Newsgroups: alt.folklore.computers From: mlipman@iname.zzz.com (Marcel Lipman) Subject: Multics Sender: news@ahisinfr.xs4all.nl (news at janux) X-Newsreader: News Xpress 2.01 Organization: AH-ISS Message-ID: X-Nntp-Posting-Host: p1imwe04.wau.mis.ah.nl Date: Thu, 29 Oct 1998 11:51:01 GMT Lines: 12 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newshub.bart.net!newsgate.cistron.nl!het.net!news.cs.utwente.nl!xs4all!ahisinfr.xs4all.nl!p1imwe04 Hello, In articles about Unix, you always read that Unix was developed by the people who used to work on the Multics project. I have never come across any other references to Multics. What was Multics developed for ? Has it ever been finished ? Just curious. Marcel (has worked with Unix since 1981) (Remove .zzz from my email address when replying) ###### From: Lee Courtney Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Thu, 29 Oct 1998 09:32:08 -0800 Organization: Monterey Software Group Inc. Lines: 44 Message-ID: <3638A698.1AEA8CF5@slip.net> References: NNTP-Posting-Host: oak-hiper1a-169-233.dialup.slip.net Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="------------EBA7FF21B2B38EB400D8972F" X-Mailer: Mozilla 4.04 [en] (Win95; I) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newspeer.monmouth.com!newsfeed.slip.net!news.slip.net!not-for-mail --------------EBA7FF21B2B38EB400D8972F Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit THE place to go for Multics info is: http://www.multicians.org/multics.html Enjoy, Lee Courtney -- Monterey Software Group Inc. Electronic Mail: leec AT slip DOT net Suite 300 Voice: 408-735-0437 520 Lawrence Expressway Fax: 408-735-8346 Sunnyvale, California 94086-4025 Pager: 408-237-1705 www.editcorp.com/Businesses/MontereySoftware --------------EBA7FF21B2B38EB400D8972F Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit THE place to go for Multics info is: http://www.multicians.org/multics.html

Enjoy,

Lee Courtney
--
Monterey Software Group Inc.              Electronic Mail: leec AT slip DOT net
Suite 300                                 Voice: 408-735-0437
520 Lawrence Expressway                   Fax:   408-735-8346
Sunnyvale, California 94086-4025          Pager: 408-237-1705

www.editcorp.com/Businesses/MontereySoftware --------------EBA7FF21B2B38EB400D8972F-- ###### From: jones@cs.uiowa.edu (Douglas W. Jones,201H MLH,3193350740,3193382879) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 29 Oct 1998 16:47:17 GMT Organization: The University of Iowa Lines: 98 Message-ID: <71a66l$bfe$1@flood.weeg.uiowa.edu> References: NNTP-Posting-Host: pyrite.cs.uiowa.edu Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!howland.erols.net!netnews.com!newsfeed.enteract.com!cyclone.i1.net!uunet!in2.uu.net!news.uiowa.edu!not-for-mail From article , by mlipman@iname.zzz.com (Marcel Lipman): > > In articles about Unix, you always read that Unix was developed by > the people who used to work on the Multics project. I have never > come across any other references to Multics. Oy these youngsters! The Multics project was a big venture between MIT, GE, Bell Labs and the Department of Defense. The goal was to develop a timesharing computer suitable for a "computer utility" that could be shared by individuals, businesses and schools (and the army, and the navy, and the marines). In the mid 1960's, this was a radical idea! Today, we have MSonline, AOL, Compuserve and plenty of other businesses that run precisely the kind of computer utilities envisioned by Project Mac. Anyway, the Multics system had the following features that may have been firsts: 1) A hierarchical file system. 2) Access control lists on files. 3) On opening a file, it becomes a segment in your address space. 4) Hierarchic protection within each address space. 5) The operating system was written in a high level language. 6) Achieved C2 security in the Orange Book model. In addition, Multics incorporated a number of features that were not firsts, but that were rare: a) Virtual memory. b) Each user process runs in a separate address space. The high level language was PL/I, although there were many Multics utilities written in BCPL. These were good choices for the 1965-1970 time frame. Note that Simula 67, Algol 68, Pascal and C all date from roughly this era, but that good compilers for these weren't really available until the early 1970's. As the Multics project evolved, Honeywell bought GE's computer business and Bell Labs dropped out of the project. The GE 600 computer on which Multics was based was replaced by the Honeywell 6000 series. These were excellent large mainframes, and they remained in fairly widespread use through the 1970's and into the 1980's. Honeywell offered two operating systems for these machines, GCOS (the General Electric Commercial Operating System, with the Electric dropped after Honeywell bought the business, and Multics. Multics was a very stable and extremely reliable operating system. Multics was not intended to be ported across multiple platforms, but there was a minicomputer company, Prime, that hired a number of Multicians (common slang for Multics programmers), and this community of programmers converted Prime's operating system, PrimOS, into a near clone of Multics. I hated the first version of PrimOS I encountered, but the final version I used before moving to the VAX world was very nice indeed. Back to the Multics/Unix connection: Dennis Ritchie, Ken Thompson, et all, were part of the staff at the Murray Hill Computer Center at Bell Labs. As a participant in the Multics project, Bell got a GE 600, and this machine became a GECOS machine when Bell withdrew from the Multics project. For several years, Ritchie, Thompson and company did things like porting Multics software to GECOS; things like the QED text editor and the RUNOFF text formatter. It was while they were being paid to do this that Ken Thompson began to fiddle with a spare PDP-9 writing a personal operating system called Unix, as something of a joke, and Dennis Ritchie began to work on a personal version of BCPL called B. Bell Labs had a practical problem to solve: Production of technical memoranda was expensive and labor intensive, requiring technical typists and many drafts. Text editing systems offered some hope, and machines such as the CAT phototypesetter offered the possibility of getting good publication quality typesetting. As a result, Ritchie and Thompson were able to put together a proposal: Rewrite Unix in Ritchie's new programming language, port QED and RUNOFF to it, and hook it to the CAT phototypesetter, thus producing a good technical word processing system. They got the funding, bought a new PDP-11, and set to work: We got UNIX, plus C, ed, sed, nroff, and roff out of this project. I still do plenty of word processing on my desktop Unix system, and occasionally, I even find myself using ed and roff! (The QED editor, as it turned out, proved to be too big to port to the PDP-11, so they broke it into two pieces, an interactive editor, ed, and a stream editor, sed. Each supported a subset of the GECOS version of QED.) No doubt, Ritchie can elaborate on this story. I was an outsider to the UNIX world until the 1980's, but I was briefly at Bell Labs in the early 1970's and I still have my GCOS QED and RUNOFF manuals with the names of most of the original UNIX development group listed as authors. Doug Jones jones@cs.uiowa.edu ###### Newsgroups: alt.folklore.computers From: alderson@netcom.netcom.com (Richard M. Alderson III) Subject: Re: Multics In-Reply-To: jones@cs.uiowa.edu's message of 29 Oct 1998 16:47:17 GMT Message-ID: Sender: alderson@netcom.netcom.com Reply-To: alderson@netcom.com Organization: NETCOM On-line services References: <71a66l$bfe$1@flood.weeg.uiowa.edu> Date: Fri, 30 Oct 1998 00:27:37 GMT Lines: 13 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!ix.netcom.com!netcom!alderson In article <71a66l$bfe$1@flood.weeg.uiowa.edu> jones@cs.uiowa.edu (Douglas W. Jones,201H MLH,3193350740,3193382879) writes: >It was while they were being paid to do this that Ken Thompson began to fiddle >with a spare PDP-9 writing a personal operating system called Unix, as >something of a joke, and Dennis Ritchie began to work on a personal version of >BCPL called B. Minor quibble: PDP-7, not PDP-9. Both 18-bit, of course... -- Rich Alderson Last LOTS Tops-20 Systems Programmer, 1984-1991 Current maintainer, MIT TECO EMACS (v. 170) last name @ XKL dot COM Chief systems administrator, XKL LLC, 1998-now ###### From: jadams@NO.SPAM.PLEASE (James W. Adams) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 30 Oct 1998 08:29:40 -0500 Organization: UNIX Internals, Charlottesville, VA Lines: 79 Message-ID: <71cf04$5fk@jamesa.cstone.net> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> NNTP-Posting-Host: jamesa.cstone.net X-Trace: Skuzzy.cstone.net 909754455 30286 206.205.42.225 (30 Oct 1998 13:34:15 GMT) X-Complaints-To: usenet@news.cstone.net NNTP-Posting-Date: 30 Oct 1998 13:34:15 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newshub.northeast.verio.net!ptdnetP!newsgate.ptd.net!news1.cstone.net!news.COMET.NET!not-for-mail In article <71a66l$bfe$1@flood.weeg.uiowa.edu>, Douglas W. Jones,201H MLH,3193350740,3193382879 wrote: Just a few diffs to a good article. > 5) The operating system was written in a high level language. Not really portable, though. Ritchie's recast of UNIX in C was the first major OS to achieve any real degree of hardware independence. Indeed, the hardware dependence of MULTICS was undoubtedly a major factor in its demise. Arguably, the MULTICS PL/I compiler may have been the most powerful general-purpose high-level language model ever implemented to date. > 6) Achieved C2 security in the Orange Book model. Actually, B2, a much higher level. At the time about the only thing better was Honeywell SCOMP, a DPS-6 product which achieved A1, but ran on a 18 bit minicomputer with much less power. >As the Multics project evolved, Honeywell bought GE's computer business >and Bell Labs dropped out of the project. The GE 600 computer on which >Multics was based was replaced by the Honeywell 6000 series. These were >excellent large mainframes, and they remained in fairly widespread use >through the 1970's and into the 1980's. Honeywell offered two operating >systems for these machines, GCOS (the General Electric Commercial >Operating System, with the Electric dropped after Honeywell bought the >business, and Multics. Multics was a very stable and extremely reliable >operating system. G(E)COS= General (Electric) Comprehensive Operation Supervisor, IIRC. I can't recall whether it was "operation" or "operating" but it was definitely "supervisor." Honeywell offered three OSs for the 6000,L/6[68],DPS-8. These were GCOS (later GCOS-8 which implemented virtual memory and eliminated memory quadrant addressing limitations on process address spaces, 256 M words of 36 bit memory, IIRC), CP-6, a very interesting OS in its own right, and MULTICS. MULTICS required special hardware. The initial system was an emulation on GE-635 hardware which later ran in production on GE-645 which had additional memory protection features. Ken Thompson developed UNIX on a GE-635 running GCOS (by then the Honeywell "merger" had taken place) and cross-compiled it to run on a DEC PDP-7. It is instructive to the student of computer architecture to look at these systems. The Honeywell 6000 with EIS is perhaps the most elegant traditional mainframe architecture, with a very complex instruction set implemented entirely in hardware, with no microcode. Later versions used increasing levels of microcode. I recall having to "miniboot" a DPS-8 using microcode on a 8 inch floppy disk which was read into a writable control store. I also ran a DG NOVA/4X at one point which had writable control store. I regard microcoded mainframes and minicomputers, such as the VAX and Data General Eclipse MV/8000 as a logical progression from CISC to RISC in the way they used silicon real estate. >Dennis Ritchie, Ken Thompson, et all, were part of the staff at the >Murray Hill Computer Center at Bell Labs. As a participant in the Multics >project, Bell got a GE 600, and this machine became a GECOS machine when >Bell withdrew from the Multics project. For several years, Ritchie, >Thompson and company did things like porting Multics software to GECOS; >things like the QED text editor and the RUNOFF text formatter. Don't forget Joe Ossana, who did much of the troff work. Brian Kernighan later developed this into ditroff (device independent troff) which used postprocessors to drive laser printers and other types of phototypesetters (the Autologic APS-5, for instance). -- James W. Adams -- jamesa @ cstone.net <-- remove spaces "I became obsessed with angels and ballerinas, things of grace and beauty, otherworldly." Charlottesville, VA 22903 --C. Love ###### From: ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 30 Oct 1998 19:18:24 GMT Organization: The National Capital FreeNet Lines: 38 Message-ID: <71d3e0$b1h@freenet-news.carleton.ca> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> Reply-To: ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) NNTP-Posting-Host: freenet5.carleton.ca X-Given-Sender: ab528@freenet5.carleton.ca (Heinz W. Wiggeshoff) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!sunqbc.risq.qc.ca!torn!nott!cunews!freenet-news.carleton.ca!FreeNet.Carleton.CA!ab528 James W. Adams (jadams@NO.SPAM.PLEASE) writes: > In article <71a66l$bfe$1@flood.weeg.uiowa.edu>, > Douglas W. Jones,201H MLH,3193350740,3193382879 wrote: > > Just a few diffs to a good article. > >> 5) The operating system was written in a high level language. > > Not really portable, though. Ritchie's recast of UNIX in > C was the first major OS to achieve any real degree of hardware > independence. Indeed, the hardware dependence of MULTICS was > undoubtedly a major factor in its demise. What crap! PL/1 was defined as a language largely independent of hardware - how various implementations acted is another matter. As for C, tell me the definition of the datatype 'int': is it 7 bits + sign, 15 bits + sign, 31 bits + sign ? (Without reference to my C manual in another room): short int <= int <= long int none of which are defined in the language! Utter nonsense. Give me DECLARE X FIXED BINARY ( 7 ), Y FIXED BINARY ( 15 ), Z FIXED BINARY ( 31 ); any day. And how does C handle DECLARE A FIXED BINARY ( 12, 4 ); A = X + 1; ###### From: korpela@islay.ssl.berkeley.edu (Eric J. Korpela) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 30 Oct 1998 20:07:22 GMT Organization: Cal Berkeley-- Space Sciences Lab Lines: 49 Message-ID: <71d69q$mot$1@agate.berkeley.edu> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> NNTP-Posting-Host: islay.ssl.berkeley.edu Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsfeed.berkeley.edu!agate!islay.ssl.berkeley.edu!korpela In article <71d3e0$b1h@freenet-news.carleton.ca>, Heinz W. Wiggeshoff wrote: > > What crap! PL/1 was defined as a language largely independent of > hardware - how various implementations acted is another matter. > > As for C, tell me the definition of the datatype 'int': > is it 7 bits + sign, 15 bits + sign, 31 bits + sign ? Gee, I wish I could remember the last time I used a ones compliment machine. Most machines no longer have a "sign bit" in integer math. > > (Without reference to my C manual in another room): > > short int <= int <= long int > > none of which are defined in the language! Utter nonsense. Certainly they are! A both a short int and an int must be able to store values between -32767 and 32767. A long must be able to store values between -2147483647 and 2147483647. A signed char must be able to store values between -127 and 127. (Actually, it may be -128 and 127). Nothing, however, prohibits these types from storing a larger range values. > Give me > > DECLARE X FIXED BINARY ( 7 ), > Y FIXED BINARY ( 15 ), > Z FIXED BINARY ( 31 ); > > any day. The question should be why you need these specific data types. Any value you could give X would fit in a C char, any value you could give Y would fit in a C int, any value you could give Z would fit in a C long. Eric -- Eric Korpela | An object at rest can never be korpela@ssl.berkeley.edu | stopped. Click for home page. ###### From: atbowler@thinkage.on.ca (Alan Bowler) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 30 Oct 1998 20:13:28 GMT Organization: Thinkage Ltd. Lines: 14 Message-ID: <71d6l8$3da$1@nntp2.uunet.ca> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> NNTP-Posting-Host: 192.102.11.4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsfeed.direct.ca!news.uunet.ca!atbowler In article <71cf04$5fk@jamesa.cstone.net> jadams@NO.SPAM.PLEASE (James W. Adams) writes: > >Honeywell offered three OSs for the 6000,L/6[68],DPS-8. These were >GCOS (later GCOS-8 which implemented virtual memory and eliminated >memory quadrant addressing limitations on process address spaces, >256 M words of 36 bit memory, IIRC), CP-6, a very interesting OS >in its own right, and MULTICS. For some reason it seems to be an architecture that encouraged operating system designs. Besides the ones you mention, there is/was also DTSS and MarkIII. Of all the operating systems only Gcos-8 and MarkIII are still undergoing active development today. (Although there is some consideration being given to porting CP-6 to the current hardware platform (dps-9000g)). ###### From: ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 30 Oct 1998 20:49:26 GMT Organization: The National Capital FreeNet Lines: 43 Message-ID: <71d8om$d30@freenet-news.carleton.ca> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> <71d69q$mot$1@agate.berkeley.edu> Reply-To: ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) NNTP-Posting-Host: freenet3.carleton.ca X-Given-Sender: ab528@freenet3.carleton.ca (Heinz W. Wiggeshoff) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-dc.