrote: > > > >Bubble sorts for small lists! Yay! :) > > > > There's never an excuse other than mental laziness for a bubble > > sort. *ALL* other sorts are faster! > > Not really. On very short stuff the low setup of the bubble sorts > (I'm including "insertion sort" here), will win out. However, > the list last to very short. Shellsort is only just a tad more > overhead and wins somewhere between 5 and 20 items. In cases, > where there is a high probability that the data is already sorted > a bubbble sort can be a clear winner even for longer lists. According to ME, the optimum GP sort is mergesort. It handles arbitrary length lists, and you never have to set up funny arrays of pointers etc. Just gobble in the data from anywhere and stuff it into a list. Sort. Maybe even sort again on another field (mergesort is stable). Dump. Done. All I have to design is one stinking little comparison function. All I need to do is build the data records with an unused 'next' field. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <3E791A67.541A3D9D@ev1.net> <3E790C5F.9060304@jetnet.ab.ca> X-Newsreader: Forte Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 47 Date: Thu, 20 Mar 2003 06:10:10 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news1.calgary.shaw.ca 1048140610 24.71.223.147 (Wed, 19 Mar 2003 23:10:10 MST) NNTP-Posting-Date: Wed, 19 Mar 2003 23:10:10 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-fra1.dfn.de!news.f.de.plusline.net!feed.news.nacamar.de!easynet-quince!easynet.net!cox.net!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news1.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132720 On Wed, 19 Mar 2003 17:33:35 -0700 in alt.folklore.computers, bfranchuk@jetnet.ab.ca wrote: >Charles Richmond wrote: > >> I am *not* advocating the wide use of bubble sort. I am >> only saying that there is a useful place for bubble sort. Only if you can guarantee that you are sorting ten items or less in almost sorted order: otherwise your choice is regrettable, and likely to be regretted by your users at some point. >> Additional bonus: bubble sort is a stable sort, quicksort >> is *not*. What's the big deal with stable sorts? Unstable sorts can be made stable by adding the element address, offset, or index to the comparison. >You really have to know the problem too at had. I really >don't think many people have a Basic version of quick sort >just lieing around :) I used Shell sort when I needed a Q&D general internal sort, in Basic or any language without one available, and couldn't avoid it by changing the processing; in C I just use qsort(3) if I really can't avoid sorting by using different data structures; better than "creating beautiful new impediments to understanding" until I find a real need for something faster. I find I only need an internal sort when I need to process the same set of data structures in two different orders, the probability of needing the alternate order is relatively low (estimated use < 50% of cases), or the overhead of maintaining another set of pointers for the alternate order is relatively high (estimated average length of second list traversal for insertion > 50%, or comparison function for alternate order is very expensive compared with primary order), or often both. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### Message-ID: <3E7983CB.F8B21B73@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <3E791A67.541A3D9D@ev1.net> <3E790C5F.9060304@jetnet.ab.ca> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 22 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: sccrnsc02 1048143901 12.241.15.59 (Thu, 20 Mar 2003 07:05:01 GMT) NNTP-Posting-Date: Thu, 20 Mar 2003 07:05:01 GMT Date: Thu, 20 Mar 2003 07:05:01 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newsfeed00.sul.t-online.de!t-online.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!wn14feed!wn13feed!wn12feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!sccrnsc02.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132847 Brian Inglis wrote: > > >Charles Richmond wrote: > > > >> I am *not* advocating the wide use of bubble sort. I am > >> only saying that there is a useful place for bubble sort. > > Only if you can guarantee that you are sorting ten items or less > in almost sorted order: otherwise your choice is regrettable, and > likely to be regretted by your users at some point. > I worked on a program where I *could* guarantee that there would be a dozen items or fewer. And I *did* use bubble sort. The program had to sort products by product code on a bill of lading. My users were so technologically ignorant that all they could regret or care about...would be the program failing to work all together. -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: <162j7vgpe89602fibjrt2ao7odohimpt8l@4ax.com> References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <3E791A67.541A3D9D@ev1.net> <3E790C5F.9060304@jetnet.ab.ca> <3E7983CB.F8B21B73@ev1.net> X-Newsreader: Forte Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 68 Date: Thu, 20 Mar 2003 10:02:41 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news1.calgary.shaw.ca 1048154561 24.71.223.147 (Thu, 20 Mar 2003 03:02:41 MST) NNTP-Posting-Date: Thu, 20 Mar 2003 03:02:41 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!uio.no!feed.news.nacamar.de!newsfeed01.sul.t-online.de!t-online.de!newspeer1-gui.server.ntli.net!ntli.net!cox.net!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news1.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132838 On Thu, 20 Mar 2003 07:05:01 GMT in alt.folklore.computers, Charles Richmond wrote: >Brian Inglis wrote: >> >> >Charles Richmond wrote: >> > >> >> I am *not* advocating the wide use of bubble sort. I am >> >> only saying that there is a useful place for bubble sort. >> >> Only if you can guarantee that you are sorting ten items or less >> in almost sorted order: otherwise your choice is regrettable, and >> likely to be regretted by your users at some point. >> >I worked on a program where I *could* guarantee that there would >be a dozen items or fewer. And I *did* use bubble sort. The >program had to sort products by product code on a bill of lading. >My users were so technologically ignorant that all they could >regret or care about...would be the program failing to work all >together. You haven't met the "Bastard Testers from Hell (TM)"! The guys I worked with were extremely sick puppies and really *enjoyed* testing everything to destruction and finding subtle bugs: they *hated* programmers who provided low hanging fruit as that was too easy and didn't challenge them. The new corporate email filters only lasted a few hours before they could tell us what the holes were, as it happened to be upgraded during a lull in testing. Your code would have been fed a gigantic bunch of different products all with the same product code, and you'd have a defect report in your incoming mail, regardless of how your code handled it. I kept them working hard by creating pessimally bad sets of data in the development and unit test environment so I knew my code would keep working sensibly regardless of the combinations of junk thrown at it. I was able to mock them into finding only three minor defects in my code during three months of testing: they found one each in the first month. A number of my warnings were reclassified as fatal errors before release, as the business did not want processing to continue when the data was that bad. We were all rewarded by a system implementation that went flawlessly, with zero internal or external support calls or defect reports (and a few nice end of project gifts), thanks to similar diligence by others developing and testing the implementation, business process testing, and internal and external training. The heavy testing is required because a mistake by one user one day cost their company a megabuck that day and another company profited at their expense by at least that amount, probably a multiple, and the customer support staff had to be able to prove that the system acted correctly on the user's instructions and fairly within the contract terms in dealing with the resulting effects. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <3E791A67.541A3D9D@ev1.net> From: Morten Reistad X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: mrr@acer.reistad.priv.no (Morten Reistad) Message-ID: Lines: 44 Date: Thu, 20 Mar 2003 11:06:36 +0100 NNTP-Posting-Host: 80.111.42.204 X-Complaints-To: abuse@chello.no X-Trace: amstwist00 1048158001 80.111.42.204 (Thu, 20 Mar 2003 12:00:01 MET) NNTP-Posting-Date: Thu, 20 Mar 2003 12:00:01 MET Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!news.tele.dk!news.tele.dk!small.news.tele.dk!proxad.net!proxad.net!news-hub.cableinet.net!blueyonder!amsnews01.chello.com!amstwist00.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132813 According to Charles Richmond : >Brian Inglis wrote: >> >> On Tue, 18 Mar 2003 15:19:28 +0000 (UTC) in >> alt.folklore.computers, hawk@slytherin.ds.psu.edu (Dr. Richard E. >> Hawkins) wrote: >> >> > >> > [snip...] [snip...] [snip...] >> > >> > Bubble sorts for small lists! Yay! :) >> >> There's never an excuse other than mental laziness for a bubble >> sort. *ALL* other sorts are faster! >> >Au contraire...the bubble sort is faster than a quicksort >if you only have to sort 10 items. Also, you have 5000 >items that are *almost* sorted already, bubble sort can >do better than quicksort. Quicksort has a worst-case >performance on "almost sorted" input data. Almost sorted >is a real aid for bubble sort. Also, a bubble sort works in a linear fashion, and will work well with primary and secondary cache on modern processors. Shellsort and quicksort will not work well with a cache. "Almost sorted, just need to insert a few items" is a very common situation. If you give the bubblesort a little stack of out-of-sequence (but the stack is sorted) it will stick to it's cache-friendly behaviour but not take more than a single pass for nonsorted items. On stack overflow just bail out to another sort. >Also, a bubble sort is *so* easy to implement...it is less >likely to have a bug in it. > >I am *not* advocating the wide use of bubble sort. I am >only saying that there is a useful place for bubble sort. >Additional bonus: bubble sort is a stable sort, quicksort >is *not*. This strategy is more "select-the-sort" than using a bubble-sort. -- mrr ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 20 Mar 03 12:40:32 GMT Organization: UltraNet Communications, Inc. Lines: 45 Message-ID: References: <3e78cf31$0$83864$a1866201@newsreader.visi.com> X-Trace: UmFuZG9tSVaPsMgN0wDPdiEiutXi5H1tie+uM2w2QEJwyN+P6AzxbKtW5vrSBrbI X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 20 Mar 2003 13:27:46 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-209 Xref: chonsp.franklin.ch alt.folklore.computers:132784 In article <3e78cf31$0$83864$a1866201@newsreader.visi.com>, Michael T Pins wrote: >Charles Shannon Hendrix writes: > >>In article , jmfbahciv@aol.com wrote: >>> In article , >>> Charles Shannon Hendrix wrote: >>>>In article , jmfbahciv@aol.com wrote: >>>> >>>>> That command truly created superstition. There were >>>>> days that it sure felt like a ^T was necessary in order for >>>>> the machine to pay you some attention. Even those who knew >>>>> the code intimately would ^T to get attention. >>>> >>>>Not much different than: >>>> >>>>% sync >>>>% sync >>>>% sync >>>> >>>>....on a UNIX machine, just in case that first sync missed something. >>> >>> It would. Think about getting a car into first gear while > still >>> moving. > >>I don't follow you. > >>The extra calls to sync don't do anything for you. > >Actually, they do. >On some OSs, sync will return before it's actually done sync'ing. If you >then halt the machine you can have a problem. In theory, the second sync >won't return until the first one is finished, but whether this is true >depends upon who wrote sync. The real effect of typing it three times is >that hopefully the first one will finish before you're done typing the >third one. > Exactly :-). Knocking wood three times would do the same thing (after the first sync is typed) but no one had wood in their offices. I had wood in my office but I had to pull the dead body threat to keep it. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 20 Mar 03 12:42:00 GMT Organization: UltraNet Communications, Inc. Lines: 23 Message-ID: References: <3E6E0E58.E976BA50@yahoo.com> X-Trace: UmFuZG9tSVbVGgd5mtqnQUqyeOseSlS2dklL3ck0rbk6ruyXcw5H+wDDlw+P0i6m X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 20 Mar 2003 13:29:13 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!207-172-216-209 Xref: chonsp.franklin.ch alt.folklore.computers:132774 In article , Charles Shannon Hendrix wrote: >In article , Dr. Richard E. Hawkins wrote: > >> Or when it's too tight. My girlfriend and I drove 80 miles the day >> after my clutch was replaced, off to meet her parents. Partway there, >> the engine would start racing if I went to fast. hmm. Her father >> looked at it, drove 10 feet, and told me what the problem was. > >Heh heh... > >I had a slipping clutch in a truck I used to have, a 1994 Toyota. On a >trip to Washington DC, I had to use I95 North. It's a very fast stretch >of road, where it isn't uncommon for the average speed to be 80mph, and >there are some going well over 100mph at times. They are the nuts. Whoa, Nellie! Am I glad I was driving south to D.C. /BAH Subtract a hundred and four for e-mail. ###### Message-ID: <3E791B1B.79601FDD@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 14 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc53 1048117100 12.241.15.59 (Wed, 19 Mar 2003 23:38:20 GMT) NNTP-Posting-Date: Wed, 19 Mar 2003 23:38:20 GMT Date: Wed, 19 Mar 2003 23:38:20 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!newsfeed.arcor-online.net!newsfeed.freenet.de!194.168.222.21.MISMATCH!newspeer1-gui.server.ntli.net!ntli.net!cox.net!cyclone1.gnilink.net!wn14feed!wn13feed!wn12feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!rwcrnsc53.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132846 "Dr. Richard E. Hawkins" wrote: > > [snip...] [snip...] [snip...] > > "There is no lower limit to human intelligence." --me > "Only two things are infinite, the universe and human stupidity, and I'm not sure about the former." -- Albert Einstein -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E6E0E58.E976BA50@yahoo.com> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 27 Date: Wed, 19 Mar 2003 22:35:35 -0500 NNTP-Posting-Host: 209.96.185.72 X-Trace: sydney.visi.net 1048134303 209.96.185.72 (Wed, 19 Mar 2003 23:25:03 EST) NNTP-Posting-Date: Wed, 19 Mar 2003 23:25:03 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!newsfeed.gamma.ru!Gamma.RU!feed2.news.rcn.net!rcn!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:132723 In article , Dr. Richard E. Hawkins wrote: > Or when it's too tight. My girlfriend and I drove 80 miles the day > after my clutch was replaced, off to meet her parents. Partway there, > the engine would start racing if I went to fast. hmm. Her father > looked at it, drove 10 feet, and told me what the problem was. Heh heh... I had a slipping clutch in a truck I used to have, a 1994 Toyota. On a trip to Washington DC, I had to use I95 North. It's a very fast stretch of road, where it isn't uncommon for the average speed to be 80mph, and there are some going well over 100mph at times. They are the nuts. I would hover around 80mph (people were passing me) until I could smell the clutch. Back down to 65 to cool it off, and then back again. I tried to drive slow enough to keep the clutch cool, but when I tried to hold it on 65mph, I nearly got killed. I95 is nuts sometimes. > It was an interesting drive back, as the maximum speed kept dropping, > eventually down to 53 . . . Have you ever driven a Chevy Chevette in the mountains? If you haven't, don't try. ###### Message-ID: <3E7932E9.DD3B7F19@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E776132.2B2A2AAD@yahoo.com> <2qah7vc7dfm559m3hdpe6o0m0kmpve2ajr@4ax.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 35 Date: Thu, 20 Mar 2003 03:44:41 GMT NNTP-Posting-Host: 12.90.167.217 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc04-news.ops.worldnet.att.net 1048131881 12.90.167.217 (Thu, 20 Mar 2003 03:44:41 GMT) NNTP-Posting-Date: Thu, 20 Mar 2003 03:44:41 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news.linkpendium.com!nntp-relay.ihug.net!ihug.co.nz!west.cox.net!cox.net!cyclone1.gnilink.net!wn14feed!wn13feed!worldnet.att.net!bgtnsc04-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132721 Brian Inglis wrote: > ... snip ... > > Once went into a brake place to make an appointment, heard the > boss tell a (presumably new) mechanic to just replace the drums > or rotors or whatever, "'coz that's how we do things here". Said > I'd be back later and left to spread the word. OTOH a few years ago I had a problem with something suddenly snaffling all the fluid out of my automatic transmission. Various people scratched their heads and made dire predictions (the car was over 20 years old) with numbers in the $500 to $1000 range. Then I tried this local transmission shop. He said "It's the whozits filter - costs about $10". He did the job in 15 minutes. I've never had the occasion to go back, but I must have sent him at least 20 other customers. Hamden Transmission, for those in the CT. area. However the whozits had let fluid into the engine to be burnt smokily, and that took the catalytic converter, which blocked. Curing that is another story. I got another 50,000 miles out of it, but then the roof and floor both rusted through. I replaced it with a relatively new '88 VW with 150,000 miles on it, that needed a clutch cable, front tires, alignment and a rear window wiper motor, for $1000, and I've put another 30 or 40000 miles on that since. Oh yes, I had to get an exhaust system, otherwise virtually maintenance free. I think I need wiper blades now. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### Message-ID: <3E798549.AFC1275D@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E776132.2B2A2AAD@yahoo.com> <2qah7vc7dfm559m3hdpe6o0m0kmpve2ajr@4ax.com> <3E7932E9.DD3B7F19@yahoo.