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsfeed.cwix.com!130.185.14.36!torn!nott!cunews!freenet-news.carleton.ca!FreeNet.Carleton.CA!ab528 Eric J. Korpela (korpela@islay.ssl.berkeley.edu) writes: > > Gee, I wish I could remember the last time I used a ones compliment > machine. Most machines no longer have a "sign bit" in integer > math. All of my PL/1 was done on IBM System 360 and 370 - normally _two's complement_ machine for fixed point binary operations. > > Certainly they are! > > A both a short int and an int must be able to store values between > -32767 and 32767. A long must be able to store values between > -2147483647 and 2147483647. A signed char must be able to store > values between -127 and 127. (Actually, it may be -128 and 127). > Nothing, however, prohibits these types from storing a larger range > values. Exactly. Absolutely undefined, as in whatever the compiler writer wishes to do. > > The question should be why you need these specific data types. > Any value you could give X would fit in a C char, any value you > could give Y would fit in a C int, any value you could give Z > would fit in a C long. FIXED BINARY are two 'attributes' in PL/1 speak. Fixed point arithmetic, and binary representation, (as opposed to FLOAT and DECIMAL). X, Y, and Z were given the maximum precision for a byte, halfword and word in that example. By default, the binary point is to the right. DECLARE P FIXED BINARY ( 20, 8 ) INITIAL ( 0 ); gives us 000000000000.00000000 which (in the IBM m/f world) fits into a 32-bit word. Arithmetic operations obviously require taking the binary point into consideration. Ooops - there goes the bell. ###### From: jones@cs.uiowa.edu (Douglas W. Jones,201H MLH,3193350740,3193382879) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 30 Oct 1998 22:15:20 GMT Organization: The University of Iowa Lines: 34 Message-ID: <71ddpo$gek$1@flood.weeg.uiowa.edu> References: <71d3e0$b1h@freenet-news.carleton.ca> NNTP-Posting-Host: pyrite.cs.uiowa.edu Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!news.maxwell.syr.edu!newspeer.monmouth.com!feeder1.wwnet.net!streamer1.cleveland.iagnet.net!NewsNG.Chicago.Qual.Net!news.uiowa.edu!not-for-mail From article <71d3e0$b1h@freenet-news.carleton.ca>, by ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff): > James W. Adams (jadams@NO.SPAM.PLEASE) writes: >>> 5) The operating system was written in a high level language. >> >> Not really portable, though. Ritchie's recast of UNIX in >> C was the first major OS to achieve any real degree of hardware >> independence. Indeed, the hardware dependence of MULTICS was >> undoubtedly a major factor in its demise. > What crap! PL/1 was defined as a language largely independent of > hardware - how various implementations acted is another matter. Not crap at all! The fact that Multics wasn't portable had nothing to do with any lack of portability in PL/I code. It had everything to do with the hierarchic protection model required by Multics. If the MMU on your machine didn't support this model, you couldn't run the system on your machine. The reason UNIX proved to be portable rests in part on the fact that UNIX makes minimal assumptions about the host's protection mechanisms. All it assumes is that there is a way to protect the kernel from the user, and in virtual memory implementations, it assumes you can give each user a distinct memory address space. Multics, on the other hand, requires a ring-number as a field of every page-table entry, it requires a ring-number field in the PSW, and it requires interesting logic comparing these fields on every memory reference. Lacking that, you can't port Multics. Doug Jones jones@cs.uiowa.edu ###### From: "George R. Gonzalez" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Fri, 30 Oct 1998 16:46:10 -0600 Organization: University of Minnesota, Twin Cities Campus Lines: 29 Message-ID: <71dfq0$iug$1@news1.tc.umn.edu> References: <71d3e0$b1h@freenet-news.carleton.ca> <71ddpo$gek$1@flood.weeg.uiowa.edu> NNTP-Posting-Host: grg2.micro.umn.edu Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Newsreader: Microsoft Outlook Express 4.72.3110.5 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news-feed.inet.tele.dk!bofh.vszbr.cz!mr.net!newshub.tc.umn.edu!news1.tc.umn.edu!not-for-mail Douglas W. Jones,201H MLH,3193350740,3193382879 wrote in message >Multics, on the other hand, requires a ring-number as a field of every >page-table entry, it requires a ring-number field in the PSW, and it >requires interesting logic comparing these fields on every memory >reference. Lacking that, you can't port Multics. > > Doug Jones > jones@cs.uiowa.edu IIRC it wasnt all done by hardware-- either the compiler had to generate some extra instructions, or if you were programming in assembler, there were some rather complex macros that hid the required levels of indirection. But it's been many years since I read Organick's book, so ICBR. Also a bit back some Multics guru hinted that there was a fair amount of word-size dependency built into the code. Just because the PL/I dialect made it possible to write machine-independent code doesnt mean every line of code was written with machine independence in mind. Which brings up a point-- if one were to want to port Multics to a more modern architecture, and assuming it could be done, how much networking support was there in Multics? Was there a TCP/IP stack? ###### From: ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 30 Oct 1998 23:58:27 GMT Organization: The National Capital FreeNet Lines: 29 Message-ID: <71djr3$hk8@freenet-news.carleton.ca> References: <71d3e0$b1h@freenet-news.carleton.ca> <71ddpo$gek$1@flood.weeg.uiowa.edu> Reply-To: ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) NNTP-Posting-Host: freenet2.carleton.ca X-Given-Sender: ab528@freenet2.carleton.ca (Heinz W. Wiggeshoff) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!sunqbc.risq.qc.ca!torn!nott!cunews!freenet-news.carleton.ca!FreeNet.Carleton.CA!ab528 Douglas W. Jones,201H MLH,3193350740,3193382879 (jones@cs.uiowa.edu) writes: > > Not crap at all! The fact that Multics wasn't portable had nothing > to do with any lack of portability in PL/I code. It had everything > to do with the hierarchic protection model required by Multics. If > the MMU on your machine didn't support this model, you couldn't > run the system on your machine. Not having any practical exposure to Multics, except after the era literature, my comments apply to the idea the PL/1 (my spelling B-) is portable IF THE IMPLEMENTATION FOLLOWS THE LANGUAGE DEFINITION. Does this sound FORTRAN, APL, COBOL, JAVA, C ... familiar? I've never read a definition of the PL/1 language that (for example) requires an Intel 286/386/486 protection scheme to run - although the SUBSCRIPTRANGE condition might be a cinch on those architectures. Think BOUND instruction. If a PL/1 program addresses a specific environment, for example: DECLARE TCBPTR POINTER, TCB FIXED BINARY ( 31 ) INITIAL ( 16 ); TCBPTR = UNSPEC ( TCB ); etc. etc., to step through the OS/360 MVT control blocks, don't expect logical results on other hardware/software. (BTW - this example worked in the Good Old Days.) ###### From: greg@sherrill.kiva.net (Gregory Travis) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 30 Oct 1998 20:16:10 -0500 Organization: Kiva Networking Lines: 19 Message-ID: <71docq$71l$1@sherrill.kiva.net> References: <71d3e0$b1h@freenet-news.carleton.ca> <71ddpo$gek$1@flood.weeg.uiowa.edu> NNTP-Posting-Host: sherrill.kiva.net Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news-peer.gip.net!news.gsl.net!gip.net!newsfeed.cwix.com!206.97.64.40!news.kiva.net!not-for-mail In article <71ddpo$gek$1@flood.weeg.uiowa.edu>, Douglas W. Jones,201H MLH,3193350740,3193382879 wrote: > >The reason UNIX proved to be portable rests in part on the fact that >UNIX makes minimal assumptions about the host's protection mechanisms. >All it assumes is that there is a way to protect the kernel from the >user, and in virtual memory implementations, it assumes you can give >each user a distinct memory address space. I'm not sure it requires even that. The ability to load a program at an arbitrary address (relocation) is about all that it requires - something that can be done in the compiler as it is, for example, on the Amiga. A minimal MMU adds protection against kernel corruption but that's it. greg -- gregory travis |"If you're going to kill someone there isn't much reason greg@littlebear.com|to get angry. You just pull the trigger. We need to smile |with Novell when we pull the trigger." MSFT's Jim Allchin ###### Newsgroups: alt.folklore.computers Subject: Re: Multics References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> Organization: Plethora Internet X-Newsreader: trn 4.0-test69 (20 September 1998) From: seebs@plethora.net (Peter Seebach) Lines: 23 Message-ID: Date: Sat, 31 Oct 1998 22:29:00 GMT NNTP-Posting-Host: 205.166.146.8 X-Trace: ptah.visi.com 909872940 205.166.146.8 (Sat, 31 Oct 1998 16:29:00 CDT) NNTP-Posting-Date: Sat, 31 Oct 1998 16:29:00 CDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!chippy.visi.com!news-out.visi.com!ptah.visi.com!not-for-mail In article <71d3e0$b1h@freenet-news.carleton.ca>, Heinz W. Wiggeshoff wrote: > As for C, tell me the definition of the datatype 'int': > is it 7 bits + sign, 15 bits + sign, 31 bits + sign ? No. It is a type capable of representing values from INT_MIN through INT_MAX. > short int <= int <= long int Typically. > none of which are defined in the language! They are defined enough. You know that they hold at least enough bits to be interesting. -s -- Copyright 1998, All rights reserved. Peter Seebach / seebs@plethora.net C/Unix wizard, Pro-commerce radical, Spam fighter. Boycott Spamazon! Seeking interesting programming projects. Not interested in commuting. Visit my new ISP --- More Net, Less Spam! ###### From: ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 31 Oct 1998 23:55:26 GMT Organization: The National Capital FreeNet Lines: 13 Message-ID: <71g81e$4gu@freenet-news.carleton.ca> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> Reply-To: ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) NNTP-Posting-Host: freenet6.carleton.ca X-Given-Sender: ab528@freenet6.carleton.ca (Heinz W. Wiggeshoff) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!isdnet!newsfeed.cwix.com!128.230.129.106!news.maxwell.syr.edu!news.new-york.net!nntp.frontiernet.net!node17.frontiernet.net!xcski.com!freenet-news.carleton.ca!FreeNet.Carleton.CA!ab528 Peter Seebach (seebs@plethora.net) writes: > > No. It is a type capable of representing values from INT_MIN through > INT_MAX. What part of the C language definition is INT_MIN/MAX ? > ^^^^^^^^^^^^^^^^^^^ > They are defined enough. You know that they hold at least enough bits to > be interesting. This is funnier than the Pope pronouncing on birth control. "You no playa da game, ..." ###### From: ftit@engin.umich.edu (Sergej Roytman) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 1 Nov 1998 00:52:19 GMT Organization: University of Michigan, College of Engineering Lines: 35 Message-ID: <71gbc3$nkr@srvr1.engin.umich.edu> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> NNTP-Posting-Host: az.engin.umich.edu Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!howland.erols.net!newsxfer.itd.umich.edu!srvr1.engin.umich.edu!ftit In article <71a66l$bfe$1@flood.weeg.uiowa.edu>, Douglas W. Jones,201H MLH,3193350740,3193382879 wrote: >From article , >by mlipman@iname.zzz.com (Marcel Lipman): >> In articles about Unix, you always read that Unix was developed by >> the people who used to work on the Multics project. I have never >> come across any other references to Multics. >Oy these youngsters! >The Multics project was a big venture between MIT, GE, Bell Labs and the > [some very interesting history snipped] Try http://www.best.com/~thvv/multics.html A friend of mine cut her teeth at the Oakland U. (Michigan, USA) Multics installation. Once when I was complaining about the latest offerings from Redmond (and face it, there's a lot to complain about!) and talking about how great Unix is, she basically did a, "Oy these youngsters!" at me and directed me to this site. Some fascinating stuff! It's a great pity, I think, that Multics proved too unwieldy for the technology of the time, though now it looks like modern Unices are integrating some of the same sorts of things. Despite what I hear in another thread on this newsgroup, I do feel sorry that I missed the Good Ol' Days. Some of the stuff I'm doing now is extremely interesting, but it's not about programming computers. It's about programming a very sophisticated virtual machine that sits _on top of_ computers. On the other hand, I am running into writing some C this weekend, and that's something more like it. Maybe I should invest in an M6811 or so... -- Sergej Roytman ###### Newsgroups: alt.folklore.computers Subject: Re: Multics From: riplin@kcbbs.gen.nz (Richard Plinston) Date: 1 Nov 98 03:01:04 GMT Message-ID: <3298304.10864.23275@kcbbs.gen.nz> References: <71d69q$mot$1@agate.berkeley.edu> Organization: Kappa Crucis Unix BBS, Auckland, New Zealand Lines: 18 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsfeed.cwix.com!203.97.37.7!newsfeed.clear.net.nz!status.gen.nz!kcbbs!riplin In message <<71d69q$mot$1@agate.berkeley.edu>> korpela@islay.ssl.berkeley.edu writes: > Heinz W. Wiggeshoff wrote: > > > > What crap! PL/1 was defined as a language largely independent of > > hardware - how various implementations acted is another matter. > > > > As for C, tell me the definition of the datatype 'int': > > is it 7 bits + sign, 15 bits + sign, 31 bits + sign ? > > Gee, I wish I could remember the last time I used a ones compliment > machine. Most machines no longer have a "sign bit" in integer > math. Having a bit that when set indicates that the value is negative and when unset indicates positive does not imply 'ones compliment' or 'twos compliment', or indeed anything else about how the value is represented. ###### Newsgroups: alt.folklore.computers Subject: Re: Multics From: riplin@kcbbs.gen.nz (Richard Plinston) Date: 1 Nov 98 03:10:30 GMT Message-ID: <3298304.11430.23577@kcbbs.gen.nz> References: <71d69q$mot$1@agate.berkeley.edu> Organization: Kappa Crucis Unix BBS, Auckland, New Zealand Lines: 24 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!212.63.192.161.MISMATCH!newshub.bart.net!news.tele2.nl!newsfeed1.swip.net!swipnet!newshub.northeast.verio.net!newspeer.monmouth.com!newsfeed.clear.net.nz!status.gen.nz!kcbbs!riplin In message <<71d69q$mot$1@agate.berkeley.edu>> korpela@islay.ssl.berkeley.edu writes: > > > > DECLARE X FIXED BINARY ( 7 ), > > Y FIXED BINARY ( 15 ), > > Z FIXED BINARY ( 31 ); > > > > any day. > > The question should be why you need these specific data types. > Any value you could give X would fit in a C char, any value you > could give Y would fit in a C int, any value you could give Z > would fit in a C long. How about: Because when these specific data type are in a record or struct, say on a disk file, the size is guaranteed so when the record is read back on another system the data is actually readable. When various implementations of a language give different sizes for the data types it becomes difficult to portably use data files, or even structs in memory, without converting the data to some other types where the size can be guaranteed. ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 1 Nov 1998 15:26:06 GMT Lines: 59 Message-ID: <71huie$meo$2@irk.zetnet.co.uk> References: <71d3e0$b1h@freenet-news.carleton.ca> NNTP-Posting-Host: man-164.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 909933966 23000 194.247.40.208 (1 Nov 1998 15:26:06 GMT) NNTP-Posting-Date: 1 Nov 1998 15:26:06 GMT X-Everything: Net-Tamer V 1.08X Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!news5.cableinet.net!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-10-30 ab528@FreeNet.Carleton.CA(HeinzW.Wiggeshoff) said: :> Not really portable, though. Ritchie's recast of UNIX in :> C was the first major OS to achieve any real degree of hardware :> independence. Indeed, the hardware dependence of MULTICS was :> undoubtedly a major factor in its demise. :What crap! What crap, exactly? Are you saying that Multics was portable? How can you justify that in the face of a fair few hardened Multics hackers essentially saying "it died with the GE range"? Or are you saying that C wasn't the factor which made for Unix's portability? (arguing against a point not made, though it's an arguable point) Or...? :PL/1 was defined as a language largely independent of :hardware - how various implementations acted is another matter. :As for C, tell me the definition of the datatype 'int': :is it 7 bits + sign, 15 bits + sign, 31 bits + sign ? :(Without reference to my C manual in another room): :short int <= int <= long int :none of which are defined in the language! Utter nonsense. I believe it cannot be less than 15 bits plus sign, but DQM. As for the modifications which give short and long, in most implementations what you'd probably find is that int is one or the other (at least 16 or at least 32), and if you need to be sure about it use a long, but if you just need a number (eg. a loop counter) use an int. What bit-length is the value returned by ANS Forth? Why is that solution not usable? After all, ANSI C has defined constants indicating word length. :Give me :DECLARE X FIXED BINARY ( 7 ), :Y FIXED BINARY ( 15 ), :Z FIXED BINARY ( 31 ); :any day. :And how does C handle :DECLARE A FIXED BINARY ( 12, 4 ); :A = X + 1; What you just wrote in PL/1 is syntactic sugar. Much as C++ wouldn't force aWindow += aTextLabel; to be compiled using the assembly-level instructions, the PL/1 stuff would have compiled to LDA X ADD #16 STA A (if it compiled to such instructions at all ;> ). I get the impression that since C was written to implement Unix in, they wanted to have everything in the language as simple (ie. close to the hardware) as possible. That tends to exclude syntactic sugar. If you want that, use C++ instead; you can have it to devastating excess... -- Communa (together) we remember... we'll see you falling you know soft spoken changes nothing to sing within her... ###### Date: Sun, 01 Nov 1998 12:22:02 -0800 From: thvv@multicians.org (Tom Van Vleck) Newsgroups: alt.folklore.computers Subject: Re: Multics Message-ID: References: <71d3e0$b1h@freenet-news.carleton.ca> <71ddpo$gek$1@flood.weeg.uiowa.edu> Organization: Multicians X-Newsreader: MT-NewsWatcher 2.4.1 X-url: http://www.multicians.org/ Lines: 50 NNTP-Posting-Host: thvv.vip.best.com X-Trace: nntp1.ba.best.com 909951733 29744 206.86.18.21 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsfeed.direct.ca!peerfeed.ncal.verio.net!news1.best.com!news3.best.com!nntp1.ba.best.com!not-for-mail Douglas W. Jones wrote: > Multics, on the other hand, requires a ring-number as a field of every > page-table entry, it requires a ring-number field in the PSW, and it > requires interesting logic comparing these fields on every memory > reference. Lacking that, you can't port Multics. A couple of points about the portability of Multics: 1. Remember we're dealing with a system that has lasted many years. The original goal was a system that would be hardware independent. Over time this goal was compromised. 2. The system was in fact ported to several different CPU and system architectures during its history, as GE hardware was replaced by several generations of Honeywell hardware. It was still Multics. 3. In particular, the ring-number field was added in the transition from the 645 (described in Organick's book) to the 6180 (about 1972). We considered ports to architectures such as the Honeywell NSA which lacked ring number fields; some features were harder to port than others. In the system designers' view, having a ring number field in pointers is not an _essential_ feature of Multics. (Neither, really, is paging; that's an economic strategy. Segmentation, on the other hand, is fundamental.) The essential feature would be that inner rings are protected from attempts by outer rings to trick them into operating on objects they shouldn't. Hardware assist like that in the 6180 is nice and simplifies the verification of the supervisor. 4. As one poster pointed out, a major problem with a "simple" port would be the dependency on a 36-bit word that crept into many data structures. The ability in PL/I to specify the bit size of the containers for values led programmers to declare data pointer sizes so that two 18-bit pointers could fit per word, for example. Many such references might be merely inefficient on a 64-bit machine, but every one would have to be scrutinized for possible bugs. A deeper problem in porting Multics to a modern architecture would be that storage and memory sizes have gotten a lot bigger. We had to expand disk address fields a few times as capacities grew; the 1MB max size for a segment, that seemed so enormous when we began, began to pinch the system in the 80s, and would certainly have to be overcome for modern machines. 5. Porting Multics to the Intel architecture was studied in the mid 80s, and two projects to do so were blocked by Honeywell management. See http://www.multicians.org/history.html near the end. ###### X-Old-trace: 909913285 A01OARAUVD437CCD1C usenet80.supernews.com Subject: Re: Multics Approved: clc@plethora.net References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> From: genew@vip.net (Gene Wirchenko) X-Newsreader: Forte Free Agent 1.1/32.230 Organization: http://www.supernews.com, The World's Usenet: Discussions Start Here Return-Path: Lines: 41 X-Old-complaints-to: newsabuse@supernews.com Reply-To: genew@vip.net Newsgroups: alt.folklore.computers,comp.lang.c.moderated Message-ID: Originator: clcm@plethora.net (comp.lang.c.moderated) Date: Sun, 01 Nov 1998 18:52:44 GMT NNTP-Posting-Host: 205.166.146.1 X-Trace: ptah.visi.com 909946364 205.166.146.1 (Sun, 01 Nov 1998 12:52:44 CDT) NNTP-Posting-Date: Sun, 01 Nov 1998 12:52:44 CDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!chippy.visi.com!news-out.visi.com!ptah.visi.com!not-for-mail seebs@plethora.net (Peter Seebach) wrote: >In article <71d3e0$b1h@freenet-news.carleton.ca>, >Heinz W. Wiggeshoff wrote: >> As for C, tell me the definition of the datatype 'int': >> is it 7 bits + sign, 15 bits + sign, 31 bits + sign ? > >No. It is a type capable of representing values from INT_MIN through >INT_MAX. > >> short int <= int <= long int > >Typically. Not always? Please give a conforming exception. >> none of which are defined in the language! > >They are defined enough. You know that they hold at least enough bits to >be interesting. Um, the "INT" of "INT_MIN" and "INT_MAX" stands for "integer" not "interesting". Seriously, yes, but you don't know how interesting. It would good to know exactly how interesting and without having to jump through hoops. I'd like to be able to declare int(999999) thingy; to portably declare the smallest integer type that will hold 999999. I've added comp.lang.c.moderated to the newsgroup list. Since Peter is its moderator, this could be interesting. Sincerely, Gene Wirchenko Computerese Irregular Verb Conjugation: I have preferences. You have biases. He/She has prejudices. -- comp.lang.c.moderated - clcm@plethora.net ###### From: `David O'Lantern' Subject: Re: Multics Newsgroups: alt.folklore.computers References: <71d3e0$b1h@freenet-news.carleton.ca> <71ddpo$gek$1@flood.weeg.uiowa.edu> Organization: Gehenna-By-The-Sea Lines: 26 Message-ID: <718%1.41$EY3.302@news6.ispnews.com> NNTP-Posting-Host: 207.201.34.8 X-Trace: news6.ispnews.com 909970627 207.201.34.8 (Sun, 01 Nov 1998 20:37:07 EDT) NNTP-Posting-Date: Sun, 01 Nov 1998 20:37:07 EDT Date: Mon, 02 Nov 1998 01:37:07 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!isdnet!sunqbc.risq.qc.ca!newsfeed.cwix.com!209.44.33.119!hub1.ispnews.com!news6.ispnews.com.POSTED!not-for-mail Below: a Holy War between dinosaurs about ancient history. (When was that defining Serbian battle again?) >>>> 5) The operating system was written in a high level language. >>> >>> Not really portable, though. Ritchie's recast of UNIX in >>> C was the first major OS to achieve any real degree of hardware >>> independence. Indeed, the hardware dependence of MULTICS was >>> undoubtedly a major factor in its demise. >> What crap! PL/1 was defined as a language largely independent of >> hardware - how various implementations acted is another matter. > Not crap at all! The fact that Multics wasn't portable had nothing > to do with any lack of portability in PL/I code. I'm glad y'all 'Old Guys' can still be as funny as freshman. Gives me something to look forward to. D. -- "Academia is intellectual foot-binding. Tenure is an anagram for neuter." --Roly Poly Man, on alt.angst ###### From: ftit@engin.umich.edu (Sergej Roytman) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 2 Nov 1998 03:47:00 GMT Organization: University of Michigan, College of Engineering Lines: 16 Message-ID: <71j9vk$7b6@srvr1.engin.umich.edu> References: <71d3e0$b1h@freenet-news.carleton.ca> <71ddpo$gek$1@flood.weeg.uiowa.edu> NNTP-Posting-Host: az.engin.umich.edu Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!newspump.monmouth.com!newspeer.monmouth.com!newsxfer.itd.umich.edu!srvr1.engin.umich.edu!ftit In article , Tom Van Vleck wrote: >Douglas W. Jones wrote: >> Multics, on the other hand, requires a ring-number as a field of every >A couple of points about the portability of Multics: [...] >5. Porting Multics to the Intel architecture was studied in the >mid 80s, and two projects to do so were blocked by Honeywell management. >See http://www.multicians.org/history.html near the end. Looked. Didn't get why (sorry). Please explain? -- Sergej Roytman ###### Date: Mon, 02 Nov 1998 07:11:16 -0800 From: thvv@multicians.org (Tom Van Vleck) Newsgroups: alt.folklore.computers Subject: Re: Multics Message-ID: References: <71d3e0$b1h@freenet-news.carleton.ca> <71ddpo$gek$1@flood.weeg.uiowa.edu> <71j9vk$7b6@srvr1.engin.umich.edu> Organization: Multicians X-Newsreader: MT-NewsWatcher 2.4.1 X-url: http://www.multicians.org/ Lines: 15 NNTP-Posting-Host: thvv.vip.best.com X-Trace: nntp1.ba.best.com 910019485 29749 206.86.18.21 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!howland.erols.net!newsfeed.wli.net!su-news-hub1.bbnplanet.com!news.gtei.net!news1.best.com!news3.best.com!nntp1.ba.best.com!not-for-mail Sergej Roytman wrote: > Tom Van Vleck wrote: > >5. Porting Multics to the Intel architecture was studied in the > >mid 80s, and two projects to do so were blocked by Honeywell management. > >See http://www.multicians.org/history.html near the end. > > Looked. Didn't get why (sorry). Please explain? > I can only conjecture, not being a mind reader. But they wre in business to make money, and allowing the Multics design to be used to build a system they wouldn't profit from probably didn't sound like a good idea. Some folks also thought that Honeywell managment, having prophesied that Multics wouldn't work, wished to ensure that the prophecy came true. ###### From: tstyles@my-dejanews.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Mon, 02 Nov 1998 08:42:40 GMT Organization: Deja News - The Leader in Internet Discussion Lines: 29 Message-ID: <71jr9v$qub$1@nnrp1.dejanews.com> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> <71d69q$mot$1@agate.berkeley.edu> <71d8om$d30@freenet-news.carleton.ca> NNTP-Posting-Host: 193.132.79.30 X-Article-Creation-Date: Mon Nov 02 08:42:40 1998 GMT X-Http-User-Agent: Mozilla/4.07 [en] (WinNT; I ;Nav) X-Http-Proxy: 1.0 x9.dejanews.com:80 (Squid/1.1.22) for client 193.132.79.30 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!diablo.theplanet.net!london-news-feed1.bbnplanet.com!news.bbnplanet.com!cam-news-hub1.bbnplanet.com!news.gtei.net!newsfeed.xcom.net!news.maxwell.syr.edu!nntp2.dejanews.com!nnrp1.dejanews.com!not-for-mail In article <71d8om$d30@freenet-news.carleton.ca>, ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) wrote: > > Eric J. Korpela (korpela@islay.ssl.berkeley.edu) writes: [snip] > > > > A both a short int and an int must be able to store values between > > -32767 and 32767. A long must be able to store values between > > -2147483647 and 2147483647. A signed char must be able to store > > values between -127 and 127. (Actually, it may be -128 and 127). > > Nothing, however, prohibits these types from storing a larger range > > values. > > Exactly. Absolutely undefined, as in whatever the compiler writer > wishes to do. Your dictionary of undefined seems very strange, a compiler write *must* implement at least the minimum ranges for each type, that's hardly undefined. [snip] -- Tristan Styles Failure is not an option It is Standard Operating Procedure -----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Mon, 02 Nov 98 12:39:29 GMT Organization: UltraNet Communications, Inc. Lines: 16 Message-ID: <71ka61$dp9$1@ligarius.ultra.net> References: <71d3e0$b1h@freenet-news.carleton.ca> <71ddpo$gek$1@flood.weeg.uiowa.edu> NNTP-Posting-Host: d3.dial-11.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 2 Nov 1998 12:56:33 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!newsfeed.wirehub.nl!howland.erols.net!newspeer.monmouth.com!news.ultranet.com!d3 In article , thvv@multicians.org (Tom Van Vleck) wrote: >5. Porting Multics to the Intel architecture was studied in the >mid 80s, and two projects to do so were blocked by Honeywell management. >See http://www.multicians.org/history.html near the end. Wow. My assumption has been that this disease was unique to Digital. Could it have been something they [management] ate? /BAH Subtract a hundred and four for e-mail. ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Mon, 02 Nov 1998 13:56:02 GMT Organization: . Lines: 39 Message-ID: <3642b1bd.2067272@news.uunet.be> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> <71d69q$mot$1@agate.berkeley.edu> <71d8om$d30@freenet-news.carleton.ca> <71jr9v$qub$1@nnrp1.dejanews.com> NNTP-Posting-Host: pool02b-194-7-146-21.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!diablo.theplanet.net!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach tstyles@my-dejanews.com on Mon, 02 Nov 1998 08:42:40 GMT to alt.folklore.computers: > In article <71d8om$d30@freenet-news.carleton.ca>, > ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) wrote: > > Eric J. Korpela (korpela@islay.ssl.berkeley.edu) writes: > > Exactly. Absolutely undefined, as in whatever the compiler writer > > wishes to do. > > Your dictionary of undefined seems very strange, a compiler write *must* > implement at least the minimum ranges for each type, that's hardly undefined. From their own point of view, both statements have something in them. One view is that it's defined, because a minimum is defined. This implies that a C program will always handle the minimum set of data correctly, and the compiler can pick the way that's most efficient on the given hardware to do it. The other view is that it's undefined because it's open-ended. As long as things stay in memory and members of structures etc. are accessed the way members are supposed to be accessed, it will run fine, but when you start writing the raw binary data to disk, portability (to other systems, and even to other compilers on the same hardware) is far away. I suppose that's why a lot of unix programs keep their data in ASCII files. On most current systems I think it wouldn't be a problem to switch to strictly defined sizes, because most of them use a multiple of 8 bits as word size. OTOH we have to remember that C was intended to be useable on all imaginable platforms: what would you do with 9-bit, 12-bit etc. machines if there was a fixed 8-bit based size requirement? Simply not using the remaining bits would be a waste of resources (and processing power), and C would never have become the universal language you find on all possible machines, like it has now. ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 2 Nov 1998 17:00:27 GMT Lines: 17 Message-ID: <71kofb$ouu$5@irk.zetnet.co.uk> References: <71djr3$hk8@freenet-news.carleton.ca> NNTP-Posting-Host: man-029.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 910026027 25566 194.247.41.36 (2 Nov 1998 17:00:27 GMT) NNTP-Posting-Date: 2 Nov 1998 17:00:27 GMT X-Everything: Net-Tamer V 1.08X Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!diablo.theplanet.net!btnet-peer!btnet-feed1!btnet!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-10-30 ab528@FreeNet.Carleton.CA(HeinzW.Wiggeshoff) said: :I've never read a definition of the PL/1 language that (for example) :requires an Intel 286/386/486 protection scheme to run - although :the SUBSCRIPTRANGE condition might be a cinch on those :architectures. Think BOUND instruction. From what I've read, the BOUND instruction is icky and never used. One of the main problems with it appears to be that it returns to the same instruction, rather than the following one - a bit silly; why on earth would you want to repeat the BOUND test after it's failed once? Also, I seem to recall it isn't consistent across x86, but that could be an soft read error on my part. ;> -- Communa (together) we remember... we'll see you falling you know soft spoken changes nothing to sing within her... ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 2 Nov 1998 17:00:32 GMT Lines: 24 Message-ID: <71kofg$ouu$6@irk.zetnet.co.uk> References: <71g81e$4gu@freenet-news.carleton.ca> NNTP-Posting-Host: man-029.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 910026032 25566 194.247.41.36 (2 Nov 1998 17:00:32 GMT) NNTP-Posting-Date: 2 Nov 1998 17:00:32 GMT X-Everything: Net-Tamer V 1.08X Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!btnet-peer!btnet-feed1!btnet!