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 47 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: sccrnsc02 1048144283 12.241.15.59 (Thu, 20 Mar 2003 07:11:23 GMT) NNTP-Posting-Date: Thu, 20 Mar 2003 07:11:23 GMT Date: Thu, 20 Mar 2003 07:11:23 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!wn14feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!sccrnsc02.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132844 CBFalconer wrote: > > Brian Inglis wrote: > > > ... snip ... > > > > Once went into a brake place to make an appointment, heard the > > boss tell a (presumably new) mechanic to just replace the drums > > or rotors or whatever, "'coz that's how we do things here". Said > > I'd be back later and left to spread the word. > > OTOH a few years ago I had a problem with something suddenly > snaffling all the fluid out of my automatic transmission. Various > people scratched their heads and made dire predictions (the car > was over 20 years old) with numbers in the $500 to $1000 range. > Then I tried this local transmission shop. He said "It's the > whozits filter - costs about $10". He did the job in 15 minutes. > I've never had the occasion to go back, but I must have sent him > at least 20 other customers. Hamden Transmission, for those in > the CT. area. > A female friend of my owned a 1970 model Ford Pinto Pony when she lived in the Denver Colorado area. In the mid 1970's she decided to go to Louisiana to graduate school. So she had the car tuned up and whipped into shape. On the road, the engine kept overheating. She had a new water pump put on somewhere in Kansas. When she stopped by the Dallas area where I lived...she was still having trouble. She went to this garage, and they put on a new waterpump...still did *not* fix it. The mechanics were debating about what drastic measures to take...rod out the water traces in the engine block, or even have her buy a rebuilt engine. Then an off-duty mechanic showed up for work. He took one look at it and figured out that whoever had overhauled the car in Colorado...had put the radiator fan on *backwards*. This reduced the air flow through the radiator to nil, and caused all the trouble. The fan was turned around and the trouble was fixed. Amazing how *stupid* mechanics can cost one a fortune for *no* damn good reason!!! -- +-------------------------------------------------------------+ | Charles and Francis Richmond | +-------------------------------------------------------------+ ###### Date: Thu, 20 Mar 2003 08:13:48 +0100 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <20030320081348.33ce785b.steveo@eircom.net> References: <3E6E0E58.E976BA50@yahoo.com> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 13 Organization: EuroNet Internet NNTP-Posting-Date: 20 Mar 2003 18:00:05 GMT NNTP-Posting-Host: i2020.vwr.wanadoo.nl X-Trace: 1048183205 maya.euronet.nl 139 194.134.215.236:1224 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!uio.no!feed.news.nacamar.de!newsfeed.icl.net!newsfeed.fjserv.net!kibo.news.demon.net!demon!news2.euro.net!postnews1.euro.net!maya.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132808 On 18 Mar 2003 22:19:42 -0800 Dowe Keller wrote: DK> Kind of like that $10 transistor burning out to save the $0.10 fuse. Isn't that what the transistor is there for, protectnig fuses is hard so it takes an expensive component. -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### From: Andreas Eder Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: 20 Mar 2003 21:56:40 +0100 Organization: T-Online Lines: 11 Sender: are@elgin.eder.de Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Trace: news.t-online.com 1048194003 04 12545 XbOsESEbSZAQye 030320 21:00:03 X-Complaints-To: abuse@t-online.com X-ID: Ta++hsZU8edycjpXtdjJSMYa3s1L7uHbaACsjswzsS5hPOBokwd56k User-Agent: Gnus/5.0808 (Gnus v5.8.8) XEmacs/21.4 (Portable Code) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newsmi-eu.news.garr.it!NewsITBone-GARR!feed.news.nacamar.de!newsfeed01.sul.t-online.de!newsmm00.sul.t-online.com!t-online.de!news.t-online.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132873 Brian Inglis writes: > There's never an excuse other than mental laziness for a bubble > sort. *ALL* other sorts are faster! Well, almost all. How about generating all permutations of the data and then look for the sorted one? 'Andreas -- Wherever I lay my .emacs, there´s my $HOME. ###### From: ararghNOSPAM@NOT.AT.enteract.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Thu, 20 Mar 2003 15:50:30 -0600 Organization: Not Really! Lines: 17 Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> NNTP-Posting-Host: tcr201.dynip.ripco.net Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: e250.ripco.com 1048197070 17120 209.100.226.201 (20 Mar 2003 21:51:10 GMT) X-Complaints-To: usenet@ripco.com NNTP-Posting-Date: Thu, 20 Mar 2003 21:51:10 +0000 (UTC) X-Newsreader: Forte Agent 1.92/32.572 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newsmi-eu.news.garr.it!NewsITBone-GARR!fu-berlin.de!gail.ripco.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132884 On 20 Mar 2003 21:56:40 +0100, Andreas Eder wrote: >Brian Inglis writes: > >> There's never an excuse other than mental laziness for a bubble >> sort. *ALL* other sorts are faster! > >Well, almost all. How about generating all permutations of the data >and then look for the sorted one? Isn't that the BOGO sort? -- Arargh (at arargh dot com) http://www.arargh.com To reply by email, change the domain name, and remove the garbage. (Enteract can keep the spam, they are gone anyway) ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Fri, 21 Mar 03 12:48:13 GMT Organization: UltraNet Communications, Inc. Lines: 22 Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <20030321064101.3e03896c.steveo@eircom.net> X-Trace: UmFuZG9tSVYc+/w2MIVYm1Dokn63kbAdzKDJ+6Un8eobJvL6GirRziOxKAzKq1Pj X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 21 Mar 2003 13:35:37 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-182-114 Xref: chonsp.franklin.ch alt.folklore.computers:132963 In article <20030321064101.3e03896c.steveo@eircom.net>, Steve O'Hara-Smith wrote: >On 20 Mar 2003 21:56:40 +0100 >Andreas Eder wrote: > >AE> Brian Inglis writes: >AE> >AE> > There's never an excuse other than mental laziness for a bubble >AE> > sort. *ALL* other sorts are faster! >AE> >AE> Well, almost all. How about generating all permutations of the data >AE> and then look for the sorted one? > > The slowest one is to write the data into memory and wait for >random bit flips to do the job. > When you figure out how to get Murphy to do your work for you, you'll be able to buy Misfot and fix it. /BAH Subtract a hundred and four for e-mail. ###### From: ararghNOSPAM@NOT.AT.enteract.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Thu, 20 Mar 2003 18:11:39 -0600 Organization: Not Really! Lines: 55 Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> NNTP-Posting-Host: tcr201.dynip.ripco.net Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: e250.ripco.com 1048205541 22156 209.100.226.201 (21 Mar 2003 00:12:21 GMT) X-Complaints-To: usenet@ripco.com NNTP-Posting-Date: Fri, 21 Mar 2003 00:12:21 +0000 (UTC) X-Newsreader: Forte Agent 1.92/32.572 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed.uncensored-news.com!propagator-sanjose!news-in-sanjose!in.nntp.be!gail.ripco.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133160 On 20 Mar 2003 22:24:09 GMT, cb@df.lth.se (Christian Brunschen) wrote: >In article , > wrote: >>On 20 Mar 2003 21:56:40 +0100, Andreas Eder >>wrote: >> >>>Well, almost all. How about generating all permutations of the data >>>and then look for the sorted one? >> >>Isn't that the BOGO sort? > >The description of bogosort that I recall was something akin to the >following: > > while (the array is not sorted) { > swap two random elements of the array > } > >somthing like (just typed into this message, not tested, so probably full >of bugs) > > #include > #include > static int random_int(int max) { > double tmp = (double)random() / (double)(LONG_MAX + 1); > return floor(max * tmp); > } > static int array_sorted(int array[], int n) { > int i = 1; > while (i < n && array[i-1] <= array[i]) > ; > return (i >= n); > } > void bogosort(int array[], int n) { > while (!array_sorted(array, n)) { > int i = random_int(n); > int j = random_int(n); > int tmp = array[i]; > array[i] = array[j]; > array[j] = tmp; > } > } > >The beauty of this version of bogosort is that it has no guaranteed upper >bound to its execution time - it could in theory mever terminate at all, >if the random number generator never hits on the right combination of >elements to exchange! :) Neither method sounds like one that I would ever use. :-) -- Arargh (at arargh dot com) http://www.arargh.com To reply by email, change the domain name, and remove the garbage. (Enteract can keep the spam, they are gone anyway) ###### Message-ID: <3E7A6B2C.3080608@boutel.co.nz> From: Brian Boutel User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.0.1) Gecko/20020830 X-Accept-Language: en-us, en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Lines: 49 Date: Fri, 21 Mar 2003 13:30:20 +1200 NNTP-Posting-Host: 203.96.144.148 X-Complaints-To: abuse@tsnz.net X-Trace: news02.tsnz.net 1048210220 203.96.144.148 (Fri, 21 Mar 2003 13:30:20 NZST) NNTP-Posting-Date: Fri, 21 Mar 2003 13:30:20 NZST Organization: TelstraClear Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.stanford.edu!logbridge.uoregon.edu!arclight.uoregon.edu!enews.sgi.com!news-out.spamkiller.net!propagator2-maxim!news-in.spamkiller.net!news02.tsnz.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132910 ararghNOSPAM@NOT.AT.enteract.com wrote: > On 20 Mar 2003 21:56:40 +0100, Andreas Eder > wrote: > > >>Brian Inglis writes: >> >> >>>There's never an excuse other than mental laziness for a bubble >>>sort. *ALL* other sorts are faster! >> >>Well, almost all. How about generating all permutations of the data >>and then look for the sorted one? > > > Isn't that the BOGO sort? > Even worse, sort by generating all possible in-order sequences and checking each one to see if it is a permutation of the data. If you write the first one in Prolog, you get something like (modulo syntax) sorted(x,y) if permutation(x,y) AND inorder(x) (Read "x is a sorted version of y if x is a permutation of y and x is in order". This can be run with y as input and then produces x.) Logically, you can swap the two argument of AND, getting sorted(x,y) if inorder(x) AND permutation(x,y) thereby getting the second algorithm, because Prolog, despite its claim to separate the logical spec from the implementation detail, requires semantics equivalent to a left-to-right, depth-first search. Isn't Logic Programming wonderful? --brian -- Brian Boutel Wellington, New Zealand Note the NOSPAM ###### Date: Fri, 21 Mar 2003 06:41:01 +0100 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Message-ID: <20030321064101.3e03896c.steveo@eircom.net> References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 19 Organization: EuroNet Internet NNTP-Posting-Date: 21 Mar 2003 06:31:03 GMT NNTP-Posting-Host: p2472.vwr.wanadoo.nl X-Trace: 1048228263 maya.euronet.nl 138 194.134.217.177:1263 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-lei1.dfn.de!newsfeed.freenet.de!194.168.4.91.MISMATCH!newspeer1-gui.server.ntli.net!ntli.net!news2.euro.net!postnews1.euro.net!maya.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133074 On 20 Mar 2003 21:56:40 +0100 Andreas Eder wrote: AE> Brian Inglis writes: AE> AE> > There's never an excuse other than mental laziness for a bubble AE> > sort. *ALL* other sorts are faster! AE> AE> Well, almost all. How about generating all permutations of the data AE> and then look for the sorted one? The slowest one is to write the data into memory and wait for random bit flips to do the job. -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### Message-ID: <3E7AFFE6.588B51DE@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <20030321064101.3e03896c.steveo@eircom.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 21 Date: Fri, 21 Mar 2003 12:12:46 GMT NNTP-Posting-Host: 12.90.168.132 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc04-news.ops.worldnet.att.net 1048248766 12.90.168.132 (Fri, 21 Mar 2003 12:12:46 GMT) NNTP-Posting-Date: Fri, 21 Mar 2003 12:12:46 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!news-xfer.cox.net!cox.net!cyclone1.gnilink.net!wn14feed!wn13feed!worldnet.att.net!bgtnsc04-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132898 Steve O'Hara-Smith wrote: > Andreas Eder wrote: > > AE> Brian Inglis writes: > AE> > AE> > There's never an excuse other than mental laziness for a bubble > AE> > sort. *ALL* other sorts are faster! > AE> > AE> Well, almost all. How about generating all permutations of the data > AE> and then look for the sorted one? > > The slowest one is to write the data into memory and wait for > random bit flips to do the job. with ECC installed and enabled. I can do anything slower than you. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Fri, 21 Mar 2003 16:57:31 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 20 Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1048265851 58186 146.186.61.46 (21 Mar 2003 16:57:31 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Fri, 21 Mar 2003 16:57:31 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132986 In article , Andreas Eder wrote: >Brian Inglis writes: >> There's never an excuse other than mental laziness for a bubble >> sort. *ALL* other sorts are faster! >Well, almost all. How about generating all permutations of the data >and then look for the sorted one? I thought of that as soon as I sent this, and *almost* superceded it. Yes, the random sort is very, very, bad :) O(N!) . . . hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Fri, 21 Mar 2003 16:59:33 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 33 Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1048265973 58186 146.186.61.46 (21 Mar 2003 16:59:33 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Fri, 21 Mar 2003 16:59:33 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132992 In article , Christian Brunschen wrote: >In article , > wrote: >>On 20 Mar 2003 21:56:40 +0100, Andreas Eder >>wrote: >>>Well, almost all. How about generating all permutations of the data >>>and then look for the sorted one? >>Isn't that the BOGO sort? >The description of bogosort that I recall was something akin to the >following: > while (the array is not sorted) { > swap two random elements of the array > } >The beauty of this version of bogosort is that it has no guaranteed upper >bound to its execution time - it could in theory mever terminate at all, >if the random number generator never hits on the right combination of >elements to exchange! :) Yes, but unless the random number generator is incapable of hitting the right combination, it finishes in finite time almost surely . . . hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### Date: Fri, 21 Mar 2003 19:42:01 +0100 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Message-ID: <20030321194201.39634457.steveo@eircom.net> References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <20030321064101.3e03896c.steveo@eircom.net> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 20 Organization: EuroNet Internet NNTP-Posting-Date: 21 Mar 2003 19:19:07 GMT NNTP-Posting-Host: i1491.vwr.wanadoo.nl X-Trace: 1048274347 maya.euronet.nl 157 194.134.213.217:1298 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!eusc.inter.net!mephistopheles.news.clara.net!news.clara.net!news2.euro.net!postnews1.euro.net!maya.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133058 On Fri, 21 Mar 03 12:48:13 GMT jmfbahciv@aol.com wrote: JC> In article <20030321064101.3e03896c.steveo@eircom.net>, JC> Steve O'Hara-Smith wrote: JC> >On 20 Mar 2003 21:56:40 +0100 JC> > The slowest one is to write the data into memory and wait for JC> >random bit flips to do the job. JC> > JC> When you figure out how to get Murphy to do your JC> work for you, you'll be able to buy Misfot and fix it. It's cheaper to use the same algorithm on them, and about as effective. -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### NNTP-Posting-Date: Fri, 21 Mar 2003 16:24:33 -0600 Date: Fri, 21 Mar 2003 17:22:23 -0500 From: John Stracke User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.1) Gecko/20020826 X-Accept-Language: en-us, en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <20030321064101.3e03896c.steveo@eircom.net> <3E7AFFE6.588B51DE@yahoo.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Message-ID: Lines: 20 NNTP-Posting-Host: 65.220.90.253 X-Trace: sv3-iZZd01JVGHT0BQo5rbzEvn0ISHhMXrVeHal3xE+Hm+NvMEWPZMzl4qw87i7CQAmHtLVUHOjyR5G00vZ!g4ClmRJOoDfYaoa5Hp6A4kfvvAhtZLXfwa6V1JQFcZ4rPKD34+K3XA2PC1ghSup5bv8wZw6yCbfG!nSOrkKbNIS1zUa/Tf7p/ X-Complaints-To: abuse@speakeasy.net X-DMCA-Complaints-To: abuse@speakeasy.net X-Abuse-and-DMCA-Info: Please be sure to forward a copy of ALL headers X-Abuse-and-DMCA-Info: Otherwise we will be unable to process your complaint properly X-Postfilter: 1.1 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed-east.nntpserver.com!nntpserver.com!border1.nntp.aus1.giganews.com!nntp.giganews.com!nntp3.aus1.giganews.com!nntp.speakeasy.net!news.speakeasy.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133115 CBFalconer wrote: > Steve O'Hara-Smith wrote: >> The slowest one is to write the data into memory and wait for >>random bit flips to do the job. > > with ECC installed and enabled. I can do anything slower than you. Send a robot ship out to the Oort cloud, write one datum on each comet, then wait for the comets to move into an arrangement that presents them in order (from the point of view of someone on a fixed spot on the Earth). -- /=============================================================\ |John Stracke | http://www.thibault.org |HTML OK | |Francois Thibault |========================================| |East Kingdom |So what's the gene for belief in genetic| |francis@thibault.org|determinism? | \=============================================================/ ###### Date: Sat, 22 Mar 2003 07:46:19 +0100 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Message-ID: <20030322074619.09bfa1a0.steveo@eircom.net> References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <20030321064101.3e03896c.steveo@eircom.net> <3E7AFFE6.588B51DE@yahoo.com> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 16 Organization: EuroNet Internet NNTP-Posting-Date: 22 Mar 2003 07:11:38 GMT NNTP-Posting-Host: p3359.vwr.wanadoo.nl X-Trace: 1048317098 maya.euronet.nl 140 212.129.225.39:1331 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!solnet.ch!solnet.ch!newsfeed.freenet.de!194.168.222.61.MISMATCH!newspeer1-gui.server.ntli.net!ntli.net!news2.euro.net!postnews1.euro.net!maya.