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-10-31 ab528@FreeNet.Carleton.CA(HeinzW.Wiggeshoff) said: :Newsgroups: alt.folklore.computers :Peter Seebach (seebs@plethora.net) writes: :> No. It is a type capable of representing values from INT_MIN :>through INT_MAX. :What part of the C language definition is INT_MIN/MAX ? Well, printf() isn't in the C language definition either, for that matter, but it doesn't mean it's any less part of the C standard, or any less likely to be present. In fact, the preprocessor *is* part of the C language, as far as I understand these things, and so it makes sense for various predefined macros to also be part of that language standard. long a = 0, b = -1, MAX_INT, MIN_INT; while (b < a) b = a++; MAX_INT = b; MIN_INT = a; if you're terribly concerned about it. -- Communa (bound to have screwed up. anyone tell me where?) you know soft spoken changes nothing to sing within her... ###### Newsgroups: alt.folklore.computers Subject: Re: Multics From: riplin@kcbbs.gen.nz (Richard Plinston) Date: 2 Nov 98 18:21:00 GMT Message-ID: <3298305.66060.3537@kcbbs.gen.nz> References: <71huie$meo$2@irk.zetnet.co.uk> Organization: Kappa Crucis Unix BBS, Auckland, New Zealand Lines: 22 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!212.63.192.161.MISMATCH!newshub.bart.net!news.tele2.nl!newsfeed1.swip.net!swipnet!news.maxwell.syr.edu!newsfeed.cwix.com!203.97.37.7!newsfeed.clear.net.nz!status.gen.nz!kcbbs!riplin In message <<71huie$meo$2@irk.zetnet.co.uk>> lisard@zetnet.co.uk writes: > > On 1998-10-30 ab528@FreeNet.Carleton.CA(HeinzW.Wiggeshoff) said: > :> Not really portable, though. Ritchie's recast of UNIX in > :> C was the first major OS to achieve any real degree of hardware > :> independence. Indeed, the hardware dependence of MULTICS was > :> undoubtedly a major factor in its demise. > > :What crap! > > What crap, exactly? Are you saying that Multics was portable? How can I didn't think that he was saying anything about the portability of Multics at all. AFAIK Multics died because it was horribly inefficient. It used most of the processor power for its own needs (please draw parallels with a modern OS). > you justify that in the face of a fair few hardened Multics hackers > essentially saying "it died with the GE range"? Or are you saying that C ###### Date: Mon, 02 Nov 1998 18:26:44 -0800 From: thvv@multicians.org (Tom Van Vleck) Newsgroups: alt.folklore.computers Subject: Re: Multics Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> Organization: Multicians X-Newsreader: MT-NewsWatcher 2.4.1 X-url: http://www.multicians.org/ Lines: 35 NNTP-Posting-Host: thvv.vip.best.com X-Trace: nntp1.ba.best.com 910060013 29758 206.86.18.21 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!howland.erols.net!feed1.news.rcn.net!rcn!news2.best.com!news3.best.com!nntp1.ba.best.com!not-for-mail Richard Plinston wrote: > AFAIK Multics died because it was horribly inefficient. It > used most of the processor power for its own needs (please > draw parallels with a modern OS). Perhaps you were present at a different set of benchmarks from the ones I participated in. The ones I remember in the late 70s placed us at the front of the pack, for large interactive mainframe systems. In those days, the Honeywell sales strategy was to bid GCOS, screw around trying to run the benchmark until the last two weeks, then panic and call in the Multicians. We were able to run quite a few benchmarks the competition (Including GCOS) couldn't, and won quite a few bids. Can you tell me what facts you bas this characterization of "horribly inefficient" on, and what time frame you're referring to? The Multics machines #were# horribly expensive. Multics ran well with a lot of main storage, like most operating systems. And the Honewyell pricing model made bunches of money on expensive memory. There were no alternate suppliers, you had to buy from Honewyell. Some customers bought underpowered systems, and were in a bind: the system that performed well on the benchmark began to run quite slowly as real-world load stressed it. This showed up as inefficiency, but you could argue whether it was the system's, or the users', trying to put too much water in a small bucket. For these customers, the way out would be to buy more memory.. expensive Honeywell memory. In those days Honeywell made 85% of its profit on upgrades to existing customer machines, for all the large systems. Presumably the numbers for Multics were higher, because the system encouraged workload growth. As far as parallels with modern OSes, I use virtual memory on PCs, Macs, UNIX boxes, and Photoshop every day. It's always slow. The solution nowadays is to add more (cheap) memory. ###### From: korpela@albert.ssl.berkeley.edu (Eric J. Korpela) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 2 Nov 1998 22:57:28 GMT Organization: Cal Berkeley-- Space Sciences Lab Lines: 49 Message-ID: <71ldco$kk5$1@agate.berkeley.edu> References: <71d69q$mot$1@agate.berkeley.edu> <3298304.11430.23577@kcbbs.gen.nz> NNTP-Posting-Host: albert.ssl.berkeley.edu Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsfeed.berkeley.edu!agate!albert.ssl.berkeley.edu!korpela In article <3298304.11430.23577@kcbbs.gen.nz>, Richard Plinston wrote: >In message <<71d69q$mot$1@agate.berkeley.edu>> korpela@islay.ssl.berkeley.edu writes: >> > >> > DECLARE X FIXED BINARY ( 7 ), >> > Y FIXED BINARY ( 15 ), >> > Z FIXED BINARY ( 31 ); >> > >> > any day. >> >> The question should be why you need these specific data types. >> Any value you could give X would fit in a C char, any value you >> could give Y would fit in a C int, any value you could give Z >> would fit in a C long. > >How about: Because when these specific data type are in a >record or struct, say on a disk file, the size is guaranteed >so when the record is read back on another system the data >is actually readable. Are you saying that PL/1 defines the endian-ness of binary data written to disk? the integer format? the floating point format? I don't care what language you're using or what OS, if you want to reliably read/write structures from a binary medium usable on multiple systems with incompatible binary representations, you WILL need reading and writing routines specific to machine you are using. Unless you choose to write in a fairly machine independant form such as ACSII. >When various implementations of a language give different >sizes for the data types it becomes difficult to portably >use data files, or even structs in memory, without >converting the data to some other types where the size >can be guaranteed. It's difficult even where the sizes are the same. Try reading binary data written on a Sparcstation running Linux on an i386 running Linux. It's not the sizes that are the problem. I don't know if these processors have a PL/1 implementation, but if they do, I'd be extremely suprised if the binary data files they create are interchangeable. Eric -- Eric Korpela | An object at rest can never be korpela@ssl.berkeley.edu | stopped. Click for home page. ###### Newsgroups: alt.folklore.computers Subject: Re: Multics References: <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> Organization: Wizvax Communications, LLC From: multics@wizvax.wizvax.net (Richard Shetron) NNTP-Posting-Host: wizvax.wizvax.net Message-ID: <363e818a.0@news.wizvax.net> Date: 2 Nov 1998 23:07:38 -0500 X-Trace: 2 Nov 1998 23:07:38 -0500, wizvax.wizvax.net Lines: 29 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!newspump.monmouth.com!newspeer.monmouth.com!newsfeed.wizvax.net!news.wizvax.net!wizvax.wizvax.net!multics In article <3298305.66060.3537@kcbbs.gen.nz>, Richard Plinston wrote: >In message <<71huie$meo$2@irk.zetnet.co.uk>> lisard@zetnet.co.uk writes: >> >> On 1998-10-30 ab528@FreeNet.Carleton.CA(HeinzW.Wiggeshoff) said: >> :> Not really portable, though. Ritchie's recast of UNIX in >> :> C was the first major OS to achieve any real degree of hardware >> :> independence. Indeed, the hardware dependence of MULTICS was >> :> undoubtedly a major factor in its demise. >> >> :What crap! >> >> What crap, exactly? Are you saying that Multics was portable? How can > >I didn't think that he was saying anything about the portability >of Multics at all. > >AFAIK Multics died because it was horribly inefficient. It >used most of the processor power for its own needs (please >draw parallels with a modern OS). I'm not sure of efficiency, but the RADC H-6180 Multics system with 3MB of core (yes core) supported about 30 users before you noticed you weren't the only one on the machine and there was an 8-12 graphic terminal room used by some of the R&D groups. My Linux AMD K6-233 with 128MB of RAM has about the same performance feel doing similier things though I think the Multics PL/1 compiler was faster (comparing programs that do the same thing, not lines of code). ###### Subject: Re: Multics Approved: clc@plethora.net References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> From: Auke.Reitsma@net.hcc.nl_NO_SPAM_ (Auke Reitsma) X-Newsreader: Forte Free Agent 1.1/32.230 Organization: UUNET-NL Return-Path: Lines: 16 Reply-To: Auke.Reitsma@NO_SPAM_net.hcc.nl Newsgroups: alt.folklore.computers,comp.lang.c.moderated Message-ID: Originator: clcm@plethora.net (comp.lang.c.moderated) Date: Mon, 02 Nov 1998 23:09:24 GMT NNTP-Posting-Host: 205.166.146.1 X-Trace: ptah.visi.com 910048164 205.166.146.1 (Mon, 02 Nov 1998 17:09:24 CDT) NNTP-Posting-Date: Mon, 02 Nov 1998 17:09:24 CDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news.tele2.nl!newsfeed1.swip.net!swipnet!chippy.visi.com!news-out.visi.com!ptah.visi.com!not-for-mail On Sun, 01 Nov 1998 18:52:44 GMT, genew@vip.net (Gene Wirchenko) wrote: ... > >They are defined enough. You know that they hold at least enough bits to > >be interesting. > > Um, the "INT" of "INT_MIN" and "INT_MAX" stands for "integer" not > "interesting". ROTFL ... That's a good one for the IAQ ... Greetings from _____ /_|__| Auke Reitsma, Delft, The Netherlands. / | \ ------------------------------------- Remove NO_SPAM_ from my address ... -- comp.lang.c.moderated - clcm@plethora.net ###### From: atbowler@thinkage.on.ca (Alan Bowler) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 3 Nov 1998 17:47:54 GMT Organization: Thinkage Ltd. Lines: 10 Message-ID: <71nfka$meo$1@goblin.uunet.ca> References: <71d8om$d30@freenet-news.carleton.ca> <71jr9v$qub$1@nnrp1.dejanews.com> <3642b1bd.2067272@news.uunet.be> NNTP-Posting-Host: 192.102.11.4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news-feed.inet.tele.dk!bofh.vszbr.cz!newsfeed.direct.ca!news.vphos.net!news1.van.metronet.ca!news1.cal.metronet.ca!news.uunet.ca!atbowler In article <3642b1bd.2067272@news.uunet.be> lucvdv@null.net (Luc Van der Veken) writes: >OTOH we have to remember that C >was intended to be useable on all imaginable platforms: what >would you do with 9-bit, 12-bit etc. machines if there was a >fixed 8-bit based size requirement? A little to sweeping. C certainly targets a wide array of hardware architectures, but certainly not "all imaginable platforms". Many of us just have to remember using (not imagine) architectures for which C is unsuitable. (e.g. IBM 1620) ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Tue, 03 Nov 1998 19:04:56 GMT Organization: . Lines: 20 Message-ID: <36464aec.6858331@news.uunet.be> References: <71g81e$4gu@freenet-news.carleton.ca> <71kofg$ouu$6@irk.zetnet.co.uk> NNTP-Posting-Host: pool02b-194-7-145-94.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!diablo.theplanet.net!diablo1!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Communa said, only yesterday, in alt.folklore.computers: > long a = 0, b = -1, MAX_INT, MIN_INT; > while (b < a) b = a++; > MAX_INT = b; MIN_INT = a; > > if you're terribly concerned about it. > -- > Communa (bound to have screwed up. anyone tell me where?) [think you did it on purpose. again.] MAX_LONG, MIN_LONG? And BTW, in VC6 it's INT_MIN and INT_MAX - but that one comes from MS, they could have just reinvented them (I seem to remember older compilers where it was MIN_INT and MAX_INT). Does rise the question though: how standardized is it? ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 3 Nov 1998 20:05:50 GMT Lines: 17 Message-ID: <71nnmu$t2f$2@irk.zetnet.co.uk> References: <71ka61$dp9$1@ligarius.ultra.net> NNTP-Posting-Host: man-117.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 910123550 29775 194.247.40.150 (3 Nov 1998 20:05:50 GMT) NNTP-Posting-Date: 3 Nov 1998 20:05:50 GMT X-Everything: Net-Tamer V 1.08X Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!diablo.theplanet.net!diablo1!btnet-peer!btnet!demeter.clara.net!news.clara.net!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-11-02 jmfbahciv@aol.com said: :thvv@multicians.org (Tom Van Vleck) wrote: : :>5. Porting Multics to the Intel architecture was studied in the :>mid 80s, and two projects to do so were blocked by Honeywell :>management. See http://www.multicians.org/history.html near the :end. :Wow. My assumption has been that this disease was unique to :Digital. Could it have been something they [management] ate? You mean they *eat* bullshit, too...? -- Communa (together) we remember... we'll see you falling you know soft spoken changes nothing to sing within her... ###### Date: Tue, 03 Nov 1998 22:19:45 -0800 From: thvv@multicians.org (Tom Van Vleck) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> Organization: Multicians X-Newsreader: MT-NewsWatcher 2.4.1 X-url: http://www.multicians.org/ Lines: 39 NNTP-Posting-Host: thvv.vip.best.com X-Trace: nntp1.ba.best.com 910160394 29760 206.86.18.21 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!212.63.192.161.MISMATCH!newshub.bart.net!news.tele2.nl!newsfeed1.swip.net!swipnet!feed1.news.rcn.net!rcn!news2.best.com!news3.best.com!nntp1.ba.best.com!not-for-mail Richard Plinston wrote: > AFAIK Multics died because it was horribly inefficient. Multics died because Honeywell management didn't see a way to make money from it in the mid 80s. Maybe they wre right. Multics is a mainframe timesharing operating system. In some sense the mainframe became obsolete when the minicomputer came out in the early 70s, and we were too stubborn to notice. When PCs came out in the late 70s and early 80s, the minicomputer became obsolete as well. Each new smaller generation of machine began with the idea that a simple operating system would be more efficient and would be adequate for the new environemnt. In each case, the complexity that was pruned off grew back. (Some would say, in deformed and debased versions.) By the mid 80s, Multics was suffering because it was running on CPUs at least a hardware generation behind the (mainframe) state of the art. (Because the product was low volume compared to GCOS, Multics CPUs took longer to build and required more hand testing and debugging. Factory capacity for Multics CPUs was limited by the number of "test cells," where newly built CPUs were tested and debugged. I remember losing the battle for another cell in the late 70s; each cell tied up a lot of capital on the factory floor in Deer Valley.) The Multics subcommittee of HLSUA (Honewyell Large System Users' Association) pressured management for a new machine, and several projects were begun, and then canceled. Without a new hardware base, sales forecasts dropped to near zero. Management stopped development of the software. I don't see this as a sinister process, just arithmetic. The economies of scale that mainframe systems optimize were no longer important. Mainframe operating systems tried to use the expensive CPU as efficiently as possible. Now we have cheap CPUs a hundred times as fast sitting idle, running screen savers, playing Life. Sure, we could have rewritten Multics to run on a large Sequent multi-386. Sure, we could rewrite it to run on a PowerPC. But it would take a lot of redesign, and that costs money and time. Honeywell management didn't see it paying off. (Look where Honeywell is now.) ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 4 Nov 1998 00:04:51 GMT Lines: 34 Message-ID: <71o5n3$tpr$1@irk.zetnet.co.uk> References: <36464aec.6858331@news.uunet.be> NNTP-Posting-Host: man-037.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 910137891 30523 194.247.41.46 (4 Nov 1998 00:04:51 GMT) NNTP-Posting-Date: 4 Nov 1998 00:04:51 GMT X-Everything: Net-Tamer V 1.08X Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!newsfeed.wirehub.nl!ayres.ftech.net!news.ftech.net!demeter.clara.net!news.clara.net!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-11-03 lucvdv@null.net(LucVanderVeken) said: :Communa said, only yesterday, in alt.folklore.computers: :> long a = 0, b = -1, MAX_INT, MIN_INT; :> while (b < a) b = a++; :> MAX_INT = b; MIN_INT = a; :> if you're terribly concerned about it. :> -- :> Communa (bound to have screwed up. anyone tell me where?) :[think you did it on purpose. again.] :MAX_LONG, MIN_LONG? *blush* Actually, I think it would be LONG_MIN and LONG_MAX. But you're right... although if you aren't criticising the general gist of what I'm saying I'm not so unhappy. (And no, I never do it on purpose. I'm just careless. C'est moi.) :And BTW, in VC6 it's INT_MIN and INT_MAX - but that one comes :from MS, they could have just reinvented them (I seem to remember :older compilers where it was MIN_INT and MAX_INT). I believe that it was INT_MIN and INT_MAX earlier in the thread as well. Whoops. :Does rise the question though: how standardized is it? Well, it'll probably work on anything with a 2's complement architecture where the compiled C code ignores overflow (as I believe compiled C code is supposed to). -- Communa (together) we remember... we'll see you falling you know soft spoken changes nothing to sing within her... ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 4 Nov 1998 00:04:53 GMT Lines: 33 Message-ID: <71o5n5$tpr$2@irk.zetnet.co.uk> References: NNTP-Posting-Host: man-037.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 910137893 30523 194.247.41.46 (4 Nov 1998 00:04:53 GMT) NNTP-Posting-Date: 4 Nov 1998 00:04:53 GMT X-Everything: Net-Tamer V 1.08X Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!newsfeed.wirehub.nl!ayres.ftech.net!news.ftech.net!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-11-02 thvv@multicians.org(TomVanVleck) said: :Richard Plinston wrote: :> AFAIK Multics died because it was horribly inefficient. It :> used most of the processor power for its own needs (please :> draw parallels with a modern OS). :Perhaps you were present at a different set of benchmarks from :the ones I participated in. The ones I remember in the late 70s :placed us at the front of the pack, for large interactive mainframe :systems. In those days, the Honeywell sales strategy was to bid :GCOS, screw around trying to run the benchmark until the last :two weeks, then panic and call in the Multicians. We were able to :run quite a few benchmarks the competition (Including GCOS) :couldn't, and won quite a few bids. So it sounds like what you're all saying about Multics is: (a) in today's terms, it's a paragon of efficiency; [a bit of a red herring, since it wasn't targeted at today's systems] (b) it ran just fine on high-end systems, making it an excellent high-end timesharing OS - which was the intention of the designers; (c) it grumbled a bit in resource-starved situations. Can we agree, then, that it was "inefficient on small systems" without implying inefficiency in general? -- Communa (together) we remember... we'll see you falling you know soft spoken changes nothing to sing within her... ####### Newsgroups: alt.folklore.computers Subject: Re: Multics References: <71g81e$4gu@freenet-news.carleton.ca> <71kofg$ouu$6@irk.zetnet.co.uk> Organization: Plethora Internet X-Newsreader: trn 4.0-test69 (20 September 1998) From: seebs@plethora.net (Peter Seebach) Lines: 23 Message-ID: Date: Wed, 04 Nov 1998 01:05:34 GMT NNTP-Posting-Host: 205.166.146.8 X-Trace: ptah.visi.com 910141534 205.166.146.8 (Tue, 03 Nov 1998 19:05:34 CDT) NNTP-Posting-Date: Tue, 03 Nov 1998 19:05:34 CDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!194.165.93.117.MISMATCH!newshub.bart.net!newsfeed.wirehub.nl!howland.erols.net!feed1.news.rcn.net!rcn!chippy.visi.com!news-out.visi.com!ptah.visi.com!not-for-mail In article <71kofg$ouu$6@irk.zetnet.co.uk>, wrote: >Well, printf() isn't in the C language definition either, Sure, it is! The whole standard, including the "library" section, is the language definition. >language, as far as I understand these things, and so it makes sense for >various predefined macros to also be part of that language standard. Yes. . > long a = 0, b = -1, MAX_INT, MIN_INT; > while (b < a) b = a++; > MAX_INT = b; MIN_INT = a; This isn't guaranteed to work. It invokes undefined behavior. -s -- Copyright 1998, All rights reserved. Peter Seebach / seebs@plethora.net C/Unix wizard, Pro-commerce radical, Spam fighter. Boycott Spamazon! Seeking interesting programming projects. Not interested in commuting. Visit my new ISP --- More Net, Less Spam! ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Wed, 04 Nov 98 12:33:59 GMT Organization: UltraNet Communications, Inc. Lines: 23 Message-ID: <71pika$qh6$2@strato.ultra.net> References: <71ka61$dp9$1@ligarius.ultra.net> <71nnmu$t2f$2@irk.zetnet.co.uk> NNTP-Posting-Host: d16.dial-17.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 4 Nov 1998 12:51:22 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news-feed.inet.tele.dk!bofh.vszbr.cz!howland.erols.net!feed1.news.rcn.net!rcn!news.ultranet.com!d16 In article <71nnmu$t2f$2@irk.zetnet.co.uk>, lisard@zetnet.co.uk wrote: > > >On 1998-11-02 jmfbahciv@aol.com said: > :thvv@multicians.org (Tom Van Vleck) wrote: > : > :>5. Porting Multics to the Intel architecture was studied in the > :>mid 80s, and two projects to do so were blocked by Honeywell > :>management. See http://www.multicians.org/history.html near the > :end. > > :Wow. My assumption has been that this disease was unique to > :Digital. Could it have been something they [management] ate? > >You mean they *eat* bullshit, too...? ROTFL. That's a good one. They must have eaten it. How else could they have deposited it on our heads? They sure didn't know how to fork it over. /BAH Subtract a hundred and four for e-mail. ###### From: Dewayne Basnett Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Wed, 04 Nov 1998 12:45:23 -0500 Organization: Network Visions, Inc. Lines: 82 Message-ID: <364092B2.8277B176@nvi.to> References: <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> Reply-To: spamnojunkno_dewayne@nvi.to Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="------------9C47921897AC0522541F9C36" X-Trace: IkEMiUmKw02XdUIU57yfW+XaydCHqdNpTigahenn060= X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 4 Nov 1998 17:45:26 GMT X-Mailer: Mozilla 4.05 [en] (Win95; I) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!howland.erols.net!master.news.rcn.net!not-for-mail This is a multi-part message in MIME format. --------------9C47921897AC0522541F9C36 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I think you could have stopped at "Honeywell management didn't see". Look at Honeywells computers now. Tom Van Vleck wrote: > Richard Plinston wrote: > > AFAIK Multics died because it was horribly inefficient. > > Multics died because Honeywell management didn't see a way > to make money from it in the mid 80s. Maybe they wre right. > > Multics is a mainframe timesharing operating system. > In some sense the mainframe became obsolete when the minicomputer > came out in the early 70s, and we were too stubborn to notice. > When PCs came out in the late 70s and early 80s, the minicomputer > became obsolete as well. Each new smaller generation of machine > began with the idea that a simple operating system would be > more efficient and would be adequate for the new environemnt. > In each case, the complexity that was pruned off grew back. > (Some would say, in deformed and debased versions.) > > By the mid 80s, Multics was suffering because > it was running on CPUs at least a hardware generation behind the > (mainframe) state of the art. (Because the product was low volume > compared to GCOS, Multics CPUs took longer to build and required > more hand testing and debugging. Factory capacity for Multics CPUs > was limited by the number of "test cells," where newly built CPUs > were tested and debugged. I remember losing the battle for another > cell in the late 70s; each cell tied up a lot of capital on the > factory floor in Deer Valley.) The Multics subcommittee of HLSUA > (Honewyell Large System Users' Association) pressured management > for a new machine, and several projects were begun, and then > canceled. Without a new hardware base, sales forecasts dropped > to near zero. Management stopped development of the software. > I don't see this as a sinister process, just arithmetic. > > The economies of scale that mainframe systems optimize were no longer > important. Mainframe operating systems tried to use the expensive CPU > as efficiently as possible. Now we have cheap CPUs a hundred times as > fast sitting idle, running screen savers, playing Life. Sure, we > could have rewritten Multics to run on a large Sequent multi-386. > Sure, we could rewrite it to run on a PowerPC. But it would take > a lot of redesign, and that costs money and time. Honeywell > management didn't see it paying off. (Look where Honeywell is now.) -- Dewayne Basnett - dewayne@nvi.to - www.nvi.to phone: 610.415.0170 - pager: 888.233.7395 pin 1911091 "Golf has probably kept more people sane than psychiatrists have." Harvey Penick --------------9C47921897AC0522541F9C36 Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf" Content-Transfer-Encoding: 7bit Content-Description: Card for Dewayne Basnett Content-Disposition: attachment; filename="vcard.vcf" begin: vcard fn: Dewayne Basnett n: Basnett;Dewayne org: Network Visions, Inc. adr: 117 Main Street;;;Phoenixville;PA;19460;USA email;internet: dewayne@nvi.to tel;work: 610.415.0170 tel;fax: 610.415.0172 x-mozilla-cpt: ;0 x-mozilla-html: FALSE version: 2.1 end: vcard --------------9C47921897AC0522541F9C36-- ###### From: Barry Margolin Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics References: <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> Organization: GTE Internetworking, Cambridge, MA X-Newsreader: trn 4.0-test60 (5 October 1997) Originator: barmar@bbnplanet.com (Barry Margolin) Lines: 30 Message-ID: Date: Wed, 04 Nov 1998 18:35:56 GMT NNTP-Posting-Host: 4.2.32.97 X-Trace: burlma1-snr1.gtei.net 910204556 4.2.32.97 (Wed, 04 Nov 1998 18:35:56 GMT) NNTP-Posting-Date: Wed, 04 Nov 1998 18:35:56 GMT Distribution: world Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!cpk-news-hub1.bbnplanet.com!news.news.gtei.net!burlma1-snr1.gtei.net!not-for-mail In article , Tom Van Vleck wrote: >Richard Plinston wrote: >> AFAIK Multics died because it was horribly inefficient. > >Multics died because Honeywell management didn't see a way >to make money from it in the mid 80s. Maybe they wre right. > >Multics is a mainframe timesharing operating system. Blaming Multics's death on the fact that it was a mainframe system doesn't seem right. GCOS was a mainframe system, and Honeywell was still doing OK selling it, and developing new processors for it, when they canned Multics. At the time Multics was cancelled (1986), PC's weren't yet widespread; I think the top-of-the-line PC was the PC/AT or PC/XT, the OS of choice was MS-DOS 3.x, and there weren't yet lots of applications for it (they weren't used for much more than spreadsheets, primitive word processing, and terminal emulation). Multics was struggling within Honeywell long before PCs came along; customers have told us about occasions where they had to convince their sales rep to sell them a Multics system instead of GCOS. The problem was managerial, not technical. -- Barry Margolin, barmar@bbnplanet.com GTE Internetworking, Powered by BBN, Burlington, MA *** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups. Don't bother cc'ing followups to me. ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Wed, 04 Nov 1998 20:17:19 GMT Organization: . Lines: 23 Message-ID: <3641a632.2735513@news.uunet.be> References: <36464aec.6858331@news.uunet.be> <71o5n3$tpr$1@irk.zetnet.co.uk> NNTP-Posting-Host: pool02b-194-7-144-27.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!diablo.theplanet.net!diablo2!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach lisard@zetnet.co.uk on 4 Nov 1998 00:04:51 GMT to alt.folklore.computers: > :Does rise the question though: how standardized is it? > > Well, it'll probably work on anything with a 2's complement architecture > where the compiled C code ignores overflow (as I believe compiled C code > is supposed to). I meant it a little differently - how standardized are MIN_INT and MAX_INT: do all compilers know them? Are they ANSI standard? (Especially because of the MIN_INT / INT_MIN matter: I'm still not sure, but it feels like I've met both of them - as well as MININT and MAXINT without the underscores.) As far as I could see, your algoritm should work fine for 2's complement as well as 1's complement (if that still exists somewhere) hardware, and even for unsigned integral types as well. On 1-s complement hardware it might flunk if you let it start below +0 though (or is -0 < +0?) ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Wed, 04 Nov 1998 20:17:19 GMT Organization: . Lines: 14 Message-ID: <3646b63b.6840015@news.uunet.be> References: <71o5n5$tpr$2@irk.zetnet.co.uk> NNTP-Posting-Host: pool02b-194-7-144-27.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!newsfeed.icl.net!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach lisard@zetnet.co.uk on 4 Nov 1998 00:04:53 GMT to alt.folklore.computers: > Can we agree, then, that it was "inefficient on small systems" without > implying inefficiency in general? [I should be staying out of here - never knew multics] That would fit in nicely with what I read about Unix (initially) being something like "multics for small systems". (if I interpreted it right) Luc "seem to be running into you a lot this week" Van der Veken. ###### Date: Wed, 04 Nov 1998 20:56:53 -0800 From: thvv@multicians.org (Tom Van Vleck) Newsgroups: alt.folklore.computers,alt.os.multics Distribution: world Subject: Re: Multics Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> Organization: Multicians X-Newsreader: MT-NewsWatcher 2.4.1 X-url: http://www.multicians.org/ Lines: 15 NNTP-Posting-Host: thvv.vip.best.com X-Trace: nntp1.ba.best.com 910241820 29761 206.86.18.21 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!peerfeed.ncal.verio.net!news1.best.com!news3.best.com!nntp1.ba.best.com!not-for-mail Barry Margolin wrote: > Multics was struggling within Honeywell long before PCs came along; > customers have told us about occasions where they had to convince their > sales rep to sell them a Multics system instead of GCOS. > > The problem was managerial, not technical. Fair enough. My summary of the story is, "They starved us for years, and then canceled us because we were too thin." Wes Burner, MIT's 645 salesman and later Director of the MIT Information Processing Center, used to say of the GE/Honeywell management, "Those are the kind of guys who could screw up a free lunch." And in fact, Honeywell was given a free lunch, a big head start in operating systems, when they got Multics with the GE merger, and they wasted it. ###### From: bhahn@spam-spam.go-away.com (Brendan Hahn) Newsgroups: alt.folklore.computers Subject: Re: Multics Message-ID: References: <36464aec.6858331@news.uunet.be> <71o5n3$tpr$1@irk.zetnet.co.uk> <3641a632.2735513@news.uunet.be> Organization: Transoft Corp Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-transfer-encoding: 8bit Reply-To: bhahn@transoft.mangle.net (unmangle address to reply) X-Newsreader: Yet Another NewsWatcher 2.4.0 Lines: 13 Date: Thu, 05 Nov 1998 01:38:24 GMT NNTP-Posting-Host: 205.180.87.35 NNTP-Posting-Date: Wed, 04 Nov 1998 17:38:24 PDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news-feed.inet.tele.dk!bofh.vszbr.cz!newshub.northeast.verio.net!news-pen-3.sprintlink.net!news-in-east1.sprintlink.net!news.sprintlink.net!netnews1.nw.verio.net!netnews.nwnet.net!nnrp2.ni.net!not-for-mail lucvdv@null.net (Luc Van der Veken) wrote: >I meant it a little differently - how standardized are MIN_INT >and MAX_INT: do all compilers know them? Are they ANSI standard? >(Especially because of the MIN_INT / INT_MIN matter: I'm still >not sure, but it feels like I've met both of them - as well as >MININT and MAXINT without the underscores.) INT_MIN and INT_MAX (along with CHAR, LONG, ULONG, etc. counterparts) are ANSI. ISO adds WCHAR_MIN and WCHAR_MAX for the wchar_t type, I believe. There's also CHAR_BIT, and a bunch of similar definitions for characteristics of the floating-point implementation. bhahn@transoft.mangle.net <-- unmangle to reply ###### Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics References: <71huie$meo$2@irk.zetnet.co.uk> Organization: Wizvax Communications, LLC From: multics@wizvax.wizvax.net (Richard Shetron) NNTP-Posting-Host: wizvax.wizvax.net Message-ID: <36419fe8.0@news.wizvax.net> Date: 5 Nov 1998 07:54:00 -0500 X-Trace: 5 Nov 1998 07:54:00 -0500, wizvax.wizvax.net Lines: 25 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!newspump.monmouth.com!newspeer.monmouth.com!newsfeed.wizvax.net!news.wizvax.net!wizvax.wizvax.net!multics In article , Tom Van Vleck wrote: >Barry Margolin wrote: >> Multics was struggling within Honeywell long before PCs came along; >> customers have told us about occasions where they had to convince their >> sales rep to sell them a Multics system instead of GCOS. >> >> The problem was managerial, not technical. > >Fair enough. My summary of the story is, "They starved us for years, >and then canceled us because we were too thin." > >Wes Burner, MIT's 645 salesman and later Director of the MIT Information >Processing Center, used to say of the GE/Honeywell management, "Those >are the kind of guys who could screw up a free lunch." And in fact, >Honeywell was given a free lunch, a big head start in operating >systems, when they got Multics with the GE merger, and they wasted it. The story I was told back in the late 70's is it took USL several months to even get the local HIS office to even admit Multics existed and had to threaten to buy a NON-HIS system to get the sale. I was