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133090 On Fri, 21 Mar 2003 12:12:46 GMT CBFalconer wrote: C> Steve O'Hara-Smith wrote: C> > The slowest one is to write the data into memory and wait for C> > random bit flips to do the job. C> C> with ECC installed and enabled. I can do anything slower than you. Y'see ECC does degrade system performance significantly. -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### Message-ID: <3E7C2484.407E6889@yahoo.com> From: CBFalconer Reply-To: cbfalconer@worldnet.att.net Organization: Ched Research X-Mailer: Mozilla 4.75 [en] (Win98; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <20030321064101.3e03896c.steveo@eircom.net> <3E7AFFE6.588B51DE@yahoo.com> <20030322074619.09bfa1a0.steveo@eircom.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 21 Date: Sat, 22 Mar 2003 08:55:37 GMT NNTP-Posting-Host: 12.90.167.40 X-Complaints-To: abuse@worldnet.att.net X-Trace: bgtnsc05-news.ops.worldnet.att.net 1048323337 12.90.167.40 (Sat, 22 Mar 2003 08:55:37 GMT) NNTP-Posting-Date: Sat, 22 Mar 2003 08:55:37 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!news.webusenet.com!cyclone1.gnilink.net!wn14feed!wn13feed!worldnet.att.net!bgtnsc05-news.ops.worldnet.att.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133212 Steve O'Hara-Smith wrote: > CBFalconer wrote: > C> Steve O'Hara-Smith wrote: > > C> > The slowest one is to write the data into memory and wait for > C> > random bit flips to do the job. > C> > C> with ECC installed and enabled. I can do anything slower than you. > > Y'see ECC does degrade system performance significantly. I think you miss the point. A fraction of 1% slowdown is not it. The lack of random bit flips in a properly protected memory system is. This can well extend the above sort timing by many orders of magnitude. -- Chuck F (cbfalconer@yahoo.com) (cbfalconer@worldnet.att.net) Available for consulting/temporary embedded and systems. USE worldnet address! ###### Date: Sat, 22 Mar 2003 13:49:58 +0100 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Message-ID: <20030322134958.0b153594.steveo@eircom.net> References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <20030321064101.3e03896c.steveo@eircom.net> <3E7AFFE6.588B51DE@yahoo.com> <20030322074619.09bfa1a0.steveo@eircom.net> <3E7C2484.407E6889@yahoo.com> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 36 Organization: EuroNet Internet NNTP-Posting-Date: 22 Mar 2003 12:52:43 GMT NNTP-Posting-Host: p3594.vwr.wanadoo.nl X-Trace: 1048337563 maya.euronet.nl 154 212.129.226.18:1357 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newsfeed00.sul.t-online.de!t-online.de!newsfeed.freenet.de!194.168.222.21.MISMATCH!newspeer1-gui.server.ntli.net!ntli.net!news2.euro.net!postnews1.euro.net!maya.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133069 On Sat, 22 Mar 2003 08:55:37 GMT CBFalconer wrote: C> Steve O'Hara-Smith wrote: C> > CBFalconer wrote: C> > C> Steve O'Hara-Smith wrote: C> > C> > C> > The slowest one is to write the data into memory and wait for C> > C> > random bit flips to do the job. C> > C> C> > C> with ECC installed and enabled. I can do anything slower than you. C> > Y'see ECC does degrade system performance significantly. C> I think you miss the point. A fraction of 1% slowdown is not it. No I don't - the slowdown is just rather more extreme than the normal fraction of one percent in this case :) C> The lack of random bit flips in a properly protected memory system They're not lacking - just very low incidence rate it'll get there given long enough. C> is. This can well extend the above sort timing by many orders of C> magnitude. I suspect with ECC we'd be lucky to finish that sort this side of the heat death. Even with overclocked gamers RAM it'd be unlikely to finish before the machine dies. -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 22 Mar 03 13:16:14 GMT Organization: UltraNet Communications, Inc. Lines: 22 Message-ID: References: <3E7B801A.F11C88CC@computer.org> X-Trace: UmFuZG9tSVZBpGdf5tpXqaof40JyrdjvHSnUNfYkuQroTs9/mrWAjfqLFqV+XHdw X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 22 Mar 2003 14:03:49 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-182-13 Xref: chonsp.franklin.ch alt.folklore.computers:132994 In article <3E7B801A.F11C88CC@computer.org>, Sam Yorko wrote: >jmfbahciv@aol.com wrote: >> >> >> Huh? Who would be foolish enough to do that? Fishing line is too >> thin. >> >> /BAH >> > >Depends on the weight. Fishing line can get into some pretty high test >weight sizes. Sure. But I don't think I'd hang a five pound thingie on 15# line. Not only does that stuff cut itself, but I'd understood that the poundage is based on temporary poundage and not constant over a long time. /BAH Subtract a hundred and four for e-mail. ###### From: Sam Yorko Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 21 Mar 2003 12:41:00 -0800 Lines: 30 Message-ID: <3E7B78DB.852C6CDA@computer.org> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> NNTP-Posting-Host: 63.145.233.34 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: fu-berlin.de 1048279237 75868429 63.145.233.34 (16 [71567]) X-Mailer: Mozilla 4.79 [en] (Windows NT 5.0; U) X-Accept-Language: en Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!63.145.233.34!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132933 Ben Hutchings wrote: > > In article , > I wrote: > > > Some versions of DOS and Windows have a limit of 2 GB. > > I remembered what this limit is. It's the limit on file-system size > in FAT16 - the maximum block size is 32K and block numbers are 16-bit. > This limit applies to the original Windows 95, NT 4 and all earlier MS > operating systems using FAT. > > > Later BIOSes invent a new fictitious geometry for the disk that suits > > the disk API, and translate addresses. This raises the limit to > > 1024 * 255 * 63 sectors ~= 8 GB. > > > This is assuming they get translation right. Apparently some buggy > BIOSes have a limit somewhere between 500 MB and 8 GB. > > -- > Ben Hutchings | personal web site: http://womble.decadentplace.org.uk/ > 73.46% of all statistics are made up. IIRC, there was some sort of limitation at the 520MB level. It may have been DOS and WFW, or something else. My SCSI disk controller warns me if I attempt to use a disk that is bigger, and asks if I want to do some sort of magic to work around it. Sam ###### From: Sam Yorko Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 21 Mar 2003 12:42:19 -0800 Lines: 51 Message-ID: <3E7B792B.1428428E@computer.org> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> NNTP-Posting-Host: 63.145.233.34 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: fu-berlin.de 1048279313 75868429 63.145.233.34 (16 [71567]) X-Mailer: Mozilla 4.79 [en] (Windows NT 5.0; U) X-Accept-Language: en Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed00.sul.t-online.de!newsfeed01.sul.t-online.de!t-online.de!fu-berlin.de!uni-berlin.de!63.145.233.34!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132916 Ben Hutchings wrote: > > In article , > Charles Shannon Hendrix wrote: > > In article > >, Ben > > Hutchings wrote: > > > >>> The exact limitations aren't even the same across all BIOS, which makes > >>> the problem even more "interesting". > >> > >> There are two different limits on drive geometry: the original BIOS > >> disk API is limited to 1024 cylinders, 255 heads (!) and 63 sectors; > >> the original AT HD controller (which ATA/IDE disks remain compatible > >> with) is limited to 16383 cylinders, 16 heads and 63 sectors. Combine > >> these limitations and you get just under 1024 * 16 * 63 sectors ~= > >> 500 MB. > > > > As I said, I'm not talking about drive geometry. > > > > I'm talking about the size of the image that the BIOS function your > > 446/512 byte loader calls. > > > > I thought there was a limit to that. > > Except in the case of a floppy disk or a virtual floppy disk on a > CD-ROM, the size limitation is that of the BIOS disk API. > > > If not, then why do kernels have to be small, > > They don't. > > > and/or why do other OS put so much effort into an interrim stage > > which loads the kernels? > > > The boot loader needs to work with any hard disk, so it has to use > BIOS calls to load, so it has to run in real mode, so it's limited to > accessing the first 1 MB of memory. Well, actually it could switch > back and forth between real and protected mode, but that would be a > pain. > > -- > Ben Hutchings | personal web site: http://womble.decadentplace.org.uk/ > 73.46% of all statistics are made up. There is some sort of limitation at the 2G level for the OS. There is a warning that the core files for NT4.0 and W2K have to reside in the first two G of the hard disk, or the system can't boot. Sam ###### From: Sam Yorko Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 21 Mar 2003 13:11:54 -0800 Lines: 13 Message-ID: <3E7B801A.F11C88CC@computer.org> References: NNTP-Posting-Host: 63.145.233.34 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: fu-berlin.de 1048281088 75374511 63.145.233.34 (16 [71567]) X-Mailer: Mozilla 4.79 [en] (Windows NT 5.0; U) X-Accept-Language: en Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!63.145.233.34!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132928 jmfbahciv@aol.com wrote: > > > Huh? Who would be foolish enough to do that? Fishing line is too > thin. > > /BAH > Depends on the weight. Fishing line can get into some pretty high test weight sizes. Sam ###### From: peter@taronga.com (Peter da Silva) Newsgroups: alt.folklore.computers,alt.sys.pdp10 Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 22 Mar 2003 23:04:58 GMT Organization: TSS Inc. Lines: 19 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <6o9r5vc2ekhvfnku4cd86fkso6uejbc2n1@4ax.com> NNTP-Posting-Host: citadel.in.taronga.com X-Trace: citadel.in.taronga.com 1048374298 82412 10.0.0.43 (22 Mar 2003 23:04:58 GMT) X-Complaints-To: usenet@taronga.com NNTP-Posting-Date: 22 Mar 2003 23:04:58 GMT X-Newsreader: trn 4.0-test72 (19 April 1999) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news.ip-plus.net!news.tesion.net!news.belwue.de!newsfeed01.sul.t-online.de!newsfeed00.sul.t-online.de!t-online.de!news.tele.dk!news.tele.dk!small.news.tele.dk!newsfeed.stanford.edu!news-proxy.abbnm.com!citadel.in.taronga.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133110 alt.sys.pdp10:12754 In article , Heiko Berges wrote: >Peter da Silva wrote: >>Here's the closest thing to the real thing... >>http://www.bostic.com/vi/ >http://ex-vi.berlios.de Heh. Going back to the original reminds me why I was so impressed by nvi's unlimited undo. It's better than the real thing. The only thing that I'd like to add is elvis' "v" move mode. -- Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD? "Be conservative in what you generate, and liberal in what you accept" -- Matthew 10:16 (l.trans) ###### From: Don Quixote Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Reply-To: donq+afc@eclipse.net Message-ID: <0t1q7vs5obkmnedr4oh6dn2rrpur6ak1h5@fishead.news.meowkitty.com> References: <6k4i3b.hhc2.ln@via.reistad.priv.no> <3E7B792B.1428428E@computer.org> X-Newsreader: Forte Agent 1.93/32.576 English (US) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Virus: Hi! I'm a header virus. Copy me to your headers and I'll spread. X-Org-1: If the Organization: header does not read 'wanted', then this post X-Org-2: has been modified from it's originally posted form and should not X-Org-3: be trusted. Lines: 17 Date: Sat, 22 Mar 2003 20:05:45 -0500 NNTP-Posting-Host: 207.207.243.252 X-Complaints-To: news@netcarrier.com X-Trace: news.netcarrier.net 1048388859 207.207.243.252 (Sat, 22 Mar 2003 22:07:39 EST) NNTP-Posting-Date: Sat, 22 Mar 2003 22:07:39 EST Organization: NetCarrier Internet Services Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!news.tele.dk!news.tele.dk!small.news.tele.dk!newsfeed.icl.net!newsfeed.fjserv.net!newsfeed.icl.net!newsfeed.fjserv.net!newsfeed1.cidera.com!Cidera!news.netcarrier.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133179 Sam Yorko might have said: >... >There is a >warning that the core files for NT4.0 and W2K have to reside in the >first two G of the hard disk, or the system can't boot. With regard to NT 4.0 I think the number is four, not two. Which is also the limit of the partition size that can be created while performing an NT installation. I think. I could be wrong, and whether or not I am, someone will be along to correct me shortly. -- 'After I started drinking yesterday I didn't do anything else stupid. That would seem to speak for itself.' Adam, in the Monastery. ###### From: William Hamblen Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <6feq7v4ufqr8svusdcsesp67i1ofcvp1h9@4ax.com> References: <3E7B792B.1428428E@computer.org> <0t1q7vs5obkmnedr4oh6dn2rrpur6ak1h5@fishead.news.meowkitty.com> X-Newsreader: Forte Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 13 Date: Sun, 23 Mar 2003 04:39:22 GMT NNTP-Posting-Host: 216.80.152.224 X-Complaints-To: abuse@earthlink.net X-Trace: newsread2.prod.itd.earthlink.net 1048394362 216.80.152.224 (Sat, 22 Mar 2003 20:39:22 PST) NNTP-Posting-Date: Sat, 22 Mar 2003 20:39:22 PST Organization: EarthLink Inc. -- http://www.EarthLink.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-fra1.dfn.de!news0.de.colt.net!peernews3.colt.net!colt.net!peernews-us.colt.net!newsfeed.news2me.com!newsfeed2.earthlink.net!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!newsread2.prod.itd.earthlink.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133203 On Sat, 22 Mar 2003 20:05:45 -0500, Don Quixote wrote: >With regard to NT 4.0 I think the number is four, not two. Which is >also the limit of the partition size that can be created while >performing an NT installation. I think. I could be wrong, and >whether or not I am, someone will be along to correct me shortly. By default the x86 NT 4.0 installer makes a FAT16 partition at the beginning of the disk to install the operating system in and the largest size a FAT16 partition can have is 2 Gbytes. The Alpha version probably does things differently. ###### From: Sam Yorko Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sun, 23 Mar 2003 00:00:28 -0800 Lines: 21 Message-ID: <3E7D699C.5FAA8E22@computer.org> References: <3E7B792B.1428428E@computer.org> <0t1q7vs5obkmnedr4oh6dn2rrpur6ak1h5@fishead.news.meowkitty.com> <6feq7v4ufqr8svusdcsesp67i1ofcvp1h9@4ax.com> NNTP-Posting-Host: w034.z065106070.sjc-ca.dsl.cnc.net (65.106.70.34) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: fu-berlin.de 1048406404 74786775 65.106.70.34 (16 [71567]) X-Mailer: Mozilla 4.79 [en] (Windows NT 5.0; U) X-Accept-Language: en Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!uio.no!feed.news.nacamar.de!fu-berlin.de!uni-berlin.de!w034.z065106070.sjc-ca.dsl.cnc.NET!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:132930 William Hamblen wrote: > > On Sat, 22 Mar 2003 20:05:45 -0500, Don Quixote > wrote: > > >With regard to NT 4.0 I think the number is four, not two. Which is > >also the limit of the partition size that can be created while > >performing an NT installation. I think. I could be wrong, and > >whether or not I am, someone will be along to correct me shortly. > > By default the x86 NT 4.0 installer makes a FAT16 partition at the > beginning of the disk to install the operating system in and the > largest size a FAT16 partition can have is 2 Gbytes. The Alpha > version probably does things differently. Naw, even if you make the first partition NTFS during initial install, the OS boot loader can't see anything past the first two Gig. I once somehow managed to get a critical file past the 2 G limit, and I think I had to boot into debug mode to fix it. Sam ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Mon, 24 Mar 2003 18:01:16 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 36 Message-ID: References: <1by94bd27o.fsf@cs.nmsu.edu> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1048528876 44924 146.186.61.46 (24 Mar 2003 18:01:16 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Mon, 24 Mar 2003 18:01:16 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133279 In article , wrote: >Dr. Richard E. Hawkins wrote: >> In article , >> Brian Inglis wrote: >>>There's never an excuse other than mental laziness for a bubble >>>sort. *ALL* other sorts are faster! >> All other sorts have a better O(). For small enough lists, the overhead >> on each action can (sometimes) make a bubble sort faster, iirc. >As far as I am aware, neither of your sentences are true. ahh, swell :) >It is not >necessarily true that more advanced sorts have better O() in the worst >case I'm thinking of general cases :) And then there are the various random sorts . . . >and it is not true that a bubble sort has the smallest K of any >possible sorts. K? For some very small sorts, the cheaper execution/comparison can offset the explosive growth in the number of comparisons. hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: Giles Todd Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Tue, 25 Mar 2003 06:04:44 +0100 Organization: The action of organizing or putting into systematic form; the arranging and co-ordinating of parts into a systematic whole. Reply-To: gt@at-dot.org Message-ID: Cancel-Lock: sha1:gh91G3LP5hg/Sj/i5HeD+QJ4Dtg= References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <3E791A67.541A3D9D@ev1.net> X-Newsreader: Forte Agent 1.93/32.576 English (American) X-No-Archive: yes MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-NFilter: 1.2.0 Lines: 13 NNTP-Posting-Date: 25 Mar 2003 06:05:16 CET NNTP-Posting-Host: 80.127.198.243 X-Trace: 1048568716 news.xs4all.nl 49103 80.127.198.243:4416 X-Complaints-To: abuse@xs4all.