also told that even after the 1977 announcement, selling a Multics system was not included in the sales goals/performance for HIS offices, only GCOS counted as a sale. ###### From: jadams@NO.SPAM.PLEASE (James W. Adams) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 5 Nov 1998 14:17:19 -0500 Organization: UNIX Internals, Charlottesville, VA Lines: 65 Message-ID: <71stjv$fio@jamesa.cstone.net> References: <71huie$meo$2@irk.zetnet.co.uk> <364092B2.8277B176@nvi.to> NNTP-Posting-Host: jamesa.cstone.net X-Trace: Skuzzy.cstone.net 910293718 27948 206.205.42.225 (5 Nov 1998 19:21:58 GMT) X-Complaints-To: usenet@news.cstone.net NNTP-Posting-Date: 5 Nov 1998 19:21:58 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!howland.erols.net!dca1-hub1.news.digex.net!dca1-feed4.news.digex.net!digex!news1.cstone.net!news.COMET.NET!not-for-mail >Tom Van Vleck wrote: > > Multics died because Honeywell management didn't see a way > to make money from it in the mid 80s. Maybe they wre right. I would never claim to have Tom's insight into MULTICS management, but I would claim to have a pretty good window into Honeywell LISD management in those days as an important customer. I disagree strongly with some of these statements. One wildcard in the mix was that by the early 1980s, CPU design had been given over to NEC, who had no particular interest in MULTICS and seemed to favor GCOS. I had schematics for the board level design of the S-800 series of CPUs, and they all had NEC all over them. They went on to produce the S/1000 CPU which was the fastest thing on the planet for a while. IIRC, it ran GCOS. The minicomputer issue is a no-op because VMS only prevailed by internal politics over Jupiter, and shipped with virtually no applications. Billerica already had a well established mini business, which didn't compete with DPS/8 because L/66 and DPS-8 reflected hardware which could kick IBM's collective ass. Honeywell's marketing arm were mushroom-managed so they had no clue why a MULTICS system might meet a customer's needs better than a GCOS system. Buying a MULTICS system required the customer to extract it by brute force from the sales reps. The differences between DPS-8 and DPS-8M were pretty trivial in the overall scheme of things. Agreeably, LISD management were clueless, but this was largely a marketing issue, not a hardware one. OTOH, I have to place a certain level of blame at the Multicians' feet. UNIX had already shown the advantages of OS portability, yet MULTICS had yet to be rewritten to run on commodity hardware. It depended totally on Level/68 hardware which was overpriced and had performance roughly equal to the VAX. Just remember, though, that the Jupiter had better performance than the Venus and was canned for political reasons. OTOH, GCOS-8 ran on DPS-88 and DPS-90 hardware which had extremely fast ECL hardware. PL/I needed portability. I think the real reasons for the demise of MULTICS were a lack of corporate vision, a lack of vision on the part of Multicians of the need to separate their OS from the hardware, and a generally increasing societal lack of appreciation for elegance which is ongoing and parasitic. The "dumbing down" of society, so to speak. Pathetic, and it leaves those of us with talent, intelligence and ingenuity at a loss. Particularly those of us with any sort of moral scruples. Perhaps the real reason for the demise of MULTICS was the sheer lack of people who could appreciate it amid a generation of "yuppies" whose main goal seemed to be "mobilizing" the capital "ballast" from the American economy and snorting it up their noses. -- James W. Adams -- jamesa @ cstone.net <-- remove spaces "I became obsessed with angels and ballerinas, things of grace and beauty, otherworldly." Charlottesville, VA 22903 --C. Love ###### From: Barry Margolin Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics References: <71huie$meo$2@irk.zetnet.co.uk> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> Organization: GTE Internetworking, Cambridge, MA X-Newsreader: trn 4.0-test60 (5 October 1997) Originator: barmar@bbnplanet.com (Barry Margolin) Lines: 33 Message-ID: <9xo02.75$KS2.941224@burlma1-snr1.gtei.net> Date: Thu, 05 Nov 1998 21:12:37 GMT NNTP-Posting-Host: 4.2.32.97 X-Trace: burlma1-snr1.gtei.net 910300357 4.2.32.97 (Thu, 05 Nov 1998 21:12:37 GMT) NNTP-Posting-Date: Thu, 05 Nov 1998 21:12:37 GMT Distribution: world Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!howland.erols.net!cpk-news-hub1.bbnplanet.com!news.news.gtei.net!burlma1-snr1.gtei.net!not-for-mail In article <71stjv$fio@jamesa.cstone.net>, James W. Adams wrote: >OTOH, I have to place a certain level of blame at the Multicians' >feet. UNIX had already shown the advantages of OS portability, >yet MULTICS had yet to be rewritten to run on commodity hardware. >It depended totally on Level/68 hardware which was overpriced and >had performance roughly equal to the VAX. Just remember, though, >that the Jupiter had better performance than the Venus and was >canned for political reasons. OTOH, GCOS-8 ran on DPS-88 and DPS-90 >hardware which had extremely fast ECL hardware. PL/I needed portability. This was well understood. Towards the end, we investigated porting Multics to commodity Intel chips. When Honeywell showed no interest in funding this, Olin Sibert started a company with the hope of licensing the Multics source code so they could port it, but Honeywell (or was it Bull by that time?) wouldn't make a deal he could afford. Note that making Multics portable wouldn't have been easy. The biggest problem is that PL/I requires you to specify sizes of variables in bits, so the source code was riddled with things like 'fixed bin(35)' and 'fixed bin(17)', and structures with padding bits to fill out 36-bit words. Porting to a 32-bit processor requires finding all of these and fixing them up. The Opus project, which the Cambridge Multicians were working on after Multics was cancelled, anticipated adopting some code from Multics, so they were developing tools to help with this, as well as a macro facility for the compiler so that we wouldn't make the same mistake again, but I don't recall how complete they were. -- Barry Margolin, barmar@bbnplanet.com GTE Internetworking, Powered by BBN, Burlington, MA *** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups. Don't bother cc'ing followups to me. ###### From: karger@watson.ibm.com (Paul A. Karger) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 6 Nov 1998 02:08:40 GMT Organization: IBM T.J. Watson Research Center Lines: 63 Message-ID: <71tln8$qmm$1@mdnews.btv.ibm.com> References: <71huie$meo$2@irk.zetnet.co.uk> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> NNTP-Posting-Host: watpub1.watson.ibm.com X-Newsreader: xrn 8.01 Originator: karger@watson.ibm.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news.idt.net!nyd.news.ans.net!newsjunkie.ans.net!newsfeeds.ans.net!news.chips.ibm.com!newsfeed.btv.ibm.com!mdnews.btv.ibm.com!karger In article <71stjv$fio@jamesa.cstone.net>, jadams@NO.SPAM.PLEASE (James W. Adams) writes: |> |> OTOH, I have to place a certain level of blame at the Multicians' |> feet. UNIX had already shown the advantages of OS portability, |> yet MULTICS had yet to be rewritten to run on commodity hardware. |> It depended totally on Level/68 hardware which was overpriced and |> had performance roughly equal to the VAX. Just remember, though, |> that the Jupiter had better performance than the Venus and was |> canned for political reasons. OTOH, GCOS-8 ran on DPS-88 and DPS-90 |> hardware which had extremely fast ECL hardware. PL/I needed portability. |> I was at Digital at the time Jupiter was cancelled, and the direct cause of the cancellation was technical, rather than political. The political causes were actually more indirect. By the time Jupiter was being finished, DEC management had realized that killing off the 36-bit line was not good for the company, and they really were backing it. There was a lot of enthusiasm throughout the company for Jupiter and everyone was surprised and disappointed with what happened. Jupiter (the follow-on DECSystem-20 processor) was faster than Venus (8600) on some instructions, but some instructions, crucial to TOPS-20 performance were not faster. The overall performance gain was much less than had been expected. Delaying the CPU another year or so to fix the problems did not seem viable when it was already so late. That was the direct technical reason. The indirect political reason requires going back further in Digital history. Prior to Jupiter, there was another KL-10 follow-on that was cancelled in favor of the VAX 11/780. As a result of that cancellation, many of the key 36-bit CPU designers either left or transfered to other projects. In particular, the key 36-bit designer (who I shall leave nameless in the interests of his privacy, since I don't have permission to name him here) had transfered to the Venus (8600) project, because it was very late. As a result, the lead designer on Jupiter was someone who hadn't a lot of DECSystem-20 experience, and didn't know which instructions were most crucial to TOPS-20 performance. The 36-bit designer did get the 8600 fixed and shipping, albeit quite late. The 8600 being late meant that Nautilus (8500/8700/8800) came out only a few months later with essentially similar performance in a much lower cost box. The Nautilus designers also got a lot better performance than they had planned on - an unexpected happy surprise. On the question of Multics portability: At the time Multics was designed, the idea of writing an operating system in a higher level language was very radical, and making it portable was totally unheard of. Furthermore, there were no other processors at the time that could have run Multics. The 360/67 was the only other CPU with paging and segmentation, and the 360 had been explicitly rejected for Multics. Multics also was pushing the state of the art in so many ways, that to add yet another radical requirement - portability - would probably have doomed it to total failure. As it was, it took a LOT longer to get it working than any of the original designers had expected. When UNIX was first ported from the PDP-11 to the Interdata computers, this was a major breakthrough in computer science. No one had EVER ported an operating system to an entirely different CPU before. By that time, Multics was already quite mature, and porting would have been a huge effort in an environment where Honeywell never was willing to fund Multics properly just to survive, let alone be ported to another CPU. ###### From: jws@jws.nospam.ni.net (Jim Stephens (root@jws.ni.net)) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Message-ID: <3642a15b.1541332@news.ni.net> References: <71huie$meo$2@irk.zetnet.co.uk> <36419fe8.0@news.wizvax.net> X-Newsreader: Forte Free Agent 1.11/32.235 Lines: 54 Date: Fri, 06 Nov 1998 07:19:26 GMT NNTP-Posting-Host: 206.16.201.187 NNTP-Posting-Date: Thu, 05 Nov 1998 23:19:26 PDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsfeed.cwix.com!192.220.250.21!netnews1.nw.verio.net!netnews.nwnet.net!nnrp2.ni.net!not-for-mail On 5 Nov 1998 07:54:00 -0500, multics@wizvax.wizvax.net (Richard Shetron) wrote: >In article , >Tom Van Vleck wrote: >>Barry Margolin wrote: >>> Multics was struggling within Honeywell long before PCs came along; >>> customers have told us about occasions where they had to convince their >>> sales rep to sell them a Multics system instead of GCOS. >>> >>> The problem was managerial, not technical. >> >>Fair enough. My summary of the story is, "They starved us for years, >>and then canceled us because we were too thin." >> >>Wes Burner, MIT's 645 salesman and later Director of the MIT Information >>Processing Center, used to say of the GE/Honeywell management, "Those >>are the kind of guys who could screw up a free lunch." And in fact, >>Honeywell was given a free lunch, a big head start in operating >>systems, when they got Multics with the GE merger, and they wasted it. > >The story I was told back in the late 70's is it took USL several months >to even get the local HIS office to even admit Multics existed and had >to threaten to buy a NON-HIS system to get the sale. > I heard on some authority, that they had only to find the right connection to sell the system. The story went that there was a large scale purchase done for a state university and that it was LSU's, Baton Rouge's turn. The were waiting for some sort of IBM mainframe, maybe a 360/91 or something like that to come out and had defered till it was available. the money was taken for that due to deadlines on getting the money spent before budget authority lapsed and was used for the Multics system. LSU got a shot the next year. I do not think anyone I ever knew from the top to the bottom of the USL staff wanted any other system than the multics, because of the user friendliness it would offer. The system they used was an RCA Spectra 70 of some old sort event at the time, and they had no use for any upgrade that stayed with the IBM architecture that it respresented. The main sales force in Louisianna was for other Honeywell lines such as air conditioning, etc, which at the time was still mixed in with the computer lines. BTW the USL Multics was up and online when I started there in Jun 1975, and had been up a few months at that time. >I was also told that even after the 1977 announcement, selling a Multics >system was not included in the sales goals/performance for HIS offices, >only GCOS counted as a sale. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Fri, 06 Nov 98 13:13:21 GMT Organization: UltraNet Communications, Inc. Lines: 35 Message-ID: <71utmo$r21$2@ligarius.ultra.net> References: <71huie$meo$2@irk.zetnet.co.uk> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> NNTP-Posting-Host: d2.dial-12.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 6 Nov 1998 13:31:04 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-dc.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsfeed.cwix.com!209.244.253.199!newsfeed.xcom.net!news.ultranet.com!d2 In article <71tln8$qmm$1@mdnews.btv.ibm.com>, karger@watson.ibm.com (Paul A. Karger) wrote: > >In article <71stjv$fio@jamesa.cstone.net>, jadams@NO.SPAM.PLEASE (James W. Adams) writes: >|> >|> OTOH, I have to place a certain level of blame at the Multicians' >|> feet. UNIX had already shown the advantages of OS portability, >|> yet MULTICS had yet to be rewritten to run on commodity hardware. >|> It depended totally on Level/68 hardware which was overpriced and >|> had performance roughly equal to the VAX. Just remember, though, >|> that the Jupiter had better performance than the Venus and was >|> canned for political reasons. OTOH, GCOS-8 ran on DPS-88 and DPS-90 >|> hardware which had extremely fast ECL hardware. PL/I needed portability. >|> > >I was at Digital at the time Jupiter was cancelled, and the direct cause of >the cancellation was technical, rather than political. The political causes >were actually more indirect. By the time Jupiter was being finished, DEC >management had realized that killing off the 36-bit line was not good for >the company, and they really were backing it. There was a lot of enthusiasm >throughout the company for Jupiter and everyone was surprised and disappointed >with what happened. In what plant were you working when this stuff happened? /BAH Subtract a hundred and four for e-mail. ###### From: atbowler@thinkage.on.ca (Alan Bowler) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 6 Nov 1998 18:22:24 GMT Organization: Thinkage Ltd. Lines: 9 Message-ID: <71vep0$cu9$1@demon.uunet.ca> References: <36464aec.6858331@news.uunet.be> <71o5n3$tpr$1@irk.zetnet.co.uk> NNTP-Posting-Host: 192.102.11.4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsfeed.direct.ca!news.uunet.ca!atbowler In article <71o5n3$tpr$1@irk.zetnet.co.uk> lisard@zetnet.co.uk writes: > >Well, it'll probably work on anything with a 2's complement architecture >where the compiled C code ignores overflow (as I believe compiled C code >is supposed to). Not quite. The C standard does not require recognizing or ignoring overflow. However, if an implementor decides to ignore overflow then a number of standard optimizations are no longer prohibited. ###### Newsgroups: alt.folklore.computers,alt.os.multics From: ehrice@his.com (Edward Rice) Subject: Re: Multics Message-ID: Organization: NDS Date: Fri, 06 Nov 1998 23:15:54 -0500 References: <71huie$meo$2@irk.zetnet.co.uk> <36419fe8.0@news.wizvax.net> NNTP-Posting-Host: ehrice.his.com Lines: 19 X-Authenticated-User: ehrice Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!212.63.192.161.MISMATCH!newshub.bart.net!news.tele2.nl!newsfeed1.swip.net!swipnet!news.idt.net!netnews.com!news-xfer.newsread.com!netaxs.com!newsread.com!news4.his.com!user In article <36419fe8.0@news.wizvax.net>, multics@wizvax.wizvax.net (Richard Shetron) wrote: > The story I was told back in the late 70's is it took USL several months > to even get the local HIS office to even admit Multics existed and had > to threaten to buy a NON-HIS system to get the sale. Obviously an exaggeration. At least the first part, about admitting it existed. > I was also told that even after the 1977 announcement, selling a Multics > system was not included in the sales goals/performance for HIS offices, > only GCOS counted as a sale. Not true. I assure you, the sales force was more than willing to collect checks based on Multics sales. ###### From: Barry Margolin Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics References: <71huie$meo$2@irk.zetnet.co.uk> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> Organization: GTE Internetworking, Cambridge, MA Originator: barmar@bbnplanet.com (Barry Margolin) Lines: 30 Message-ID: Date: Sat, 07 Nov 1998 00:02:49 GMT NNTP-Posting-Host: 4.2.32.97 X-Trace: burlma1-snr1.gtei.net 910396969 4.2.32.97 (Sat, 07 Nov 1998 00:02:49 GMT) NNTP-Posting-Date: Sat, 07 Nov 1998 00:02:49 GMT Distribution: world Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!cpk-news-hub1.bbnplanet.com!news.news.gtei.net!burlma1-snr1.gtei.net!not-for-mail In article <71tln8$qmm$1@mdnews.btv.ibm.com>, >When UNIX was first ported from the PDP-11 to the Interdata computers, this was >a major breakthrough in computer science. No one had EVER ported an operating >system to an entirely different CPU before. By that time, Multics was already >quite mature, and porting would have been a huge effort in an environment where >Honeywell never was willing to fund Multics properly just to survive, let alone >be ported to another CPU. And don't forget that Honeywell considered themselves primarily a hardware vendor, not OS vendor. Unix's other radical feature was that it was written by a third party, not an organization with a stake in the success of a particular computer system design. How many portable OSes have been written by hardware manufacturers? VAX/VMS was developed long after Unix had proven that it could be done, but AFAIK they didn't design it to be portable, either. For that matter, how many portable OSes have been implemented by *anyone*? MacOS runs on 68K and PPC processors, but it makes heavy use of a 68K emulator. I think NT runs on x86 and Alpha architecures. Symbolics Genera was ported to their Ivory processors, but there's quite a bit of conditional code in many low-level operators. It's nearly 3 decades after Unix was implemented, yet portable OSes other than Unix are still not really mainstream. Faulting the Multics developers for not doing this 20 years ago is really pushing it. -- Barry Margolin, barmar@bbnplanet.com GTE Internetworking, Powered by BBN, Burlington, MA *** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups. Don't bother cc'ing followups to me. ###### From: James howard Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Sat, 07 Nov 1998 02:17:57 -0500 Organization: University of Maryland Lines: 61 Message-ID: <3643F425.EE71E539@wam.umd.edu> References: <71huie$meo$2@irk.zetnet.co.uk> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> NNTP-Posting-Host: byzantine.student.umd.edu Mime-Version: 1.0 Content-Type: multipart/signed; protocol="application/x-pkcs7-signature"; micalg=sha1; boundary="------------ms4D8D5CE0D2B9B9F6ECE0A17B" X-Mailer: Mozilla 4.5 [en] (X11; U; FreeBSD 3.0-RELEASE i386) X-Accept-Language: en Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!cpk-news-hub1.bbnplanet.com!news.news.gtei.net!cpk-news-feed4.bbnplanet.com!cpk-news-feed1.bbnplanet.com!news.gtei.net!news.ums.edu!haven.umd.edu!hecate.umd.edu!not-for-mail This is a cryptographically signed message in MIME format. --------------ms4D8D5CE0D2B9B9F6ECE0A17B Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Barry Margolin wrote: > For that matter, how many portable OSes have been implemented by *anyone*? > MacOS runs on 68K and PPC processors, but it makes heavy use of a 68K > emulator. I think NT runs on x86 and Alpha architecures. Symbolics Genera NT also runs on PPC and MIPS (at least, 3.x did). Apple is supposedly porting Mac to x86 for OS X. Plan 9 is portable too. --------------ms4D8D5CE0D2B9B9F6ECE0A17B Content-Type: application/x-pkcs7-signature; name="smime.p7s" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="smime.p7s" Content-Description: S/MIME Cryptographic Signature MIIH2gYJKoZIhvcNAQcCoIIHyzCCB8cCAQExCzAJBgUrDgMCGgUAMAsGCSqGSIb3DQEHAaCC BcgwggKHMIIB8KADAgECAgI8rTANBgkqhkiG9w0BAQQFADCBuTELMAkGA1UEBhMCWkExFTAT BgNVBAgTDFdlc3Rlcm4gQ2FwZTEUMBIGA1UEBxMLRHVyYmFudmlsbGUxGjAYBgNVBAoTEVRo YXd0ZSBDb25zdWx0aW5nMSkwJwYDVQQLEyBUaGF3dGUgUEYgUlNBIElLIDE5OTguOS4xNiAx Nzo1NTE2MDQGA1UEAxMtVGhhd3RlIFBlcnNvbmFsIEZyZWVtYWlsIFJTQSBJc3N1ZXIgMTk5 OC45LjE2MB4XDTk4MDkyNzEzNDkwMloXDTk5MDkyNzEzNDkwMlowRjEfMB0GA1UEAxMWVGhh d3RlIEZyZWVtYWlsIE1lbWJlcjEjMCEGCSqGSIb3DQEJARYUaG93YXJkanBAd2FtLnVtZC5l ZHUwXDANBgkqhkiG9w0BAQEFAANLADBIAkEAvZ0pBGnHOH/T4duG4K6trTwXRo3V9wc3xWvB 2Dawe80iDOzMWq/J9zWl8dpmD9QooNbLqnjkRMmpxt38LAWFEwIDAQABo1QwUjARBglghkgB hvhCAQEEBAMCBaAwDgYDVR0PAQH/BAQDAgWgMAwGA1UdEwEB/wQCMAAwHwYDVR0jBBgwFoAU /j5gnGuMD7DYM8bKxh5YsHE4teAwDQYJKoZIhvcNAQEEBQADgYEAeRe85ROHZlyJ0af7gTHl D759LSjaZbhfGJdIGdUl/XLI5OAM4hoOeuXJgEMTEHFeL1QOw7Zhz5t3ZI0Qy5rYmP/zjv33 a77WLGagRIXnaYv+thpgbrGioYpAIeSa4gxaHQFfiu01c+g4jaZPQBrD9yr9hk55QiFhPZZU lzhtwOMwggM5MIICoqADAgECAgEKMA0GCSqGSIb3DQEBBAUAMIHRMQswCQYDVQQGEwJaQTEV MBMGA1UECBMMV2VzdGVybiBDYXBlMRIwEAYDVQQHEwlDYXBlIFRvd24xGjAYBgNVBAoTEVRo YXd0ZSBDb25zdWx0aW5nMSgwJgYDVQQLEx9DZXJ0aWZpY2F0aW9uIFNlcnZpY2VzIERpdmlz aW9uMSQwIgYDVQQDExtUaGF3dGUgUGVyc29uYWwgRnJlZW1haWwgQ0ExKzApBgkqhkiG9w0B CQEWHHBlcnNvbmFsLWZyZWVtYWlsQHRoYXd0ZS5jb20wHhcNOTgwOTE2MTc1NTM0WhcNMDAw OTE1MTc1NTM0WjCBuTELMAkGA1UEBhMCWkExFTATBgNVBAgTDFdlc3Rlcm4gQ2FwZTEUMBIG A1UEBxMLRHVyYmFudmlsbGUxGjAYBgNVBAoTEVRoYXd0ZSBDb25zdWx0aW5nMSkwJwYDVQQL EyBUaGF3dGUgUEYgUlNBIElLIDE5OTguOS4xNiAxNzo1NTE2MDQGA1UEAxMtVGhhd3RlIFBl cnNvbmFsIEZyZWVtYWlsIFJTQSBJc3N1ZXIgMTk5OC45LjE2MIGfMA0GCSqGSIb3DQEBAQUA A4GNADCBiQKBgQDEpeXU1NBfCALuByF9JL+ra44e6yAHAhWEa4/QkyQfG53uaLK5LE/pk2cX EBceoflDQSO5MKp2l7vz5/2BwLUxi/amUCZU8pUo6xmkHpcesOK4m8EEmjLQPAlsT+Q1T/B2 vwATA09FCGDz/LTQkAGKEsmcun9S6iqTNTY8POQ1LwIDAQABozcwNTASBgNVHRMBAf8ECDAG AQH/AgEAMB8GA1UdIwQYMBaAFHJJwnM0xlX0C3ZygX539IfnxrIOMA0GCSqGSIb3DQEBBAUA A4GBACzHgh8BQz4Hj+5pXKlkgvjAlq2TK8ubUNdAmoHCuqZ2nTyVQNxVweFVgnmrCimm1Qzh Vyg+j/m71d8Nk1iqWy2LjzPk3VgVNXZyFSm9QvRakgt3X50n25otThuCBo7SjVa7ld7bDGUF 3pWeAt1TF76+/GvDGiJ6FCthvcKfXnpaMYIB2jCCAdYCAQEwgcAwgbkxCzAJBgNVBAYTAlpB MRUwEwYDVQQIEwxXZXN0ZXJuIENhcGUxFDASBgNVBAcTC0R1cmJhbnZpbGxlMRowGAYDVQQK ExFUaGF3dGUgQ29uc3VsdGluZzEpMCcGA1UECxMgVGhhd3RlIFBGIFJTQSBJSyAxOTk4Ljku MTYgMTc6NTUxNjA0BgNVBAMTLVRoYXd0ZSBQZXJzb25hbCBGcmVlbWFpbCBSU0EgSXNzdWVy IDE5OTguOS4xNgICPK0wCQYFKw4DAhoFAKCBsTAYBgkqhkiG9w0BCQMxCwYJKoZIhvcNAQcB MBwGCSqGSIb3DQEJBTEPFw05ODExMDcwNzE3NTdaMCMGCSqGSIb3DQEJBDEWBBR779/80ki/ CKNx+NPAy89MEzxbtDBSBgkqhkiG9w0BCQ8xRTBDMAoGCCqGSIb3DQMHMA4GCCqGSIb3DQMC AgIAgDAHBgUrDgMCBzANBggqhkiG9w0DAgIBQDANBggqhkiG9w0DAgIBKDANBgkqhkiG9w0B AQEFAARAi3KBMgesOTyp1iydm41rx1VfHmKjrKMwuuH1oBO+aYoiUPaIuyIElajBzDWiYXFs BA2yTmIhucUwifqolYWF6Q== --------------ms4D8D5CE0D2B9B9F6ECE0A17B-- ###### Newsgroups: alt.folklore.computers,alt.os.multics From: exxdgdc@bath.ac.uk (Douglas Clark) Subject: Re: Multics Organization: Guest of the University of Bath Computing Services, UK Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <36419fe8.0@news.wizvax.net> Date: Sat, 7 Nov 1998 07:27:27 GMT Lines: 8 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!diablo.theplanet.net!dispose.news.demon.net!demon!nntp.news.xara.net!xara.net!rill.news.pipex.net!pipex!server1.netnews.ja.net!hgmp.mrc.ac.uk!pegasus.csx.cam.ac.uk!bath.ac.uk!exxdgdc I remember Quackenboss selling us Multics for the Avon Universities Computer Centre (Bristol and Bath) and he really put on a fabulous performance. -- Douglas Clark voice: +44 1225 427104 69 Hillcrest Drive, mailto: D.G.D.Clark@bath.ac.uk Bath, Somerset, BA2 1HD, UK Lynx: http://www.bath.ac.uk/~exxdgdc/lynx.html ###### Newsgroups: alt.folklore.computers,alt.os.multics From: hoh@approve.se.NO_JUNK_EMAIL (Goran Larsson) Subject: Re: Multics Message-ID: <1998Nov7.113115.24424@lorelei.approve.se> Originator: hoh@approve.se.NO_JUNK_EMAIL (Goran Larsson) Sender: hoh@lorelei.approve.se.NO_JUNK_EMAIL (Goran Larsson) Organization: [1] + 5934 done /bin/rm -rf ~/ & X-No-Archive: yes X-Newsreader: trn 4.0-test66 (4 June 1998) References: <71huie$meo$2@irk.zetnet.co.uk> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> Lines: 11 NNTP-Posting-Host: 130.244.106.133 X-Complaints-To: news-abuse@swip.net X-Trace: nntpserver.swip.net 910438709 130.244.106.133 (Sat, 07 Nov 1998 12:38:29 MET DST) NNTP-Posting-Date: Sat, 07 Nov 1998 12:38:29 MET DST X-Sender: s-64812@dialup106-9-5.swipnet.se Date: Sat, 7 Nov 1998 11:31:15 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news.tele2.nl!newsfeed1.swip.net!swipnet!nntpserver.swip.net!not-for-mail In article , Barry Margolin wrote: > For that matter, how many portable OSes have been implemented by *anyone*? VxWorks. -- Goran Larsson hoh AT approve DOT se I was an atheist, http://home DOT swipnet DOT se/hoh/ until I found out I was God. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Sat, 07 Nov 98 12:44:10 GMT Organization: UltraNet Communications, Inc. Lines: 13 Message-ID: <721gcc$upm$2@strato.ultra.net> References: <71huie$meo$2@irk.zetnet.co.uk> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> <3643F425.EE71E539@wam.umd.edu> NNTP-Posting-Host: d3.dial-15.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 7 Nov 1998 13:02:04 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news.tele2.nl!newsfeed1.swip.net!swipnet!feed1.news.rcn.net!rcn!news.ultranet.com!d3 In article <3643F425.EE71E539@wam.umd.edu>, James howard wrote: >This is a cryptographically signed message in MIME format. >< Content-Type: application/x-pkcs7-signature; Filename: smime.p7s > > You have just guaranteed that I will never read any of your posts again. Keep your bits to yourself. /BAH Subtract a hundred and four for e-mail. ###### From: "John F. Couleur" Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Sat, 07 Nov 1998 12:24:19 -0700 Organization: ICGNetcom Lines: 152 Message-ID: <36449E63.65F3@ix.netcom.com> References: <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> NNTP-Posting-Host: phn-az20-30.ix.netcom.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-NETCOM-Date: Sat Nov 07 1:24:19 PM CST 1998 X-Mailer: Mozilla 3.0C-NC320 (Win95; U) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!ix.netcom.com!news Tom Van Vleck wrote: > > Richard Plinston wrote: > > AFAIK Multics died because it was horribly inefficient. > > Multics died because Honeywell management didn't see a way > to make money from it in the mid 80s. Maybe they wre right. > > Multics is a mainframe timesharing operating system. > In some sense the mainframe became obsolete when the minicomputer > came out in the early 70s, and we were too stubborn to notice. > When PCs came out in the late 70s and early 80s, the minicomputer > became obsolete as well. Each new smaller generation of machine > began with the idea that a simple operating system would be > more efficient and would be adequate for the new environemnt. > In each case, the complexity that was pruned off grew back. > (Some would say, in deformed and debased versions.) > > By the mid 80s, Multics was suffering because > it was running on CPUs at least a hardware generation behind the > (mainframe) state of the art. (Because the product was low volume > compared to GCOS, Multics CPUs took longer to build and required > more hand testing and debugging. Factory capacity for Multics CPUs > was limited by the number of "test cells," where newly built CPUs > were tested and debugged. I remember losing the battle for another > cell in the late 70s; each cell tied up a lot of capital on the > factory floor in Deer Valley.) The Multics subcommittee of HLSUA > (Honewyell Large System Users' Association) pressured management > for a new machine, and several projects were begun, and then > canceled. Without a new hardware base, sales forecasts dropped > to near zero. Management stopped development of the software. > I don't see this as a sinister process, just arithmetic. > > The economies of scale that mainframe systems optimize were no longer > important. Mainframe operating systems tried to use the expensive CPU > as efficiently as possible. Now we have cheap CPUs a hundred times as > fast sitting idle, running screen savers, playing Life. Sure, we > could have rewritten Multics to run on a large Sequent multi-386. > Sure, we could rewrite it to run on a PowerPC. But it would take > a lot of redesign, and that costs money and time. Honeywell > management didn't see it paying off. (Look where Honeywell is now.) To alt.os.multics readers: Please excuse any protocol transgressions. This is the first news letter that I have submitted. I would appreciate any corrections as to form. Hopefully, the observations are accurate after all of these years. I saw the "what happened to MULTICS?" discussion going by in the newsgroup and finally decided to stick in another oar. In 1978, I was in Dewey Manzer's product planning group in Phoenix, and was party to many of the discussions that were going on concerning what to do about MULTICS. At the time, Honeywell was producing the 6000/8000 (?) on the NSA platform but running GCOS. Only University of Waterloo and WWMCCS had complete access to the features of the NSA architecture, and no one was allowed to speak or write about features of NSA. The NSA architecture was a complete secret outside of Honeywell. Inside of Honeywell, the NSA was viewed as an architecture that corrected major flaws of the 645 - small address, page sized segments, limited protection for need to know, etc. CP-6 used the virtual memory features of NSA and GCOS was moving to using the 34 bit address. Reverting to the MULTICS architecture meant taking a giant step backward architecturally for GCOS. Dick Douglas had just taken control of the Honeywell Computer Operation in Phoenix, and was being beseiged by proposals to cap GCOS and go with MULTICS. In addition to MULTICS' obsolete architecture, a major problem legislated against replacing GCOS with MULTICS. Honeywell was selling over $500 million per year of "equivalent" GCOS systems. Very few of them were new systems. We were selling add-ons and upgrades for old systems. And, the market for sales to old customers was growing by 20% per year. It looked like we had a highly profitable money machine. It took one snip of the wire cutters to speed up the processor and increase the value of an installed system by $1 million. As far as MULTICS was concerned, any of our current customers that wanted a MULTICS system could have one and either replace his GCOS machines or run it along side of the GCOS machines. Very few did. A cutover to MULTICS would have meant throwing away a $500 million per year business for a highly questionable business. At the time, the only way that we could make MULTICS sales was to give away additional processors and memories to get the performance/cost up to acceptable levels. Obviously, none of the other studies had considered the economics of the decision. The question that Dick asked was "is there any reason for continuing MULTICS in the face of the business that we now have?" For the first time, I had occasion to examine the relationship between the MULTICS system and the GCOS system. Personally, I favored interactive computing and had no use for the complexities of programming a batch computer. Yet, a lot of people obviously did not see things that way. Apparently, there were really two worlds (at least) of computing. One world was geared to production, just like a manufacturing organization. There was a design staff and an operation staff. The output was never used by the designers. The output went to produce paychecks, ledgers, inventory reports, and financial reports that management could read on the way home. This computer was a production machine and like other production machines was evaluated on its ability to save/generate money for the company. The amount of money saved could be quantified, and therefore the cost of the machine could be justified to management. The other purpose was to support the intellectual activities of the organization. As an intellectual tool, the computing system had to conform to the needs of the user. Efficiency and performance/cost were not the metrics of this computer since its value was measured by the leverage it gave the engineer or analyst. Given its end use, the features of this system could be specified. Simplicty, a high degree of interactivity, and no separation between programming and operation. The computer/terminal had to be in the room with the person that it was supporting. It did not seem possible that a single system (hardware plus software) could serve both applications well. Therefore, both systems should be continued. GCOS would be continued for its profitability in the batch world and MULTICS would be the entre to the world of intellectual support. In order to give MULTICS every opportunity to succeed, MULTICS was put under a "czar" who was given authority to set prices and margins down to production cost. In effect, with this arrangement GCOS' high profitability would subsidize MULTICS and avoid putting MULTICS on a profit or die basis. Unfortunately, the value of intellectual support is not easily quantified and management had to take the