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-fra1.dfn.de!news.f.de.plusline.net!feed.news.nacamar.de!newsfeed.icl.net!newsfeed.fjserv.net!skynet.be!skynet.be!transit.news.xs4all.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133402 On Wed, 19 Mar 2003 23:35:21 GMT, Charles Richmond wrote in message <3E791A67.541A3D9D@ev1.net>: > Also, a bubble sort is *so* easy to implement...it is less > likely to have a bug in it. s/less/more/ Bubblesort is so trivial to implement (and it is only going to be used for those special cases) that the coder is likely to get it wrong through not paying attention. Giles. ###### Message-ID: <98u6l-oo3.ln1@chromite.brokersys.com> From: jguthrie@brokersys.com Subject: Re: PDP10 and RISC Newsgroups: alt.folklore.computers References: <1by94bd27o.fsf@cs.nmsu.edu> User-Agent: tin/1.5.17-20030301 ("Bubbles") (UNIX) (Linux/2.4.18-k7+xfs+freeswan (i686)) Date: Tue, 25 Mar 2003 15:28:09 -0600 NNTP-Posting-Host: 216.39.196.197 X-Trace: news3.mylinuxisp.com 1048631100 216.39.196.197 (25 Mar 2003 16:25:00 -0600) Lines: 55 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!uio.no!feed.news.nacamar.de!newsfeed.icl.net!newsfeed.fjserv.net!newsfeed1.cidera.com!Cidera!news3.mylinuxisp.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133505 Dr. Richard E. Hawkins wrote: > In article , > wrote: >>Dr. Richard E. Hawkins wrote: >>> In article , >>> Brian Inglis wrote: >>>>There's never an excuse other than mental laziness for a bubble >>>>sort. *ALL* other sorts are faster! >>> All other sorts have a better O(). For small enough lists, the overhead >>> on each action can (sometimes) make a bubble sort faster, iirc. >>As far as I am aware, neither of your sentences are true. > ahh, swell :) Of course, I could be wrong. (And you can usually tell when I've gone off the deep end of an argument because I leave out the weasel words to that effect.) >>It is not >>necessarily true that more advanced sorts have better O() in the worst >>case > I'm thinking of general cases :) And then there are the various random > sorts . . . >>and it is not true that a bubble sort has the smallest K of any >>possible sorts. > K? If sorting zero elements takes zero time, then the run time of a sort with O() behavior for N elements can be expressed with an equation like "T = K * + lower-order effects" for whatever describes the behavior of the sort when the number of elements becomes large. (With appropriate hand-waving to make it sound mathematical. I'm not a mathematician, I'm just an engineer with delusions of grandeur.) > For some very small sorts, the cheaper execution/comparison can offset > the explosive growth in the number of comparisons. I understand. However, if you can neglect the lower order effects, then the sort with the smallest "K" will tend to win when compared with sorts with a larger "K". When I did tests in the late 80's, I found that an extraction sort of an array of integers tends to be faster than a bubble sort for large numbers of elements. It just now occurred to me that I cannot extrapolate from those results to the results with a smaller number of elements because of the simplifying assumptions inherent in the O()-notation. Never mind. -- Jonathan Guthrie (jguthrie@brokersys.com) Sto pro veritate ###### Message-ID: <3E80FBFD.9EB0150D@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <3E791A67.541A3D9D@ev1.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 23 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: rwcrnsc53 1048633427 12.241.15.59 (Tue, 25 Mar 2003 23:03:47 GMT) NNTP-Posting-Date: Tue, 25 Mar 2003 23:03:47 GMT Date: Tue, 25 Mar 2003 23:03:47 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!news.webusenet.com!cyclone1.gnilink.net!wn14feed!wn13feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!rwcrnsc53.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133530 Giles Todd wrote: > > On Wed, 19 Mar 2003 23:35:21 GMT, Charles Richmond > wrote in message <3E791A67.541A3D9D@ev1.net>: > > > Also, a bubble sort is *so* easy to implement...it is less > > likely to have a bug in it. > > s/less/more/ > > Bubblesort is so trivial to implement (and it is only going to be used > for those special cases) that the coder is likely to get it wrong > through not paying attention. > Even if we accept this, the implementation of the bubble sort is usually so *small* that a test program can be easily put together... IMHO, it is *no* harder to write a bubble sort than to set up a comparison function for the C qsort() function. -- +----------------------------------------------------------------+ | Charles and Francis Richmond richmond at plano dot net | +----------------------------------------------------------------+ ###### From: Giles Todd Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Wed, 26 Mar 2003 05:11:47 +0100 Organization: The action of organizing or putting into systematic form; the arranging and co-ordinating of parts into a systematic whole. Reply-To: gt@at-dot.org Message-ID: Cancel-Lock: sha1:7cENJUNoKiRKco9Fmw1h2sbbgtQ= References: <1by94bd27o.fsf@cs.nmsu.edu> <1frnmvq.14xg6kv1yno1h3N%lars@bearnip.com> <3E791A67.541A3D9D@ev1.net> <3E80FBFD.9EB0150D@ev1.net> X-Newsreader: Forte Agent 1.93/32.576 English (American) X-No-Archive: yes MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-NFilter: 1.2.0 Lines: 12 NNTP-Posting-Date: 26 Mar 2003 05:12:35 CET NNTP-Posting-Host: 80.127.198.243 X-Trace: 1048651955 news.xs4all.nl 49107 80.127.198.243:2302 X-Complaints-To: abuse@xs4all.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!solnet.ch!solnet.ch!newsfeed.freenet.de!news2.euro.net!news2.euro.net!transit.news.xs4all.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133523 On Tue, 25 Mar 2003 23:03:47 GMT, Charles Richmond wrote in message <3E80FBFD.9EB0150D@ev1.net>: > Even if we accept this, the implementation of the bubble sort > is usually so *small* that a test program can be easily put > together... IMHO, it is *no* harder to write a bubble sort > than to set up a comparison function for the C qsort() function. I agree with your HO. My only argument with you is about how things turn out in practice. Giles. ###### From: Giles Todd Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Wed, 26 Mar 2003 05:19:08 +0100 Organization: The action of organizing or putting into systematic form; the arranging and co-ordinating of parts into a systematic whole. Reply-To: gt@at-dot.org Message-ID: Cancel-Lock: sha1:LttrStydNuFMqZV1cRXmDnGMCrk= References: <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com> <1frznw0.8e1waqbp4y8N%lars@bearnip.com> <851.207T1872T7555017@kltpzyxm.invalid> <1fs2bof.1xrda5x1w2b5luN%lars@bearnip.com> X-Newsreader: Forte Agent 1.93/32.576 English (American) X-No-Archive: yes MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-NFilter: 1.2.0 Lines: 12 NNTP-Posting-Date: 26 Mar 2003 05:19:11 CET NNTP-Posting-Host: 80.127.198.243 X-Trace: 1048652351 news.xs4all.nl 49116 80.127.198.243:2309 X-Complaints-To: abuse@xs4all.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!solnet.ch!solnet.ch!newsfeed.freenet.de!news2.euro.net!news2.euro.net!transit.news.xs4all.nl!newsfeed.xs4all.nl!xs4all!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133540 On Wed, 19 Mar 2003 12:16:40 -0700, lars@bearnip.com (Lars Duening) wrote in message <1fs2bof.1xrda5x1w2b5luN%lars@bearnip.com>: > It's a thought I have whenever I install the Cygwin environment on a > Windows machine: since I'm basically installing an OS compatibility > environment in order to get programs compiled, what is the point of > using Windows at all? Your executable has to run on a box with Windows installed fresh from the CD. Nothing else. Giles. ###### From: lars@bearnip.com (Lars Duening) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Wed, 26 Mar 2003 08:54:40 -0700 Lines: 15 Message-ID: <1fsejlz.htb5qx1qzuzqyN%lars@bearnip.com> References: <3E5A710D.6030403@jetnet.ab.ca> <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com> <1frznw0.8e1waqbp4y8N%lars@bearnip.com> <851.207T1872T7555017@kltpzyxm.invalid> <1fs2bof.1xrda5x1w2b5luN%lars@bearnip.com> NNTP-Posting-Host: dhcp-dsl-31-26-31.interfold.com (198.31.26.31) X-Trace: fu-berlin.de 1048694081 80332351 198.31.26.31 (16 [14293]) X-Orig-Path: lars User-Agent: MacSOUP/2.5b1 (Mac OS X version 10.2.4) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!dhcp-dsl-31-26-31.interfold.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133537 Giles Todd wrote: > On Wed, 19 Mar 2003 12:16:40 -0700, lars@bearnip.com (Lars Duening) > wrote in message <1fs2bof.1xrda5x1w2b5luN%lars@bearnip.com>: > > > It's a thought I have whenever I install the Cygwin environment on a > > Windows machine: since I'm basically installing an OS compatibility > > environment in order to get programs compiled, what is the point of > > using Windows at all? > > Your executable has to run on a box with Windows installed fresh from > the CD. Nothing else. No problem, I will just link in installer and images for a decent OS into the executable as well. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Thu, 27 Mar 03 12:36:07 GMT Organization: UltraNet Communications, Inc. Lines: 28 Message-ID: References: <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com> <1frznw0.8e1waqbp4y8N%lars@bearnip.com> <851.207T1872T7555017@kltpzyxm.invalid> <1fs2bof.1xrda5x1w2b5luN%lars@bearnip.com> <3E82150F.84D9B00F@yahoo.com> <3E825D18.4827F91A@ev1.net> X-Trace: UmFuZG9tSVZTFfQV6LHNjJQrsz8jXfBPokZR4P+FAgEMxFCGGQreo3M1zQQ1DjRs X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 27 Mar 2003 13:24:40 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!newscore.univie.ac.at!newsfeed01.sul.t-online.de!t-online.de!feed.news.nacamar.de!feed2.news.rcn.net!feed1.news.rcn.net!rcn!209-122-255-43 Xref: chonsp.franklin.ch alt.folklore.computers:133586 In article <3E825D18.4827F91A@ev1.net>, Charles Richmond wrote: >Peter Flass wrote: >> >> Giles Todd wrote: >> > >> > On Wed, 19 Mar 2003 12:16:40 -0700, lars@bearnip.com (Lars Duening) >> > wrote in message <1fs2bof.1xrda5x1w2b5luN%lars@bearnip.com>: >> > >> > > It's a thought I have whenever I install the Cygwin environment on a >> > > Windows machine: since I'm basically installing an OS compatibility >> > > environment in order to get programs compiled, what is the point of >> > > using Windows at all? >> > >> >> Is there a point? I try to avoid it at all costs. >> >The point is getting hoisted on Bill Gate's petard!!! (Gee, >I wondered if I would ever get around to using the word >"petard"...look that up in your Funk & Wagnells...) I execute petard hoists every day; I just don't talk about it. :-)))) /BAH Subtract a hundred and four for e-mail. ###### Message-ID: <3E825D18.4827F91A@ev1.net> From: Charles Richmond Reply-To: richmond@ev1.net Organization: Canine Computer Center X-Mailer: Mozilla 4.7C-CCK-MCD {C-UDP; EBM-APPLE} (Macintosh; I; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC References: <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com> <1frznw0.8e1waqbp4y8N%lars@bearnip.com> <851.207T1872T7555017@kltpzyxm.invalid> <1fs2bof.1xrda5x1w2b5luN%lars@bearnip.com> <3E82150F.84D9B00F@yahoo.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 23 NNTP-Posting-Host: 12.241.15.59 X-Complaints-To: abuse@attbi.com X-Trace: sccrnsc03 1048723823 12.241.15.59 (Thu, 27 Mar 2003 00:10:23 GMT) NNTP-Posting-Date: Thu, 27 Mar 2003 00:10:23 GMT Date: Thu, 27 Mar 2003 00:10:23 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!chi1.webusenet.com!news.webusenet.com!cyclone1.gnilink.net!wn14feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi.com!sccrnsc03.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133675 Peter Flass wrote: > > Giles Todd wrote: > > > > On Wed, 19 Mar 2003 12:16:40 -0700, lars@bearnip.com (Lars Duening) > > wrote in message <1fs2bof.1xrda5x1w2b5luN%lars@bearnip.com>: > > > > > It's a thought I have whenever I install the Cygwin environment on a > > > Windows machine: since I'm basically installing an OS compatibility > > > environment in order to get programs compiled, what is the point of > > > using Windows at all? > > > > Is there a point? I try to avoid it at all costs. > The point is getting hoisted on Bill Gate's petard!!! (Gee, I wondered if I would ever get around to using the word "petard"...look that up in your Funk & Wagnells...) -- +----------------------------------------------------------------+ | Charles and Francis Richmond richmond at plano dot net | +----------------------------------------------------------------+ ###### From: "J. Clarke" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Thu, 27 Mar 2003 14:13:58 -0500 Organization: http://extra.newsguy.com Lines: 69 Message-ID: References: <3E776132.2B2A2AAD@yahoo.com> <2qah7vc7dfm559m3hdpe6o0m0kmpve2ajr@4ax.com> NNTP-Posting-Host: p-927.newsdawg.com Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Pan/0.13.4 (She had eyes like strange sins.) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews4 Xref: chonsp.franklin.ch alt.folklore.computers:133651 On Wed, 19 Mar 2003 17:52:04 +0000, Brian Inglis wrote: > On Wed, 19 Mar 2003 17:09:58 +0000 (UTC) in alt.folklore.computers, > hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: > >>In article <3E776132.2B2A2AAD@yahoo.com>, CBFalconer >> wrote: >>>"Dr. Richard E. Hawkins" wrote: >> >>>> >> When you have lost your clutch cable, it is necessary. You start >> >>>> >Like in a VW Beetle when the clutch cable goes. Fun, trying to get >>>> >all the way home, and fudging a little at yellow lights and stop >>>> >signs... :) >> >>>> Or when it's too tight. My girlfriend and I drove 80 miles the day >>>> after my clutch was replaced, off to meet her parents. Partway >>>> there, the engine would start racing if I went to fast. hmm. Her >>>> father looked at it, drove 10 feet, and told me what the problem was. >> >>>> It was an interesting drive back, as the maximum speed kept dropping, >>>> eventually down to 53 . . . >> >>>You probably needed another new clutch, after all that slipping. That's >>>why you always keep adequate free play in the clutch pedal. You had a >>>very sloppy mechanic. >> >>Yep. I showed up on monday, and said that they needed to redo my >>clutch. >> The lady snarled, "how many miles does it have on it?" "400." >> "Oh." > > Looked like they expected you back, but not so soon: must have > overtightened the cable too much. > >> It wasn't until later that I drew the connection between going in for >>service there and the new problems that arose within a day or two. >>plugging that loose wire (a factory test wire?) into the fuse block was >>the real joy. We'd traced the entire electrical system and lifted the >>upholstery trying to find the short that caused the fuse to blow when >>the taillights went on. Suddenly inspiration hit: what if I remove >>this wire, which seems to have no purpose? Amazing how much better >>things work when you don't connect +12 to GND . . . > > Looked like they were doing their best to ensure you returned to give > them more business. > > "Never attribute to malice what can be adequately explained by > stupidity." > > A couple of problems like that is a definite indication that one, the > other, or both, is occurrring. Hope you never went back there again, and > spread the word. Just remember, "Once is happenstance, twice is circumstance, three times is enemy action." > Once went into a brake place to make an appointment, heard the boss tell > a (presumably new) mechanic to just replace the drums or rotors or > whatever, "'coz that's how we do things here". Said I'd be back later > and left to spread the word. > > Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- --John Reply to jclarke at ae tee tee global dot net (was jclarke at eye bee em dot net) ###### From: Jeff Teunissen Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Dusk To Dawn Computing Lines: 17 Message-ID: <3E842F8E.89281BD8@d2dc.net> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <3E7B78DB.852C6CDA@computer.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 4.77 [en] (X11; U; Linux 2.4.19 i686) X-Accept-Language: en-US, en NNTP-Posting-Host: 12.245.2.60 X-Complaints-To: abuse@attbi.com X-Trace: sccrnsc02 1048850401 12.245.2.60 (Fri, 28 Mar 2003 11:20:01 GMT) NNTP-Posting-Date: Fri, 28 Mar 2003 11:20:01 GMT Date: Fri, 28 Mar 2003 11:20:01 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.icl.net!newsfeed.fjserv.net!kibo.news.demon.net!demon!news.maxwell.syr.edu!out.nntp.be!propagator2-sterling!news-in.nuthinbutnews.com!cyclone1.gnilink.net!wn14feed!wn13feed!worldnet.att.net!204.127.198.203!attbi_feed3!attbi_feed4!attbi.com!sccrnsc02.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133854 Sam Yorko wrote: [snip] > IIRC, there was some sort of limitation at the 520MB level. It may have > been DOS and WFW, or something else. My SCSI disk controller warns me > if I attempt to use a disk that is bigger, and asks if I want to do some > sort of magic to work around it. There's a 504/528MB limit (BIOS, 1024/16/64 CHS), and a 2/2.1GB limit (DOS, 1024*64*64 CHS), and an 8/8.4GB limit (LBA, 1024/256/64 CHS). -- | Jeff Teunissen -=- Pres., Dusk To Dawn Computing -=- deek @ d2dc.net | GPG: 1024D/9840105A 7102 808A 7733 C2F3 097B 161B 9222 DAB8 9840 105A | Core developer, The QuakeForge Project http://www.quakeforge.net/ | Specializing in Debian GNU/Linux http://www.d2dc.net/~deek/ ###### From: Sam Yorko Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 28 Mar 2003 15:03:41 -0800 Lines: 20 Message-ID: <3E84D4CC.7DBB4E45@computer.org> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <3E7B78DB.852C6CDA@computer.org> <3E842F8E.89281BD8@d2dc.net> NNTP-Posting-Host: symsj01.sj.symbol.com (63.145.233.34) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: fu-berlin.de 1048892592 1225536 63.145.233.34 (16 [71567]) X-Mailer: Mozilla 4.79 [en] (Windows NT 5.0; U) X-Accept-Language: en Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!symsj01.sj.symbol.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:133726 Jeff Teunissen wrote: > > Sam Yorko wrote: > > [snip] > > > IIRC, there was some sort of limitation at the 520MB level. It may have > > been DOS and WFW, or something else. My SCSI disk controller warns me > > if I attempt to use a disk that is bigger, and asks if I want to do some > > sort of magic to work around it. > > There's a 504/528MB limit (BIOS, 1024/16/64 CHS), and a 2/2.1GB limit > (DOS, 1024*64*64 CHS), and an 8/8.4GB limit (LBA, 1024/256/64 CHS). > Thanks. It's nice to know that my memory isn't completely shot. So what caused the ~120G limit? Sam ###### Path: chonsp.franklin.ch!not-for-mail From: Neil Franklin Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 29 Mar 2003 16:06:39 +0100 Organization: My own Private Self Lines: 38 Message-ID: <6u1y0qyzy8.fsf@chonsp.franklin.ch> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <3E7B78DB.852C6CDA@computer.org> <3E842F8E.89281BD8@d2dc.net> <3E84D4CC.7DBB4E45@computer.org> NNTP-Posting-Host: chonsp.franklin.ch X-Trace: chonsp.franklin.ch 1048950399 465 10.0.3.2 (29 Mar 2003 15:06:39 GMT) X-Complaints-To: news@chonsp.franklin.ch NNTP-Posting-Date: 29 Mar 2003 15:06:39 GMT X-Newsreader: Gnus v5.7/Emacs 20.4 Xref: chonsp.franklin.ch alt.folklore.computers:133874 Sam Yorko writes: > Jeff Teunissen wrote: > > > > > IIRC, there was some sort of limitation at the 520MB level. It may have > > > been DOS and WFW, or something else. My SCSI disk controller warns me > > > if I attempt to use a disk that is bigger, and asks if I want to do some > > > sort of magic to work around it. > > > > There's a 504/528MB limit (BIOS, 1024/16/64 CHS), and a 2/2.1GB limit > > (DOS, 1024*64*64 CHS), and an 8/8.4GB limit (LBA, 1024/256/64 CHS). Lots of imprecise and sometimes wrongly attributed numbers there. The first of those limits is really 1024*16*63*512 = 528482304 = 504MByte. Where the 1024 and 63(!) are BIOS limits and 16 is WD/IDE/ATA controller register set limit. The second I will pass up, as I do not know the data structures involved. The third of those limits is 1024*255*63*512 = 8422686720 = 8032.5MByte. Where all 1024, 255 and 63 are BIOS limits, nothing to do with LBA or even IDE/ATA drives, SCSI also has this limit, when driven via an PC BIOS. > Thanks. It's nice to know that my memory isn't completely shot. > > So what caused the ~120G limit? That is the LBA Limit of 2^28*512 = 137438953472 = 131072MByte = 128GByte. -- Neil Franklin, neil@franklin.ch.remove http://neil.franklin.ch/ Hacker, Unix Guru, El Eng HTL/BSc, Programmer, Archer, Blacksmith - hardware runs the world, software controls the hardware code generates the software, have you coded today? ###### Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E842F8E.89281BD8@d2dc.net> <3E84D4CC.7DBB4E45@computer.org> <6u1y0qyzy8.fsf@chonsp.franklin.ch> From: Morten Reistad X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: mrr@acer.reistad.priv.no (Morten Reistad) Message-ID: <29jb6b.7vm1.ln@via.reistad.priv.no> Lines: 41 Date: Tue, 1 Apr 2003 10:40:02 +0200 NNTP-Posting-Host: 80.111.42.204 X-Complaints-To: abuse@chello.no X-Trace: amstwist00 1049187601 80.111.42.204 (Tue, 01 Apr 2003 11:00:01 MET DST) NNTP-Posting-Date: Tue, 01 Apr 2003 11:00:01 MET DST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.belwue.de!newsfeed01.sul.t-online.de!t-online.de!lnewspeer01.lnd.ops.eu.uu.net!emea.uu.net!news2.euro.net!amsnews01.chello.com!amstwist00.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134067 According to Neil Franklin : >Sam Yorko writes: > >> Jeff Teunissen wrote: >> > >> > > IIRC, there was some sort of limitation at the 520MB level. It may have >> > > been DOS and WFW, or something else. My SCSI disk controller warns me >> > > if I attempt to use a disk that is bigger, and asks if I want to do some >> > > sort of magic to work around it. >> > >> > There's a 504/528MB limit (BIOS, 1024/16/64 CHS), and a 2/2.1GB limit >> > (DOS, 1024*64*64 CHS), and an 8/8.4GB limit (LBA, 1024/256/64 CHS). > >Lots of imprecise and sometimes wrongly attributed numbers there. > >The first of those limits is really 1024*16*63*512 = 528482304 = 504MByte. > Where the 1024 and 63(!) are BIOS limits and 16 is WD/IDE/ATA controller > register set limit. > >The second I will pass up, as I do not know the data structures involved. I am aware of the 2^31-1 = 2147483647 that crops up in a lot of places, when programmers use a signed 32-bit int to represent something. It is present in lots of file systems (SunOS 4 ffs, irix pre 4.0.5 ffs) and I have seen it in SCSI disks and controllers. But I was unaware it was a PC bios bug per se, except for some boot code. >The third of those limits is 1024*255*63*512 = 8422686720 = 8032.5MByte. > Where all 1024, 255 and 63 are BIOS limits, nothing to do with LBA > or even IDE/ATA drives, SCSI also has this limit, when driven via an > PC BIOS. > > >> Thanks. It's nice to know that my memory isn't completely shot. >> >> So what caused the ~120G limit? > >That is the LBA Limit of 2^28*512 = 137438953472 = 131072MByte = 128GByte. -- mrr ###### From: Don Quixote Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Reply-To: donq+afc@eclipse.net Message-ID: References: <3E7B792B.1428428E@computer.org> <0t1q7vs5obkmnedr4oh6dn2rrpur6ak1h5@fishead.news.meowkitty.com> <6feq7v4ufqr8svusdcsesp67i1ofcvp1h9@4ax.com> <3E7D699C.5FAA8E22@computer.org> X-Newsreader: Forte Agent 1.93/32.576 English (US) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Virus: Hi! I'm a header virus. Copy me to your headers and I'll spread. X-Org-1: If the Organization: header does not read 'wanted', then this post X-Org-2: has been modified from it's originally posted form and should not X-Org-3: be trusted. Lines: 39 Date: Tue, 01 Apr 2003 11:45:59 -0500 NNTP-Posting-Host: 207.207.243.216 X-Complaints-To: news@netcarrier.com X-Trace: news.netcarrier.net 1049225656 207.207.243.216 (Tue, 01 Apr 2003 14:34:16 EST) NNTP-Posting-Date: Tue, 01 Apr 2003 14:34:16 EST Organization: NetCarrier Internet Services Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed1.cidera.com!Cidera!news.netcarrier.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134085 Sam Yorko might have said: >William Hamblen wrote: >> >> On Sat, 22 Mar 2003 20:05:45 -0500, Don Quixote >> wrote: >> >> >With regard to NT 4.0 I think the number is four, not two. Which is >> >also the limit of the partition size that can be created while >> >performing an NT installation. I think. I could be wrong, and >> >whether or not I am, someone will be along to correct me shortly. >> >> By default the x86 NT 4.0 installer makes a FAT16 partition at the >> beginning of the disk to install the operating system in and the >> largest size a FAT16 partition can have is 2 Gbytes. The Alpha >> version probably does things differently. > >Naw, even if you make the first partition NTFS during initial install, >the OS boot loader can't see anything past the first two Gig. I once >somehow managed to get a critical file past the 2 G limit, and I think I >had to boot into debug mode to fix it. Would you be saying that the boot loader is different after installation, then? Or that the boot loader is installed to the < 2G partition, with the remainder of the OS installed to wherever installer whim or MicroSoft chose? (I've had a few NT boxes, with NT on NTFS somewhere above 2G, but always with a FAT16 partition as the first partition. Sometimes the NTFS was an extended and sometimes a real partition, depending on moon phase and the colour of the candles included with the hardware.) Thanks. -- I'm watching this with the fascination I'd feel if I were watching roses sprout from my toilet; I certainly never expected to see it, and I really have trouble believing it will continue. Clifton, on Sprint cleaning house. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 02 Apr 03 12:16:42 GMT Organization: UltraNet Communications, Inc. Lines: 69 Message-ID: References: <3E7B792B.1428428E@computer.org> <0t1q7vs5obkmnedr4oh6dn2rrpur6ak1h5@fishead.news.meowkitty.com> <6feq7v4ufqr8svusdcsesp67i1ofcvp1h9@4ax.com> <3E7D699C.5FAA8E22@computer.org> <3E8A410D.BFC06CF8@computer.org> X-Trace: UmFuZG9tSVaHxwPBjc1xhz5SwqZtqUQmgg4+VoBQibqYXQ3gj8lXspcWPrLG+H/6 X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 2 Apr 2003 13:06:26 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!feed2.news.rcn.net!feed1.news.rcn.net!rcn!208-59-182-141 Xref: chonsp.franklin.ch alt.folklore.computers:134121 In article , Brian Inglis wrote: >On Tue, 01 Apr 2003 17:46:53 -0800 in alt.folklore.computers, Sam >Yorko wrote: > >>Don Quixote wrote: >>> >>> Sam Yorko might have said: >>> >>> >William Hamblen wrote: >>> >> >>> >> On Sat, 22 Mar 2003 20:05:45 -0500, Don Quixote >>> >> wrote: >>> >> >>> >> >With regard to NT 4.0 I think the number is four, not two. Which is >>> >> >also the limit of the partition size that can be created while >>> >> >performing an NT installation. I think. I could be wrong, and >>> >> >whether or not I am, someone will be along to correct me shortly. >>> >> >>> >> By default the x86 NT 4.0 installer makes a FAT16 partition at the >>> >> beginning of the disk to install the operating system in and the >>> >> largest size a FAT16 partition can have is 2 Gbytes. The Alpha >>> >> version probably does things differently. >>> > >>> >Naw, even if you make the first partition NTFS during initial install, >>> >the OS boot loader can't see anything past the first two Gig. I once >>> >somehow managed to get a critical file past the 2 G limit, and I think I >>> >had to boot into debug mode to fix it. >>> >>> Would you be saying that the boot loader is different after >>> installation, then? Or that the boot loader is installed to >>> the < 2G partition, with the remainder of the OS installed to >>> wherever installer whim or MicroSoft chose? >>> >>> (I've had a few NT boxes, with NT on NTFS somewhere above 2G, but >>> always with a FAT16 partition as the first partition. Sometimes the >>> NTFS was an extended and sometimes a real partition, depending on moon >>> phase and the colour of the candles included with the hardware.) >>> >>> Thanks. >> >>Um, I'm not quite sure I understand the assumptions behind your >>question, but I'll try to answer... >> >>The NT boot loader loads in quite a few files (using the native disk >>support), and I also get the impression that it loads in some critical >>drivers, as well. At some point, there is enough loaded in that the >>boot loader can just invoke what's been loaded. >> >>I believe that the installer assumes that it's doing the right thing, >>and installs all those critical files and drivers in the physical first >>2Gig of the hard drive. However, if you do a driver replacement after >>the initial install, the driver may be kicked up above the 2G limit >>(since the OS, when running, can access it all). >> >>Somehow, I got one of those components above the 2G limit, and NT was >>not happy. I had to boot into debug mode, and somehow rearrange things. > >Explains why MyCruft$oft is now suggesting creating a root^Wboot >partition of less than 2GB containing only Windows. Cutler never did do smart memory management. This is 11M-itis all the way. /BAH Subtract a hundred and four for e-mail. ###### From: Sam Yorko Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 01 Apr 2003 17:46:53 -0800 Lines: 54 Message-ID: <3E8A410D.BFC06CF8@computer.org> References: <3E7B792B.1428428E@computer.org> <0t1q7vs5obkmnedr4oh6dn2rrpur6ak1h5@fishead.news.meowkitty.com> <6feq7v4ufqr8svusdcsesp67i1ofcvp1h9@4ax.com> <3E7D699C.5FAA8E22@computer.org> NNTP-Posting-Host: symsj01.sj.symbol.com (63.145.233.34) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Trace: fu-berlin.de 1049247987 4281211 63.145.233.34 (16 [71567]) X-Mailer: Mozilla 4.79 [en] (Windows NT 5.0; U) X-Accept-Language: en Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!fu-berlin.de!uni-berlin.de!symsj01.sj.symbol.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134108 Don Quixote wrote: > > Sam Yorko might have said: > > >William Hamblen wrote: > >> > >> On Sat, 22 Mar 2003 20:05:45 -0500, Don Quixote > >> wrote: > >> > >> >With regard to NT 4.0 I think the number is four, not two. Which is > >> >also the limit of the partition size that can be created while > >> >performing an NT installation. I think. I could be wrong, and > >> >whether or not I am, someone will be along to correct me shortly. > >> > >> By default the x86 NT 4.0 installer makes a FAT16 partition at the > >> beginning of the disk to install the operating system in and the > >> largest size a FAT16 partition can have is 2 Gbytes. The Alpha > >> version probably does things differently. > > > >Naw, even if you make the first partition NTFS during initial install, > >the OS boot loader can't see anything past the first two Gig. I once > >somehow managed to get a critical file past the 2 G limit, and I think I > >had to boot into debug mode to fix it. > > Would you be saying that the boot loader is different after > installation, then? Or that the boot loader is installed to > the < 2G partition, with the remainder of the OS installed to > wherever installer whim or MicroSoft chose? > > (I've had a few NT boxes, with NT on NTFS somewhere above 2G, but > always with a FAT16 partition as the first partition. Sometimes the > NTFS was an extended and sometimes a real partition, depending on moon > phase and the colour of the candles included with the hardware.) > > Thanks. Um, I'm not quite sure I understand the assumptions behind your question, but I'll try to answer... The NT boot loader loads in quite a few files (using the native disk support), and I also get the impression that it loads in some critical drivers, as well. At some point, there is enough loaded in that the boot loader can just invoke what's been loaded. I believe that the installer assumes that it's doing the right thing, and installs all those critical files and drivers in the physical first 2Gig of the hard drive. However, if you do a driver replacement after the initial install, the driver may be kicked up above the 2G limit (since the OS, when running, can access it all). Somehow, I got one of those components above the 2G limit, and NT was not happy. I had to boot into debug mode, and somehow rearrange things. Sam ###### X-Trace-PostClient-IP: 68.147.131.211 From: Brian Inglis Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Organization: Systematic Software Reply-To: Brian.Inglis@SystematicSw.ab.ca Message-ID: References: <3E7B792B.1428428E@computer.org> <0t1q7vs5obkmnedr4oh6dn2rrpur6ak1h5@fishead.news.meowkitty.com> <6feq7v4ufqr8svusdcsesp67i1ofcvp1h9@4ax.com> <3E7D699C.5FAA8E22@computer.org> <3E8A410D.BFC06CF8@computer.org> X-Newsreader: Forte Agent 1.93/32.576 English (American) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 67 Date: Wed, 02 Apr 2003 02:07:18 GMT NNTP-Posting-Host: 24.71.223.147 X-Complaints-To: abuse@shaw.ca X-Trace: news3.calgary.shaw.ca 1049249238 24.71.223.147 (Tue, 01 Apr 2003 19:07:18 MST) NNTP-Posting-Date: Tue, 01 Apr 2003 19:07:18 MST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!uni-erlangen.de!news-nue1.dfn.de!news-lei1.dfn.de!newsfeed.freenet.de!194.168.4.91.MISMATCH!newspeer1-gui.server.ntli.net!ntli.net!cox.net!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news3.calgary.shaw.ca.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134181 On Tue, 01 Apr 2003 17:46:53 -0800 in alt.folklore.computers, Sam Yorko wrote: >Don Quixote wrote: >> >> Sam Yorko might have said: >> >> >William Hamblen wrote: >> >> >> >> On Sat, 22 Mar 2003 20:05:45 -0500, Don Quixote >> >> wrote: >> >> >> >> >With regard to NT 4.0 I think the number is four, not two. Which is >> >> >also the limit of the partition size that can be created while >> >> >performing an NT installation. I think. I could be wrong, and >> >> >whether or not I am, someone will be along to correct me shortly. >> >> >> >> By default the x86 NT 4.0 installer makes a FAT16 partition at the >> >> beginning of the disk to install the operating system in and the >> >> largest size a FAT16 partition can have is 2 Gbytes. The Alpha >> >> version probably does things differently. >> > >> >Naw, even if you make the first partition NTFS during initial install, >> >the OS boot loader can't see anything past the first two Gig. I once >> >somehow managed to get a critical file past the 2 G limit, and I think I >> >had to boot into debug mode to fix it. >> >> Would you be saying that the boot loader is different after >> installation, then? Or that the boot loader is installed to >> the < 2G partition, with the remainder of the OS installed to >> wherever installer whim or MicroSoft chose? >> >> (I've had a few NT boxes, with NT on NTFS somewhere above 2G, but >> always with a FAT16 partition as the first partition. Sometimes the >> NTFS was an extended and sometimes a real partition, depending on moon >> phase and the colour of the candles included with the hardware.) >> >> Thanks. > >Um, I'm not quite sure I understand the assumptions behind your >question, but I'll try to answer... > >The NT boot loader loads in quite a few files (using the native disk >support), and I also get the impression that it loads in some critical >drivers, as well. At some point, there is enough loaded in that the >boot loader can just invoke what's been loaded. > >I believe that the installer assumes that it's doing the right thing, >and installs all those critical files and drivers in the physical first >2Gig of the hard drive. However, if you do a driver replacement after >the initial install, the driver may be kicked up above the 2G limit >(since the OS, when running, can access it all). > >Somehow, I got one of those components above the 2G limit, and NT was >not happy. I had to boot into debug mode, and somehow rearrange things. Explains why MyCruft$oft is now suggesting creating a root^Wboot partition of less than 2GB containing only Windows. Thanks. Take care, Brian Inglis Calgary, Alberta, Canada -- Brian.Inglis@CSi.com (Brian dot Inglis at SystematicSw dot ab dot ca) fake address use address above to reply abuse@aol.com tosspam@aol.com abuse@att.com abuse@earthlink.com abuse@hotmail.com abuse@mci.com abuse@msn.com abuse@sprint.com abuse@yahoo.com abuse@cadvision.com abuse@shaw.ca abuse@telus.com abuse@ibsystems.com uce@ftc.gov spam traps ###### From: "Helmut P. Einfalt" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 2 Apr 2003 05:08:13 -0000 Organization: T-Online Lines: 45 Message-ID: References: <3E7B792B.1428428E@computer.org> <0t1q7vs5obkmnedr4oh6dn2rrpur6ak1h5@fishead.news.meowkitty.com> <6feq7v4ufqr8svusdcsesp67i1ofcvp1h9@4ax.com> <3E7D699C.5FAA8E22@computer.org> <3E8A410D.BFC06CF8@computer.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-Trace: news.t-online.com 1049260034 00 16758 2CBwE2VESwAs1k 030402 05:07:14 X-Complaints-To: abuse@t-online.com X-ID: Xi09l2Z6QeyNMrwnwQ-7D9rBv6xzhqOH24V80iQ+eA4Fr+N8ETa78l X-Newsreader: Microsoft Outlook Express 4.72.3110.5 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!takemy.news.telefonica.de!telefonica.de!eusc.inter.net!newsfeed01.sul.t-online.de!newsmm00.sul.t-online.com!t-online.de!news.t-online.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134149 >>I believe that the installer assumes that it's doing the right thing, >>and installs all those critical files and drivers in the physical first >>2Gig of the hard drive. However, if you do a driver replacement after >>the initial install, the driver may be kicked up above the 2G limit >>(since the OS, when running, can access it all). >> >>Somehow, I got one of those components above the 2G limit, and NT was >>not happy. I had to boot into debug mode, and somehow rearrange things. > >Explains why MyCruft$oft is now suggesting creating a root^Wboot >partition of less than 2GB containing only Windows. Quoting Bruce Chambers directly from microsoft.public.winnt.setup: WinNT's setup _cannot_ create a partition larger than 4 Gb. If you wish to use a larger (up to 7.8 Gb) system partition, you must first create the partition and format it NTFS in another WinNT machine. Windows NT Boot Process and Hard Disk Constraints http://support.microsoft.com/support/kb/articles/Q114/8/41.asp? Windows NT 4.0 Supports Maximum of 7.8-GB System Partition http://support.microsoft.com/support/kb/articles/Q224/5/26.ASP Installing Windows NT on a Large IDE Hard Disk http://support.microsoft.com/directory/article.asp?ID=kb;en-us;Q197667 HTH Helmut ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 2 Apr 2003 15:14:11 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 46 Message-ID: References: <2qah7vc7dfm559m3hdpe6o0m0kmpve2ajr@4ax.com> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1049296451 27188 146.186.61.46 (2 Apr 2003 15:14:11 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Wed, 2 Apr 2003 15:14:11 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.cis.ohio-state.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134126 In article , J. Clarke wrote: >On Wed, 19 Mar 2003 17:52:04 +0000, Brian Inglis wrote: >> On Wed, 19 Mar 2003 17:09:58 +0000 (UTC) in alt.folklore.computers, >> hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) wrote: >>> It wasn't until later that I drew the connection between going in for >>>service there and the new problems that arose within a day or two. >>>plugging that loose wire (a factory test wire?) into the fuse block was >>>the real joy. We'd traced the entire electrical system and lifted the >>>upholstery trying to find the short that caused the fuse to blow when >>>the taillights went on. Suddenly inspiration hit: what if I remove >>>this wire, which seems to have no purpose? Amazing how much better >>>things work when you don't connect +12 to GND . . . >> Looked like they were doing their best to ensure you returned to give >> them more business. >> "Never attribute to malice what can be adequately explained by >> stupidity." >> A couple of problems like that is a definite indication that one, the >> other, or both, is occurrring. Hope you never went back there again, and >> spread the word. Yes, but I apparently didn't have to do so; they shut down not all that much afterwards. >Just remember, "Once is happenstance, twice is circumstance, three times >is enemy action." Yes, especially with multiple vehicles with the same experience :) Mmm, and then there was the time I pulled in with sqeaky brakes, and the not too bright guy lit up and said, "The boss usually lets me work on breaks." As the "we'll make it fit" ad flashed through my mind, I shot out *very* fast in reverse . . . hawk -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: pechter@shell.monmouth.com (Bill/Carolyn Pechter) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 2 Apr 2003 10:23:14 -0500 Organization: Lakewood MicroSystems Lines: 28 Message-ID: References: <3E8A410D.BFC06CF8@computer.org> NNTP-Posting-Host: shell.monmouth.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newspeer.monmouth.com!news.monmouth.com!shell.monmouth.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134147 In article , wrote: >Cutler never did do smart memory management. This is 11M-itis all >the way. > >/BAH Nah... this is the limitation of the old firmware drivers in the PC Bios and older IDE issues. This is all fixed in newer versions of NT (Windows 2000/XP) etc. Actually, I think the stuff was fixed in the WindowsNT service packs -- so if you could rebuild and remaster NT CD's it would work ok (if you took the limit out of the installer and used recent hardware. SCSI disks on PC's also only went to 2gb bootable in early Adaptec firmware. You can blame a lot on Cutler... but bad PC hardware design choices aren't his fault. Bill -- +---------------------------------------------------------------------------+ | Bill and/or Carolyn Pechter | pechter@shell.monmouth.com | | Bill Gates is a Persian cat and a monocle away from being a villain in | | a James Bond movie -- Dennis Miller | +---------------------------------------------------------------------------+ ###### Path: chonsp.franklin.ch!not-for-mail From: Neil Franklin Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: 02 Apr 2003 21:54:56 +0200 Organization: My own Private Self Lines: 50 Message-ID: <6u4r5gwu7j.fsf@chonsp.franklin.ch> References: <3E8A410D.BFC06CF8@computer.org> NNTP-Posting-Host: chonsp.franklin.ch X-Trace: chonsp.franklin.ch 1049313296 1186 10.0.3.2 (2 Apr 2003 19:54:56 GMT) X-Complaints-To: news@chonsp.franklin.ch NNTP-Posting-Date: 2 Apr 2003 19:54:56 GMT X-Newsreader: Gnus v5.7/Emacs 20.4 Xref: chonsp.franklin.ch alt.folklore.computers:134195 pechter@shell.monmouth.com (Bill/Carolyn Pechter) writes: > In article , wrote: > > >Cutler never did do smart memory management. This is 11M-itis all > >the way. > > Nah... this is the limitation of the old firmware drivers in the PC > Bios and older IDE issues. Yup. > SCSI disks on PC's also only went to 2gb bootable in early Adaptec > firmware. 1GB = 1024 cylinders * 64 tracks * 32 sect * 512 bytes The 1024 is a PC BIOS limit. The 64 and 32 were arbitrarily chosen, to with the 512 make neat 1MB/cylinder, and to allow simple bit shifting to convert cyl/track/sect to SCSI block number. Later SCSI BIOSes on PCs switched to the 1024/255/63 = <8GByte that IDE also uses. Above that required the extended BIOS interface. PCs are crap! PCs are crap! PCs are crap! PCs are crap! PCs are crap! PCs are crap! PCs are crap! PCs are crap! > You can blame a lot on Cutler... but bad PC hardware design > choices aren't his fault. There far worse cooks were spoiling it, while Cutler was still at DEC making VAXen. -- Neil Franklin, neil@franklin.ch.remove http://neil.franklin.ch/ Hacker, Unix Guru, El Eng HTL/BSc, Programmer, Archer, Blacksmith - hardware runs the world, software controls the hardware code generates the software, have you coded today? ###### Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Reply-To: Anne & Lynn Wheeler From: Anne & Lynn Wheeler Message-ID: Organization: Wheeler&Wheeler User-Agent: Gnus/5.090015 (Oort Gnus v0.15) Emacs/21.2 References: <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> Cancel-Lock: sha1:oE5KIbllgu6M7hLYemYXLUDJ1Jk= MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Lines: 38 Date: Fri, 04 Apr 2003 15:29:14 GMT NNTP-Posting-Host: 63.211.247.152 X-Complaints-To: abuse@earthlink.net X-Trace: newsread2.prod.itd.earthlink.net 1049470154 63.211.247.152 (Fri, 04 Apr 2003 07:29:14 PST) NNTP-Posting-Date: Fri, 04 Apr 2003 07:29:14 PST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed2.earthlink.net!newsfeed.earthlink.net!stamper.news.pas.earthlink.net!newsread2.prod.itd.earthlink.net.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134391 "J. Clarke" writes: > Hyperthreading on the P4 is not prefetching. It relies on the fact that > the pipeline in the P4 is quite long with some duplication of function > (remember, it supports predictive branching), much of which experiences a > certain amount of idle time. So hyperthreading runs two instructions > through the pipeline simultaneously--if neither hits an operation for > which the other has the hardware tied up then it is literally processing > two instructions simultaneously. > > It's not a dual CPU because there is contention for some of the > micro-operations (at least that is my understanding) but there is a > performance benefit--instead of a dual CPU machine, from a performance > viewpoint think of it as a fractional CPU machine--1.5 CPUs say. this type of thing was worked on 30 years ago for 370/195 ... which never made it to market. there was 63 instruction pipeline and any branch (not to an instruction already in the pipeline) drained the pipeline. normal codes rarely kept the pipeline more than half full. basically, the implementation duplicated the registers and the psw ... and added a bit to the instructions in the pipeline indicating which i-stream the instruction was associated with. random past refs to 195 dual-istream project: http://www.garlic.com/~lynn/94.html#38 IBM 370/195 http://www.garlic.com/~lynn/99.html#73 The Chronology http://www.garlic.com/~lynn/99.html#97 Power4 = 2 cpu's on die? http://www.garlic.com/~lynn/2000g.html#15 360/370 instruction cycle time http://www.garlic.com/~lynn/2001b.html#38 Why SMP at all anymore? http://www.garlic.com/~lynn/2001c.html#1 Z/90, S/390, 370/ESA (slightly off topic) http://www.garlic.com/~lynn/2001j.html#27 Pentium 4 SMT "Hyperthreading" http://www.garlic.com/~lynn/2001n.html#63 Hyper-Threading Technology - Intel information. http://www.garlic.com/~lynn/2002g.html#70 Pipelining in the past http://www.garlic.com/~lynn/2002g.html#76 Pipelining in the past http://www.garlic.com/~lynn/2002p.html#58 AMP vs SMP -- Anne & Lynn Wheeler | http://www.garlic.com/~lynn/ Internet trivia 20th anv http://www.garlic.com/~lynn/rfcietff.htm ###### From: "J. Clarke" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 04 Apr 2003 08:49:10 -0500 Organization: http://extra.newsguy.com Lines: 21 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <20030306064252.64caca77.steveo@eircom.net> NNTP-Posting-Host: p-968.newsdawg.com Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Pan/0.13.4 (She had eyes like strange sins.) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!zen.net.uk!diablo.theplanet.net!newsfeed.icl.net!newsfeed.fjserv.net!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews2 Xref: chonsp.franklin.ch alt.folklore.computers:134442 On Thu, 06 Mar 2003 06:42:52 +0100, Steve O'Hara-Smith wrote: > On Thu, 6 Mar 2003 01:27:58 +0000 > Ben Hutchings wrote: > > BH> in limits of 128 GB and 2 TB respectively. Unfortunately IDE disks > BH> are already pushing the lower of these. > > Pushing ? They went through it some time ago, my favorite PC > bits sypplier doesn't list anything smaller than 60GB. There are three > at 200GB and one at 180GB (all ATA). The biggest drive (I am amazed to > discover) though is apparently a 500GB LaCie firewire drive. I'd hate to > back that up on 9-track :) The Lacie is not a drive, it's two 250 GB drives in a RAID-0 array. -- --John Reply to jclarke at ae tee tee global dot net (was jclarke at eye bee em dot net) ###### From: "J. Clarke" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 04 Apr 2003 09:17:58 -0500 Organization: http://extra.newsguy.com Lines: 223 Message-ID: References: <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> NNTP-Posting-Host: p-154.newsdawg.com Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Pan/0.13.4 (She had eyes like strange sins.) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews1 Xref: chonsp.franklin.ch alt.folklore.computers:134439 On Sun, 09 Mar 2003 11:36:23 +0000, jmfbahci wrote: > In article , > Charles Shannon Hendrix wrote: >>In article , jmfbahciv@aol.com wrote: >> >>>>>>Some filesystems are almost that big, like SGI's xfs. >>>>> >>>>> You don't need the file system in core to have one. >>>> >>>>Yes you do. There is no way to do it otherwise. >>>> >>>>Filesystems in any non-trivial system are never idle. If you tried to >>>>keep the code out of core, you'd never get any work done. >>> >>> I didn't say "code"; I said "file system". You don't need the >>> data for handling a device until somebody pushes the on-line button. >> >>Sorry, I had thought it obvious I was talking about filesystem code, not >>the filesystem itself. > > Aha! I misread your post. > > >>Filesystem data (metadata) is read as needed of course. Most of the >>time you want some of it in core all the time for performance reasons. > > The only time is if the file contains some form of data base that > requires inplace management of the bits. "File" != "File System" Some of the data contained in the file system is generally kept resident for performance reasons--this would be directory information and other data necessary to locate data on the disk, and possibly other data relating to the jouranalling system if one is in place. If you have access to an MS-DOS (not Windows of any kind) system this is easy to demonstrate--set "BUFFERS=0" in config.sys, boot up, and copy a diskette. Note the time required. Then do the same with "BUFFERS=99" and note the time, and you will see a considerable improvement. The reason for this is that with no data buffered the drive has to make two seeks per segment, one to get the next-segment information from the FAT, and the next to get the data. If you listen you can hear this all going on. >>In fact, you'll usually have a lot of cached filesystem information. If >>memory isn't in use by a program, there is no sense in letting it sit >>idle when it can be used to speed I/O operations. > > Sure. But those I/O ops are the ones happening now. One sector may contain enough pointer information for a hundred or more subsequent accesses, so there would be a performance penalty if one did not buffer that sector the first time it was read--if one did not do that then the machine would spend a lot of time waiting for the disk that it did not need to spend. >>>>> I have yet to meet a CPU that can deal with more than one thing at a >>>>> time (that's why we have multiple-CPU systems). >>>> >>>>????? >>> >>> CPUs are synchronous machines. The art of timesharing OSes is make >>> them appear to be able to do everything at once. >> >>Actually, they aren't all like that. Some of them now have >>hyperthreading: more than one instruction stream per CPU. > > This is getting into a grey area :-). I don't consider prefetching an > execution. I just consider that a set up for the computation. Things > have really been hazy with microcode involved. I won't talk about this > one since FMPOV it gets into religion and semantics. Hyperthreading on the P4 is not prefetching. It relies on the fact that the pipeline in the P4 is quite long with some duplication of function (remember, it supports predictive branching), much of which experiences a certain amount of idle time. So hyperthreading runs two instructions through the pipeline simultaneously--if neither hits an operation for which the other has the hardware tied up then it is literally processing two instructions simultaneously. It's not a dual CPU because there is contention for some of the micro-operations (at least that is my understanding) but there is a performance benefit--instead of a dual CPU machine, from a performance viewpoint think of it as a fractional CPU machine--1.5 CPUs say. >>The art of a lot of OSs is the same BTW. > > I had always assumed that this was true until certain releases :-(. > >> .. Basically any system which >>allows the CPU to be interrupted is trying to make it as invisible as >>possible. > > I'd use the word efficient. It's just struck me that "invisible" > doesn't mean what we used to mean anymore. > > >>Timesharing just brought that out to interaction, some with more success >>than others. >> >>> Scenario: /BAH starts up a task that requires the disk structure BARB >>> to be spinning. So the pack is mounted, on-line procedures are >>> completed, and /BAH goes back to office to massage a few bits. A >>> priority interrupt is generated by a PHB or a bit god or a next door >>> neighbor who just did an Ooops!. Until I do a DIRECT >>> BARB:[10,2162]FOO.*, the data needed to access the bits doesn't have >>> to be in core. This is a case where service on demand is a more >>> efficient use of core. >> >>This is not true on a UNIX system, or just about any current system. >> >>BARB is likely an IDE or SCSI hard drive or removable media. The >>drivers and filesystem code are already in memory because other drives >>and filesystems, including those needed to boot the system, are already >>in use. > > Yes. I'm not advocating a separate copy of code/job/device. > > >>Unless BARB happens to be on a drive on a unique type of controller and >>filesystem, it's just not very bloody likely that this code has not >>already been loaded. >> >>Perhaps an example or two will help: >> >>Now, on my system just now, I've not used my SCSI CD-ROM in days. When I >>put in a CD and mounted it, several things happened in the system: >> >>* the EATA SCSI driver is loaded (main controller is Adaptec) * the bus >>is initialized >>* the CD-ROM driver is loaded >>* the iso9660 filesystem code is loaded * the volume is mounted at >>/mnt/cdrom, which includes reading >> and checking at least some filesystem metadata >> >>No further actions will cause any other drivers to be loaded. My system >>is not set up to unload drivers not in use. Just my preference. Note: I >>now have dual Adaptec SCSI controllers, so the drivers is already in >>memory for both cards upon OS boot. > > OK. I think we may be talking past each again. There are days that I > hate medium. >> >>Now another example. Let's say I mount a new SCSI hard drive. When I >>mount this the system will: >> >>* mount the volume at /, which includes reading and >> checking at least some filesystem metadata >> >>The controller drivers and filesystem code are already in core, because >>this code is needed for other mounted systems. >> >>The first case I listed above is going to be VERY rare. > > It is? I would think that the last case (new device installed) would be > the rare beast. > > >>> My point is that no OS will know when or how much data is going to be >>> used. >> >>True, which is why there has been so much work on unified memory >>management in UNIX over the last ten years. > > Unified memory management? I don't think I know what you mean by this. > > >>>>Most OS routines are going to be in use almost 24/7, so I don't see >>>>much benefit in making them swappable. >>> >>> No, they're not. If you've got a compute-bound job that does no I/O, >>> the device driver code will NEVER, I mean never, be executed. >> >>This compute bound job... how did it get loaded? > > With a RUN command 30 days ago. (I'm exaggerating slightly.) > > >>The system it is running on, how did it boot and start its other jobs >>(surely it has other work to do) without loading device drivers, >>filesystem code, etc? >> >>Now for older systems, it might be more common to have this situation, >>but most modern systems are going to be pretty busy, and even >>compute-bound jobs cause at least enough I/O to get loaded. > > Sure but they're compute-bound; that, by definition, implies that the > only computing serice they require is CPU. They're not going to do I/O > by design since that takes away heavy computation time. Just because > they had some I/O at the beginning of run, it makes absolutely no sense > to keep all of that code and/or data to start the program in core using > up address space. This is all valid for a machine that is doing batch processing. Very few machines are used in that fashion any longer--while one job might not be doing I/O, another very likely will or will need to do so in the near future. >>>>Now, your USB drivers... they might only get used once in awhile. >>>>Lot's of things fall into that catagory. But the bulk of a UNIX OS >>>>and its drivers are going to be in use all the time. There just won't >>>>be a time of non-use long enough to justify moving it out. >>> >>> This tells me that you've never dealt with a real timesharing OS at >>> its instruction level. >> >>Nope, I think something has confused you. It might be my fault, I don't >>know. > > Both our faults :-). We're mixing new gear with old resource management > techniques in the conversation. And it probably has to do with the > invention of "smart" controllers. > > /BAH > > Subtract a hundred and four for e-mail. -- --John Reply to jclarke at ae tee tee global dot net (was jclarke at eye bee em dot net) ###### From: "J. Clarke" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Fri, 04 Apr 2003 09:30:47 -0500 Organization: http://extra.newsguy.com Lines: 66 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> NNTP-Posting-Host: p-531.newsdawg.com Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Pan/0.13.4 (She had eyes like strange sins.) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.he.net!cyclone-sf.pbi.net!129.250.175.17!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews2 Xref: chonsp.franklin.ch alt.folklore.computers:134433 On Wed, 05 Mar 2003 18:22:51 -0500, Charles Shannon Hendrix wrote: > In article > , Ben > Hutchings wrote: >> In article , Charles Shannon >> Hendrix wrote: >> >>> A PC BIOS can only boot an image up to 2.88MB, the size of the highest >>> density 3.5" floppy. This is true even for a CD boot. The boot image >>> on the CD is also limited to 2.88MB. >> >> >> A PC BIOS only reads the first sector off the boot disk. It's up to >> the code squeezed into those 512 bytes (446 in the case of a hard disk, >> as 66 bytes are used for the partition table) to tell it to read >> additional sectors. There is a limit on the extent of a disk >> accessible through the BIOS routines, but it's much higher than 2.88MB. > > I don't think so. > > It is true the BIOS only knows how to load the 512-byte boot code. > > This code is loaded to a specific location and is executed in real mode. > Due to the small size, bare-metal programming to get out of real mode is > nearly impossible. > > This code needs to use BIOS functions to jump to the next stage, and > these functions have limitations in placement and size of the next > stage. > > The exact limitations aren't even the same across all BIOS, which makes > the problem even more "interesting". > > I was pretty sure this second stage could only be 2.88MB, and I know > that's the case on a CD-ROM. > > I couldn't find a reference to the limit of this state anywhere. > > What I don't understand is why a modern PC BIOS doesn't have a "No DOS" > flag for when you don't need the 640K split and other DOS/Windows hooks, > and why they don't have a 32-bit native loader after all these years. A BIOS flag wouldn't really help the 640k split as there may be memory-mapped hardware immediately above that point and there almost certainly is at 704k and above (video is mapped to that general region), and there is no mechanism established which can be counted on to identify which locations between 640k and 1 meg are free. With "legacy-free" machines this might be doable, but they're rather recent innovations and very honestly I wouldn't trust the same mechanism that supports plug-and-pray to report this information accurately. The 32-bit loader would have to deal with 16-bit code, and generally speaking the Intel processors are a lot happier going from real to protected mode than they are going from protected to real mode, so building a 32-bit loader into the BIOS would be of questionable utility--either it would cause problems if you needed to load a 16-bit OS or there would have to be some kind of recognized flag on the disk to indicate that the 32-bit loader was to be activated. -- --John Reply to jclarke at ae tee tee global dot net (was jclarke at eye bee em dot net) ###### Date: Fri, 4 Apr 2003 18:59:57 +0200 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <20030404185957.7cb2c41c.steveo@eircom.net> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <20030306064252.64caca77.steveo@eircom.net> X-Newsreader: Sylpheed version 0.8.10 (GTK+ 1.2.10; i386-portbld-freebsd4.7) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 18 Organization: EuroNet Internet NNTP-Posting-Date: 04 Apr 2003 17:00:07 GMT NNTP-Posting-Host: 62.234.211.60 X-Trace: 1049475607 maya.euronet.nl 1775 62.234.211.60:2616 X-Complaints-To: abuse@euronet.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!newsfeed1.bredband.com!bredband!amsnews01.chello.com!news2.euro.net!postnews1.euro.net!news.euronet.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134432 On Fri, 04 Apr 2003 08:49:10 -0500 "J. Clarke" wrote: JC> On Thu, 06 Mar 2003 06:42:52 +0100, Steve O'Hara-Smith wrote: JC> > three at 200GB and one at 180GB (all ATA). The biggest drive (I am JC> > amazed to discover) though is apparently a 500GB LaCie firewire JC> > drive. I'd hate to back that up on 9-track :) JC> JC> The Lacie is not a drive, it's two 250 GB drives in a RAID-0 array. I had wondered if it might be something like that - too lazy to check :) -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 39 Date: Fri, 4 Apr 2003 23:33:54 -0500 NNTP-Posting-Host: 209.96.179.79 X-Trace: sydney.visi.net 1049520521 209.96.179.79 (Sat, 05 Apr 2003 00:28:41 EST) NNTP-Posting-Date: Sat, 05 Apr 2003 00:28:41 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!takemy.news.telefonica.de!telefonica.de!feed.news.schlund.de!schlund.de!newsfeed01.sul.t-online.de!t-online.de!newspeer1-gui.server.ntli.net!ntli.net!cox.net!cyclone1.gnilink.net!wn14feed!worldnet.att.net!199.184.165.233!feed2.news.rcn.net!rcn!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:134590 In article , J. Clarke wrote: > A BIOS flag wouldn't really help the 640k split as there may be > memory-mapped hardware immediately above that point and there almost > certainly is at 704k and above (video is mapped to that general region), Yes, but that's precisely because BIOS is in DOS mode. If it were not, there would be no need for these ancient fixed memory locations. PCs could finally catch up to microcomputers of the early 80s. > very honestly I wouldn't trust the same mechanism that supports > plug-and-pray to report this information accurately. That's mostly a BIOS problem, and a failure to follow specifications. Besides, there are better systems out there. Plug and Play works perfectly on my system, though I've certainly had systems where it did not. > The 32-bit loader would have to deal with 16-bit code, and generally > speaking the Intel processors are a lot happier going from real to No it wouldn't. That's my point, get rid of all real mode, period. Certainly none of the UNIX systems would need real mode. None of my hardware needs real mode. Basically, real mode is just in the way on my system. That's why you make this an option at first. In time flash upgrades will make sure expansion cards no lnoger care about real mode. Not an overnight change, but certainly we should have long ago started making the move away. ###### From: "J. Clarke" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Sat, 05 Apr 2003 00:55:52 -0500 Organization: http://extra.newsguy.com Lines: 63 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> NNTP-Posting-Host: p-822.newsdawg.com Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Pan/0.13.4 (She had eyes like strange sins.) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!nntp-out.monmouth.com!newspeer.monmouth.com!nntp1.phx1.gblx.net!nntp.gblx.net!nntp.gblx.net!pln-w!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews2 Xref: chonsp.franklin.ch alt.folklore.computers:134563 On Fri, 04 Apr 2003 23:33:54 -0500, Charles Shannon Hendrix wrote: > In article , J. Clarke > wrote: > >> A BIOS flag wouldn't really help the 640k split as there may be >> memory-mapped hardware immediately above that point and there almost >> certainly is at 704k and above (video is mapped to that general >> region), > > Yes, but that's precisely because BIOS is in DOS mode. If it were not, > there would be no need for these ancient fixed memory locations. However in that case you would have to virtualize the hardware, which is a bit much to expect from a boot loader. > PCs could finally catch up to microcomputers of the early 80s. > >> very honestly I wouldn't trust the same mechanism that supports >> plug-and-pray to report this information accurately. > > That's mostly a BIOS problem, and a failure to follow specifications. > > Besides, there are better systems out there. > > Plug and Play works perfectly on my system, though I've certainly had > systems where it did not. It works perfectly until there is time pressure, at which point it provides an excellent demonstration of the perversity of inanimate objects. >> The 32-bit loader would have to deal with 16-bit code, and generally >> speaking the Intel processors are a lot happier going from real to > > No it wouldn't. That's my point, get rid of all real mode, period. > > Certainly none of the UNIX systems would need real mode. > > None of my hardware needs real mode. Basically, real mode is just in > the way on my system. > > That's why you make this an option at first. > > In time flash upgrades will make sure expansion cards no lnoger care > about real mode. That includes expansion cards whose vendors have ceased to support them? > Not an overnight change, but certainly we should have long ago started > making the move away. Why? It is a minor inconvenience at worst. If you want a machine that is not a PC then get a Mac or an Alpha or a Z-Architecture or an RS/6000 or whatever floats your boat. But changing the design of the PC so that DOS won't run is not something that I find that there is any particularly compelling need to do. -- --John Reply to jclarke at ae tee tee global dot net (was jclarke at eye bee em dot net) ###### Newsgroups: alt.folklore.computers From: Charles Shannon Hendrix Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> Organization: Big Endian Reply-To: cshNOSPAM@NOSPAM.widomaker.com Message-ID: User-Agent: slrn/0.9.7.4 (Linux) Lines: 44 Date: Sun, 6 Apr 2003 18:57:06 -0400 NNTP-Posting-Host: 206.246.249.189 X-Trace: sydney.visi.net 1049671713 206.246.249.189 (Sun, 06 Apr 2003 19:28:33 EDT) NNTP-Posting-Date: Sun, 06 Apr 2003 19:28:33 EDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!luth.se!news.stupi.se!news.stupi.se!news.tele.dk!news.tele.dk!small.news.tele.dk!newsfeed.news2me.com!newshosting.com!news-xfer2.atl.newshosting.com!newsxfer.visi.net!sydney.visi.net!escape.shannon.net!nobody Xref: chonsp.franklin.ch alt.folklore.computers:134717 In article , J. Clarke wrote: > However in that case you would have to virtualize the hardware, which is a > bit much to expect from a boot loader. Even 80s era micros managed to handle this sort of thing. >> That's why you make this an option at first. >> >> In time flash upgrades will make sure expansion cards no lnoger care >> about real mode. > > That includes expansion cards whose vendors have ceased to support them? Irrelevant. There is no way to avoid problems with older hardware. At some point all systems have to change. >> Not an overnight change, but certainly we should have long ago started >> making the move away. > > Why? It is a minor inconvenience at worst. No, it's a major problem. Board vendors in particular hate it. > If you want a machine that is > not a PC then get a Mac or an Alpha or a Z-Architecture or an RS/6000 or > whatever floats your boat. ...or a PC which has flaws corrected. That's what going to happen, slowly, so we'll get there eventually. > But changing the design of the PC so that DOS won't run is not > something that I find that there is any particularly compelling need > to do. To you maybe. To a lot of other people, we can't wait to see this changed. Jeez, there was better hardware that mysteriously got around the problems you've mentioned back in the 80s. We certainly can do even better now. My 1986 Atari ST made more sense than a modern PC. ###### From: "J. Clarke" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Mon, 07 Apr 2003 00:03:08 -0400 Organization: http://extra.newsguy.com Lines: 70 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> NNTP-Posting-Host: p-446.newsdawg.com Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Pan/0.13.4 (She had eyes like strange sins.) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews1 Xref: chonsp.franklin.ch alt.folklore.computers:134704 On Sun, 06 Apr 2003 18:57:06 -0400, Charles Shannon Hendrix wrote: > In article , J. Clarke wrote: > >> However in that case you would have to virtualize the hardware, which is a >> bit much to expect from a boot loader. > > Even 80s era micros managed to handle this sort of thing. 80s era micros that might be booting from a third-party device? A lot of things that are easy in a single-vendor system can become fiendishly difficult in a PC. >>> That's why you make this an option at first. >>> >>> In time flash upgrades will make sure expansion cards no lnoger care >>> about real mode. >> >> That includes expansion cards whose vendors have ceased to support them? > > Irrelevant. There is no way to avoid problems with older hardware. At > some point all systems have to change. > >>> Not an overnight change, but certainly we should have long ago started >>> making the move away. >> >> Why? It is a minor inconvenience at worst. > > No, it's a major problem. Board vendors in particular hate it. I'm not quite sure I understand why board vendors other than those who sell proprietary storage controllers would give a damn about the boot loader. >> If you want a machine that is >> not a PC then get a Mac or an Alpha or a Z-Architecture or an RS/6000 or >> whatever floats your boat. > > ...or a PC which has flaws corrected. That's what going to happen, > slowly, so we'll get there eventually. Or just introduce a new set. >> But changing the design of the PC so that DOS won't run is not >> something that I find that there is any particularly compelling need >> to do. > > To you maybe. To a lot of other people, we can't wait to see this > changed. For certain values of "lot". Or perhaps for certain values of "compelling need". > Jeez, there was better hardware that mysteriously got around the > problems you've mentioned back in the 80s. We certainly can do even > better now. Better hardware that could be assembled with no two components coming from the same vendor? > My 1986 Atari ST made more sense than a modern PC. So why aren't there hundreds of millions of Atari STs out there? Technical merit doesn't sell machines in shipload lots. -- --John Reply to jclarke at ae tee tee global dot net (was jclarke at eye bee em dot net) ###### From: Pete Fenelon Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 08 Apr 2003 11:02:01 -0000 Organization: Posted via Supernews, http://www.supernews.com Message-ID: Sender: Pete Fenelon References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E92134E.DC87AEAD@ev1.net> User-Agent: tin/1.5.16-20021229 ("Spiders") (UNIX) (FreeBSD/4.8-STABLE (i386)) X-Complaints-To: abuse@supernews.com Lines: 23 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!snoopy.risq.qc.ca!newsfeed.news2me.com!newsfeed2.easynews.com!newsfeed1.easynews.com!easynews.com!easynews!sn-xit-02!sn-xit-06!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134827 Charles Richmond wrote: > > Back in the good ole USENET days...circa twelve years ago, > the groups and > were *packed* with software written by those hackers and > freely dispensed into the public domain. > I still miss my STs - spent about six years using them for work and play on a daily basis. Best thing about them was those perfect keyboards and lovely mono screens... (oh and Bubble Bobble and Balance of Power - but I've got them on Mame and Billy$hit respectively these days ;)) Did a lot of very enjoyable hacking on them - they were cheap machines made of simple components, but you could really get a lot out of them with a bit of ingenuity. Whereas the Amiga was all about learning how to do weird things with exotic hardware, the Atari was just a straightforward and very programmer-friendly little micro. pete -- pete@fenelon.com "there's no room for enigmas in built-up areas" HMHB ###### From: "Rupert Pigott" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 8 Apr 2003 13:32:40 +0100 Organization: Titanic Enterprises Unlimited Lines: 18 Message-ID: <1049808199.842340@saucer.planet.gong> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E92134E.DC87AEAD@ev1.net> NNTP-Posting-Host: darkboong.demon.co.uk X-Trace: news.demon.co.uk 1049808203 13426 80.177.7.220 (8 Apr 2003 13:23:23 GMT) X-Complaints-To: abuse@demon.net NNTP-Posting-Date: Tue, 8 Apr 2003 13:23:23 +0000 (UTC) X-MIMEOLE: Produced By Microsoft MimeOLE V6.00.2800.1106 X-Priority: 3 X-Cache: nntpcache 3.0.1 (see http://www.nntpcache.org/) X-Newsreader: Microsoft Outlook Express 6.00.2800.1106 X-MSMail-Priority: Normal Cache-Post-Path: saucer.planet.gong!unknown@voodoo.planet.gong Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.imp.ch!news.imp.ch!solnet.ch!solnet.ch!newsfeed.freenet.de!newsfeed00.sul.t-online.de!t-online.de!kibo.news.demon.net!news.demon.co.uk!demon!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134816 "Pete Fenelon" wrote in message news:v95b19ofh83971@corp.supernews.com... [SNIP] > with a bit of ingenuity. Whereas the Amiga was all about learning how > to do weird things with exotic hardware, the Atari was just a > straightforward and very programmer-friendly little micro. I didn't find the Amiga's hardware too "exotic". I had a fairly good handle on it as a PFY without too much sweat. Besides, you could mostly hide behind the libraries anyways, no need to mess. I found them fairly easy to handle. Cheers, Rupert ###### From: Pete Fenelon Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 08 Apr 2003 17:16:31 -0000 Organization: Posted via Supernews, http://www.supernews.com Message-ID: Sender: Pete Fenelon References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E92134E.DC87AEAD@ev1.net> <1049808199.842340@saucer.planet.gong> User-Agent: tin/1.5.16-20021229 ("Spiders") (UNIX) (FreeBSD/4.8-STABLE (i386)) X-Complaints-To: abuse@supernews.com Lines: 23 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!takemy.news.telefonica.de!telefonica.de!feed.news.schlund.de!schlund.de!newsfeed01.sul.t-online.de!newsfeed00.sul.t-online.de!t-online.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!sn-xit-03!sn-xit-01!sn-post-01!supernews.com!corp.supernews.com!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134843 Rupert Pigott wrote: > "Pete Fenelon" wrote in message > news:v95b19ofh83971@corp.supernews.com... > [SNIP] > >> with a bit of ingenuity. Whereas the Amiga was all about learning how >> to do weird things with exotic hardware, the Atari was just a >> straightforward and very programmer-friendly little micro. > > I didn't find the Amiga's hardware too "exotic". I had a > fairly good handle on it as a PFY without too much sweat. > Besides, you could mostly hide behind the libraries > anyways, no need to mess. I found them fairly easy to > handle. Relative to what was on the Atari, the Agnus/Denise/Portia chipset on the Amiga was *bloody* exotic. In many respects the ST was a good 8-bit micro with a 32-bit CPU and a 16 bit databus ;) pete -- pete@fenelon.com "there's no room for enigmas in built-up areas" HMHB ###### From: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Tue, 8 Apr 2003 22:08:19 +0000 (UTC) Organization: Penn State University, Center for Academic Computing Lines: 29 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1049808199.842340@saucer.planet.gong> NNTP-Posting-Host: slytherin.ds.psu.edu X-Trace: f04n12.cac.psu.edu 1049839699 26704 146.186.61.46 (8 Apr 2003 22:08:19 GMT) X-Complaints-To: usenet@f04n12.cac.psu.edu NNTP-Posting-Date: Tue, 8 Apr 2003 22:08:19 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: hawk@slytherin.ds.psu.edu (Dr. Richard E. Hawkins) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.media.kyoto-u.ac.jp!headwall.stanford.edu!newsfeed.stanford.edu!news.ems.psu.edu!news.aset.psu.edu!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134915 In article , Pete Fenelon wrote: >Rupert Pigott wrote: >> "Pete Fenelon" wrote in message >> news:v95b19ofh83971@corp.supernews.com... >> [SNIP] >>> with a bit of ingenuity. Whereas the Amiga was all about learning how >>> to do weird things with exotic hardware, the Atari was just a >>> straightforward and very programmer-friendly little micro. >> I didn't find the Amiga's hardware too "exotic". I had a >> fairly good handle on it as a PFY without too much sweat. >> Besides, you could mostly hide behind the libraries >> anyways, no need to mess. I found them fairly easy to >> handle. >Relative to what was on the Atari, the Agnus/Denise/Portia chipset on >the Amiga was *bloody* exotic. In many respects the ST was a good 8-bit >micro with a 32-bit CPU and a 16 bit databus ;) *now* you tell me. WHy didn't anyone tell me this 20 years ago :) hawk, who hasn't found much since the 8 bitters interesting -- Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign dochawk@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail These opinions will not be those of X and postings. Penn State until it pays my retainer. / \ ###### From: "J. Clarke" Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Wed, 09 Apr 2003 14:37:01 -0400 Organization: http://extra.newsguy.com Lines: 35 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <3E5C026F.5080708@jetnet.ab.ca> <6k4i3b.hhc2.ln@via.reistad.priv.no> <3E92134E.DC87AEAD@ev1.net> Reply-To: jclarke@nospam.invalid NNTP-Posting-Host: p-710.newsdawg.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7Bit User-Agent: KNode/0.7.2 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!syros.belnet.be!news.belnet.be!newsfeed.stanford.edu!logbridge.uoregon.edu!pln-e!spln!dex!extra.newsguy.com!newsp.newsguy.com!enews4 Xref: chonsp.franklin.ch alt.folklore.computers:134989 Charles Richmond wrote: > "J. Clarke" wrote: >> >> [snip...] [snip...] [snip...] >> >> So why aren't there hundreds of millions of Atari STs out there? >> Technical merit doesn't sell machines in shipload lots. >> > The Atari ST is (was) a great little hacker machine. There just > were *not* hundreds of millions of hackers out there. You made the statement "To a lot of other people, we can't wait to see this changed." If your "lot of other people" is of the same magnitude as the number of persons who bought Atari STs then you need to reevaluate your definition of the term "lot" in the context of the current microprocessor market, which aims to sell tens or hundreds of millions of processors to people who are not hackers. In that context what hackers want is pretty much irrelevant if it conflicts with what the mass-market consumers want. > Back in the good ole USENET days...circa twelve years ago, > the groups and > were *packed* with software written by those hackers and > freely dispensed into the public domain. > > > -- > +----------------------------------------------------------------+ > | Charles and Francis Richmond richmond at plano dot net | > +----------------------------------------------------------------+ -- -- John Reply to jclarke at ae tee tee global dot net (used to be jclarke at eye bee em dot net) ###### Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030310200541.12ec6fa1.steveo@eircom.net> <90mo4b.f96.ln@teabag.cbhnet> <20030313081821.6893c1d7.steveo@eircom.net> Organization: Honest Chris' Sysadmin Emporium X-Newsreader: trn 4.0-test76 (Apr 2, 2001) From: cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) Originator: cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) Date: Wed, 9 Apr 2003 22:03:54 +0100 Message-ID: Lines: 34 NNTP-Posting-Date: 09 Apr 2003 21:10:59 GMT NNTP-Posting-Host: 62.254.134.35 X-Trace: 1049922659 news.gradwell.net 66673 cbh/62.254.134.35 X-Complaints-To: news-abuse@gradwell.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!takemy.news.telefonica.de!telefonica.de!eusc.inter.net!psinet-eu-nl!news.imp.ch!news.imp.ch!zen.net.uk!news.cabal.org.uk!news-peer.gradwell.net!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:134986 According to Steve O'Hara-Smith : > I notice that I forgot to mention, SSADM and Yourdon - the > mix with OO design must have been particularly confusing :) I don't think I've encountered Yourdon, and I'm happy to say that the whole design phase was out of the way (well, mostly) by the time I arrived. It was another of those things that'd already gained legendary status, not only because of its bizarreness but because of the notoriously heated arguments between various people called David. :) > By the time I left Telford for the joys of Basingstoke they > had divided into two groups. There were a few whose brains were still > functioning and were starting to integrate it all and become useful. > The others were given harmless side tasks and treated gently. There > were a couple of people pretty much full time looking after them by > then. I was later told that these two groups corresponded to those that > put themselves on the transfer list and those that were put on it. I think that the split was still very evident when I worked there; the members of the former group tended to gradually migrate to other interesting projects, the latter ones, er, didn't. It was still an uncomfortable experience for me, I hadn't realised that some git had sold me as some hot-shot programmer and dumped me in at the deep end, I didn't even have a beard and just thought "eh, woss going on?" Didn't help that the, er, lady who ran that team was _really_ into the CYA type of management of the most preemptive style. In spite of all that, it was quite fun to work on for a while. Chris. -- "If the world was an orange it would be like much too small, y'know?" Neil, '84 Currently playing: The Stranglers - "Shah Shah A Go Go" http://www.chrishedley.com My stuff, including genealogy, other things, etc ###### Date: Thu, 10 Apr 2003 18:57:03 +0200 From: Steve O'Hara-Smith Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Message-ID: <20030410185703.37da3511.steveo@eircom.net> References: <3E05FE9D.B3A12286@vax6k.openecs.org> <20030310200541.12ec6fa1.steveo@eircom.net> <90mo4b.f96.ln@teabag.cbhnet> <20030313081821.6893c1d7.steveo@eircom.net> X-Newsreader: Sylpheed version 0.8.11 (GTK+ 1.2.10; i386-portbld-freebsd4.8) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 49 Organization: Wanadoo NNTP-Posting-Date: 10 Apr 2003 17:00:07 GMT NNTP-Posting-Host: rot2-p0790.dial.wanadoo.nl X-Trace: 1049994007 maya.euronet.nl 1770 62.234.201.22:1553 X-Complaints-To: abuse@wanadoo.nl Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!news.maxwell.syr.edu!news.tele.dk!news.tele.dk!small.news.tele.dk!newsfeed.multikabel.nl!news2.euro.net!postnews1.euro.net!news.wanadoo.nl!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:135177 On Wed, 9 Apr 2003 22:03:54 +0100 cbh@ieya.co.REMOVE_THIS.uk (Chris Hedley) wrote: CH> According to Steve O'Hara-Smith : CH> > I notice that I forgot to mention, SSADM and Yourdon - the CH> > mix with OO design must have been particularly confusing :) CH> CH> I don't think I've encountered Yourdon, and I'm happy to say that Try to avoid it - AFAICT it is well suited to writing tape format converters and that's about it. The guy giving the course to us must have got fed up with the heckling about how inexpressive it was from some members of the team - we were not impressed. CH> the whole design phase was out of the way (well, mostly) by the CH> time I arrived. It was another of those things that'd already CH> gained legendary status, not only because of its bizarreness but CH> because of the notoriously heated arguments between various people CH> called David. :) There were more than two of them ? Was Jonathon involved too ? Anyway there were never arguments when I was there - just lively discussions :) CH> I think that the split was still very evident when I worked there; CH> the members of the former group tended to gradually migrate to other CH> interesting projects, the latter ones, er, didn't. It was still an I wandered off to pastures greener (well the Beeb) because the place seemed to have run out of interesting projects. CH> uncomfortable experience for me, I hadn't realised that some git had CH> sold me as some hot-shot programmer and dumped me in at the deep end, Hey everyone was sold into that place as a hot shot programmer, it's what they wanted to hear :) CH> I didn't even have a beard and just thought "eh, woss going on?" CH> Didn't help that the, er, lady who ran that team was _really_ into the Hmm, management deteriorated then - we always had good shit shields, as well as a first class shit dispenser of our own on call. But then Doug chose the managers directly in the early days. -- C:>WIN | Directable Mirrors The computer obeys and wins. |A Better Way To Focus The Sun You lose and Bill collects. | licenses available - see: | http://www.sohara.org/ ###### From: Larry Jones Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Newsgroups: alt.folklore.computers References: <6rrd6vof9eghuu8crkbc0dn50s6r3h10c7@4ax.com> User-Agent: tin/1.4.5-20010409 ("One More Nightmare") (UNIX) (BSD/OS/4.2 (i386)) Message-ID: Lines: 21 Date: Sat, 12 Apr 2003 04:20:41 GMT NNTP-Posting-Host: 65.27.189.87 X-Complaints-To: abuse@rr.com X-Trace: twister.neo.rr.com 1050121241 65.27.189.87 (Sat, 12 Apr 2003 00:20:41 EDT) NNTP-Posting-Date: Sat, 12 Apr 2003 00:20:41 EDT Organization: Road Runner High Speed Online -- Northeast Ohio Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!news.maxwell.syr.edu!newsfeed-east.nntpserver.com!nntpserver.com!news-west.rr.com!cyclone.kc.rr.com!cyclone2.kc.rr.com!news2.kc.rr.com!twister.neo.rr.com.POSTED!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:135203 J. Clarke wrote: > > Hyperthreading on the P4 is not prefetching. It relies on the fact that > the pipeline in the P4 is quite long with some duplication of function > (remember, it supports predictive branching), much of which experiences a > certain amount of idle time. So hyperthreading runs two instructions > through the pipeline simultaneously--if neither hits an operation for > which the other has the hardware tied up then it is literally processing > two instructions simultaneously. Hyperthreading on the P4 is a hack to try to make up for the fact that the front end of the processor is too slow to keep the back end busy due to leaving out big parts of the original design in a misguided attempt to get it out the door and thus stop AMD's claim to have the fastest processor. Turning a defect into a desirable feature is the pinnacle of marketing (c.f. Heinz and ketchup that's too thick to get out of the bottle). -Larry Jones I can feel my brain beginning to atrophy already. -- Calvin ###### From: Ben Hutchings Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC Date: Sun, 20 Apr 2003 01:36:40 +0100 Lines: 39 Message-ID: References: <1l8f5v8cf0tnjv9bc6lel9c25p9p2517p7@4ax.com> <3E5A710D.6030403@jetnet.ab.ca> <1frfh0t.nm8oe5dijljmN%lars@bearnip.com> <1frirq0.3dq1ns1yfxddsN%lars@bearnip.com> <1frznw0.8e1waqbp4y8N%lars@bearnip.com> <1fs43a6.pv28rg17wjp9oN%lars@bearnip.com> Reply-To: Ben Hutchings NNTP-Posting-Host: pc3-cmbg5-3-cust175.cmbg.cable.ntl.com (81.104.194.175) X-Trace: fu-berlin.de 1050798608 4355738 81.104.194.175 (16 [70929]) X-Orig-Path: decadentplace.org.uk!nobody User-Agent: slrn/0.9.7.4 (Linux) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!irazu.switch.ch!news.maxwell.syr.edu!newsfeed.icl.net!newsfeed.fjserv.net!feed.news.nacamar.de!fu-berlin.de!uni-berlin.de!pc3-cmbg5-3-cust175.cmbg.cable.ntl.COM!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:135730 In article <1fs43a6.pv28rg17wjp9oN%lars@bearnip.com>, Lars Duening wrote: > Charles Shannon Hendrix wrote: > >> In article <1frznw0.8e1waqbp4y8N%lars@bearnip.com>, Lars Duening wrote: >> > C++ implementations have the same problem, one magnitude larger, as the >> > standard library is rather complex. >> >> It seems to have a lot more dependency issues, probably because of >> the code sharing it encourages. >> >> I would have felt better if it was a series of independent classes with >> interfaces between them, so you could use them seperately. > > I have to take another look, but I seem to remember that the basic > algorithm and datatype classes are relatively independent. The IO > libraries on the other hand are . "...not"? The biggest problem, apparently, is that of locale support. Portable C++ programs almost invariably need iostreams, and they need locale support, and it's very hard to avoid pulling in the whole of the locale library. Here are the sizes of "Hello, world" programs when compiled by gcc 3.0 on Debian Linux for i386 and with debug information stripped: C - printf - dynamically linked: 3,008 C - printf - statically linked: 384,236 C++ - printf - dynamically linked: 3,372 C++ - printf - statically linked: 394,164 C++ - iostreams - dynamically linked: 3,420 C++ - iostreams - statically linked: 442,840 -- Ben Hutchings | personal web site: http://womble.decadentplace.org.uk/ Tomorrow will be cancelled due to lack of interest. ###### From: abuse@mooli.org.uk (Peter Corlett) Newsgroups: alt.folklore.computers Subject: Re: PDP10 and RISC (was Re: vax6k.openecs.org rebirth) Date: Mon, 28 Apr 2003 12:00:38 +0000 (UTC) Organization: cabal.org.uk news service, Leeds, UK. Lines: 10 Message-ID: References: <3E05FE9D.B3A12286@vax6k.openecs.org> <1049808199.842340@saucer.planet.gong> NNTP-Posting-Host: mooli.org.uk X-Trace: mooli.org.uk 1051531238 12330 195.92.99.54 (28 Apr 2003 12:00:38 GMT) X-Complaints-To: usenet@mooli.org.uk NNTP-Posting-Date: Mon, 28 Apr 2003 12:00:38 +0000 (UTC) X-Newsreader: trn 4.0-test76 (Apr 2, 2001) Originator: abuse@mooli.org.uk (Peter Corlett) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!takemy.news.telefonica.de!telefonica.de!news1.dtag.de!newsfeed00.sul.t-online.de!t-online.de!skynet.be!skynet.be!newsfeed.icl.net!newsfeed.fjserv.net!nntp.theplanet.net!inewsm1.nntp.theplanet.net!zen.net.uk!news.cabal.org.uk!not-for-mail Xref: chonsp.franklin.ch alt.folklore.computers:136229 Pete Fenelon wrote: [...] > Relative to what was on the Atari, the Agnus/Denise/Portia chipset on the > Amiga was *bloody* exotic. s/Portia/Paula/ http://www.amigarealm.com/archives/amiga-clinic/pediatrics/chipsets.htm gives a list of all the custom chips used in Amigas.