benefits on faith. Instead of having faith, management assessed the users of the computer a share of the cost and operating expense of the computer on a minute used basis. When this computer was a mainframe class computer, the cost of the computer included 5 times more development, overhead and sales dollars than production dollars. Even with a mini computer the ratio was 3:1. The user paid $5-$10 for every $1's worth of computing that he got. Sponteneity and exploration, which were important requirements of an intellectual support computer, were destroyed by having to keep an eye on a very expensive clock. It wasn't until the microcomputer came along that computing for intellectual support became economically viable. And even then it was sold initially on the economic basis that it supported spread sheets which reduced a three man month budget process to three man hours. I hope that by bringing economics into the discussion you do not feel that I have cast aspersions on MULTICS. As James Carvelle observed, "that's what it's all about". As a further thought, has anyone besides me noticed the similarity between MULTICS and Windows9x. When my young friends from Microsoft tell me about the wonderful things that they are going to do in Washington, I have to tell them that the idea is at least 35 years old. DLLs, OLE, virtual memory, ODBC, heirarchical file system, ease of use, all were part of the original MULTICS. John Couleur jcouleur@ix.netcom.com ###### Newsgroups: alt.folklore.computers Subject: Re: Multics References: <71o5n5$tpr$2@irk.zetnet.co.uk> Organization: Wizvax Communications, LLC From: multics@wizvax.wizvax.net (Richard Shetron) NNTP-Posting-Host: wizvax.wizvax.net Message-ID: <3644a909.0@news.wizvax.net> Date: 7 Nov 1998 15:09:45 -0500 X-Trace: 7 Nov 1998 15:09:45 -0500, wizvax.wizvax.net Lines: 26 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newspeer.monmouth.com!newsfeed.wizvax.net!news.wizvax.net!wizvax.wizvax.net!multics In article <71o5n5$tpr$2@irk.zetnet.co.uk>, wrote: [snip] >So it sounds like what you're all saying about Multics is: > >(a) in today's terms, it's a paragon of efficiency; > >[a bit of a red herring, since it wasn't targeted at today's systems] I was certianly able to get much more work done per hour on a Multics system then any other system I've ever worked on in the 30 years I've been involved with computers. >(b) it ran just fine on high-end systems, making it an excellent > high-end timesharing OS - which was the intention of the designers; > >(c) it grumbled a bit in resource-starved situations. > >Can we agree, then, that it was "inefficient on small systems" without >implying inefficiency in general? The single CPU H-6180 with/3MB magnetic core and 4KW(?) cache system I worked on at Griffiss was about as responsive as win98 on my P200 with 64MB and the system at Griffiss normally had 30 or so other people on it. In General, win98 tends to be slower and less responsive. ###### From: ftit@engin.umich.edu (Sergej Roytman) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 7 Nov 1998 21:14:57 GMT Organization: University of Michigan, College of Engineering Lines: 47 Message-ID: <722d8h$qka@srvr1.engin.umich.edu> References: <71huie$meo$2@irk.zetnet.co.uk> <36449E63.65F3@ix.netcom.com> NNTP-Posting-Host: az.engin.umich.edu Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsfeed.cwix.com!128.174.5.49!vixen.cso.uiuc.edu!uwm.edu!msunews!srvr1.engin.umich.edu!ftit In article <36449E63.65F3@ix.netcom.com>, John F. Couleur wrote: >As a further thought, has anyone besides me noticed the similarity >between MULTICS and Windows9x. When my young friends from Microsoft tell >me about the wonderful things that they are going to do in Washington, I >have to tell them that the idea is at least 35 years old. DLLs, OLE, >virtual memory, ODBC, heirarchical file system, ease of use, all were >part of the original MULTICS. First of all, thank you for the different perspective on the decisions that affect the life of a computer system. I have always approached my machines from a programmer's point of view, and I guess, have sort of forgotten that not everyone thinks of programming as an art, or values an elegant design for its own sake. Strange as it may sound, I needed reminding that there is another way to view these things, and that it looks not at what a clean architecture this machine has, but at how many dollars it can generate. Not that I'm naive; I'm working at a large automobile company as I'm doing grad school, so I must have encountered this attitude before, but somehow it never really sank in. In my opinion, it's not a good move to impose restrictive structure on a creative effort. Even though it makes a certain sort of sense to commoditize everything, and assign a neat number to it, it seems like a bunch of overgrown bean-counting to me, and I think it always ends up missing something. I still like the feeling or spirit of some things, and would be sad to see them disappear. But that is as I said, my opinion. What I wanted to comment on was that last part, about Microsoft's "new" features. I was reading a recent OS book, to update my knowledge, a while ago (I do AI, so do not get much exposure to new low-level stuff unless I look for it myself). It struck me that Microsoft is just now blundering upon things that better, smarter programmers had invented *thirty* *years* *ago!* Only now, programming no longer involves squeezing the last drop of performance out of a slow, 32K mainframe, and the money keeps coming in no matter what the code looks like. Things become bloated out of all proportion. I read once that aeronautical engineers say that with a powerful enough motor, you can make a barn door fly. I wonder if this isn't happening in the software world now. I just hope that I'll continue to be able to avoid using your young friends' products. It will be a sad day in my little cubicle in Dilbert-land if they ever swap out my workstation for an NT box. -- Sergej Roytman ###### Newsgroups: alt.folklore.computers,alt.os.multics From: alderson@netcom.netcom.com (Richard M. Alderson III) Subject: Re: Multics In-Reply-To: Barry Margolin's message of Sat, 07 Nov 1998 00:02:49 GMT Message-ID: Sender: alderson@netcom.netcom.com Reply-To: alderson@netcom.com Organization: NETCOM On-line services References: <71huie$meo$2@irk.zetnet.co.uk> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> Date: Sat, 7 Nov 1998 21:40:40 GMT Lines: 21 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!howland.erols.net!ix.netcom.com!netcom!alderson In article Barry Margolin writes: >For that matter, how many portable OSes have been implemented by *anyone*? >MacOS runs on 68K and PPC processors, but it makes heavy use of a 68K >emulator. I *believe* that the latest versions (8.x) are parallel compilations, with PPC- native code for the OS. That was certainly the trend as early as 7.5, anyway. >Symbolics Genera was ported to their Ivory processors, but there's quite a bit >of conditional code in many low-level operators. That's not really a port issue in the same sense as Unix or MacOS or NT; it's much more like the move from the 360 to the 370, since Ivory was a single-chip implementation of the 36xx architecture. (I wanted a Mac IIfx with the Ivory and Explorer co-processors so bad...) -- Rich Alderson Last LOTS Tops-20 Systems Programmer, 1984-1991 Current maintainer, MIT TECO EMACS (v. 170) last name @ XKL dot COM Chief systems administrator, XKL LLC, 1998-now ###### From: eva_remove_this_ns@evans.pgh.pa.us (Arthur Evans Jr) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Sat, 07 Nov 1998 22:14:03 -0500 Organization: Ada Consulting Lines: 35 Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <36419fe8.0@news.wizvax.net> NNTP-Posting-Host: ppp43.s8.pgh.net X-Newsreader: MT-NewsWatcher 2.3.1 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!howland.erols.net!newsfeed.fast.net!news.pgh.net!ppp43.s8.pgh.net!user In article <36419fe8.0@news.wizvax.net>, multics@wizvax.wizvax.net (Richard Shetron) wrote: > The story I was told back in the late 70's is it took USL several months > to even get the local HIS office to even admit Multics existed and had > to threaten to buy a NON-HIS system to get the sale. > > I was also told that even after the 1977 announcement, selling a Multics > system was not included in the sales goals/performance for HIS offices, > only GCOS counted as a sale. In about 1972 or 3 when I worked at Lincoln Laboratory, I helped prepare an RFP for a major computer purchase. It was clear to the RFP team that some flavor of 360 and Multics were strong contenders, as well as several other machines I no longer remember. We worked _very_ hard in writing the RFP to be fair -- we knew that no vendor could meet all of the requirements. OTH, we were clear in the RFP that we would consider proposals that failed to meet all requirements. IBM wrote two proposals, one that met one subset of the requirements and one that met a different subset. Not surprisingly, it was well done. Honeywell did not bid. Years later, I found out that they saw that they couldn't meet all the requirements, concluded that the RFP had been rigged for IBM (not uncommon in those days), and forgot about it. They failed to realize that the competition also could not meet all requirements. They also failed to realize that folks spending government money in those days could get into _serious_ trouble by rigging a bid for IBM. Mostly, they failed to try very hard. We bought a 360/67. Sigh......... Art Evans ###### From: Barry Margolin Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics References: <71huie$meo$2@irk.zetnet.co.uk> <71tln8$qmm$1@mdnews.btv.ibm.com> Organization: GTE Internetworking, Cambridge, MA X-Newsreader: trn 4.0-test60 (5 October 1997) Originator: barmar@bbnplanet.com (Barry Margolin) Lines: 22 Message-ID: Date: Sat, 07 Nov 1998 22:53:08 GMT NNTP-Posting-Host: 4.2.35.145 X-Trace: burlma1-snr1.gtei.net 910479188 4.2.35.145 (Sat, 07 Nov 1998 22:53:08 GMT) NNTP-Posting-Date: Sat, 07 Nov 1998 22:53:08 GMT Distribution: world Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsfeed.cwix.com!4.1.16.34!cpk-news-hub1.bbnplanet.com!news.news.gtei.net!burlma1-snr1.gtei.net!not-for-mail In article , Richard M. Alderson III wrote: >In article Barry Margolin > writes: >>Symbolics Genera was ported to their Ivory processors, but there's quite a bit >>of conditional code in many low-level operators. > >That's not really a port issue in the same sense as Unix or MacOS or NT; it's >much more like the move from the 360 to the 370, since Ivory was a single-chip >implementation of the 36xx architecture. (I wanted a Mac IIfx with the Ivory >and Explorer co-processors so bad...) I don't think so. The Ivory instruction set is completely different from that of the 36xx architecture. You had to recompile programs with the Ivory compiler to make them run on the new system (all the standard software was distributed with two binaries: .BIN and .IBIN). -- Barry Margolin, barmar@bbnplanet.com GTE Internetworking, Powered by BBN, Burlington, MA *** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups. Don't bother cc'ing followups to me. ###### From: hellsop@execpc.com (Peter H. Coffin) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Sat, 7 Nov 1998 23:42:59 -0600 Organization: ExecPC Internet - Milwaukee, WI Lines: 15 Approved: No Message-ID: <723b0e$88j@newsops.execpc.com> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> NNTP-Posting-Host: traken-70.isdn.mke.execpc.com X-Trace: gotham-globe.newsops.execpc.com 910503758 8467 (None) 169.207.67.198 X-Complaints-To: abuse@execpc.com X-No-Archive: Yes Mail-Copies-To: never X-Face: UlYAGAXCay!kP+LL$r`^aw{/W5<<=s!"D")0|s1#}ZM2YnXH2kq@2dPlI6FBT 8wCJ7YBj2x\^)-vi["NV~D'\"(S|+^@J'':9t3kwytu!d0:"aAQ!={]`61+ X-Newsreader: MacSOUP 2.3 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!howland.erols.net!ix.netcom.com!newsfeeds.sol.net!newsops.execpc.com!posts.execpc.com!gotham-globe.newsops.execpc.com!hellsop Barry Margolin wrote: > For that matter, how many portable OSes have been implemented by *anyone*? The most current OS/400 (v4r3) will still run programs compiled on the System 38 tranparently. That's 10-15 year old object code over at least one complete architecture change. The machine also has a history of being able to run in System 36 mode, over that same architecture change. I'm not sure if that counts as a ported OS or not, and I do work for IBM, so I'll stop now... -- "We've got a kitchen by Giger, but we're too scared to go in there." ###### Newsgroups: alt.folklore.computers From: ehrice@his.com (Edward Rice) Subject: Re: Multics Message-ID: Organization: NDS Date: Sun, 08 Nov 1998 02:25:33 -0500 References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> NNTP-Posting-Host: ehrice.his.com Lines: 16 X-Authenticated-User: ehrice Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news.tele2.nl!newsfeed1.swip.net!swipnet!howland.erols.net!news-xfer.newsread.com!netaxs.com!newsread.com!news4.his.com!user In article <71cf04$5fk@jamesa.cstone.net>, jadams@NO.SPAM.PLEASE (James W. Adams) wrote: > Later versions used increasing levels of > microcode. I recall having to "miniboot" a DPS-8 using > microcode on a 8 inch floppy disk which was read into > a writable control store. I have no recollection of this -- any details? The Microprogrammed Controllers had to be booted, and conceivable one might need to be fiddled to get the system to boot if interrupts weren't being honored, somehow, but I disremember any floppy drive on any DPS-8 I used. ###### Newsgroups: alt.folklore.computers From: ehrice@his.com (Edward Rice) Subject: Re: Multics Message-ID: Organization: NDS Date: Sun, 08 Nov 1998 02:25:37 -0500 References: <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> NNTP-Posting-Host: ehrice.his.com Lines: 34 X-Authenticated-User: ehrice Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news-xfer.newsread.com!netaxs.com!newsread.com!news4.his.com!user In article , thvv@multicians.org (Tom Van Vleck) wrote: > Perhaps you were present at a different set of benchmarks from > the ones I participated in. The ones I remember in the late 70s > placed us at the front of the pack, for large interactive mainframe > systems. In those days, the Honeywell sales strategy was to bid > GCOS, screw around trying to run the benchmark until the last > two weeks, then panic and call in the Multicians. We were able to > run quite a few benchmarks the competition (Including GCOS) couldn't, > and won quite a few bids. Uh... what bids did Multics pull out of the fire in the last two weeks, that GCOS couldn't handle? I think FSO had its fair share of competitive benchmarks, and I can't recall a single instance like that. I don't remember any competitive benchmark on which platforms were switched in the last /ten/ weeks. Multics was able to perform some workloads better than GCOS, and even perform some workloads that GCOS could not. *A SMALL PART* of that superiority was the artificial nature of benchmarks, which placed an unreasonably high premium on code-sharing. This was one of several things that bit customers when they got their own machines installed -- they found that compiling forty copies of a single program at once was very quick, while compiling forty comparable programs that were different was not so quick. Multics provided a feature set that GCOS couldn't touch, nor could any other system around. But when it came to raw COBOL and Fortran performance, Multics was not ahead of GCOS. Not even with FAST and DFAST, which came pretty late in the game. ###### From: Anne & Lynn Wheeler Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 08 Nov 1998 08:26:38 -0800 Organization: Wheeler&Wheeler Lines: 20 Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <36419fe8.0@news.wizvax.net> Reply-To: Anne & Lynn Wheeler NNTP-Posting-Host: lynn-18.garlic.com X-Newsreader: Gnus v5.6.44/Emacs 20.3 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news-feed.inet.tele.dk!bofh.vszbr.cz!news.maxwell.syr.edu!news.shore.net!uunet!in4.uu.net!bulb.garlic.com!not-for-mail >We bought a 360/67. Sigh......... >Art Evans LL had two processer 360/67 from 1967 that ran CP/67. about 73, LL discontinued one of the processors and sent it back. CSC had half of a two processor system .... and wanted an upgrade. After the processor had been picked up at LL, one of the people at CSC called the trucker and directed him to deliver to 545 Tech sq (instead of taking it back to plant). Took a long time before the plant was able to track it down. -- -- Anne & Lynn Wheeler | lynn@garlic.com, finger for pgp key http://www.garlic.com/~lynn/ ###### Newsgroups: alt.folklore.computers,alt.os.multics From: ehrice@his.com (Edward Rice) Subject: Re: Multics Message-ID: Organization: NDS Date: Mon, 09 Nov 1998 00:06:31 -0500 References: <71huie$meo$2@irk.zetnet.co.uk> <36419fe8.0@news.wizvax.net> NNTP-Posting-Host: ehrice.his.com Lines: 64 X-Authenticated-User: ehrice Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!isdnet!howland.erols.net!news-xfer.newsread.com!netaxs.com!newsread.com!news4.his.com!user In article , eva_remove_this_ns@evans.pgh.pa.us (Arthur Evans Jr) wrote: > Honeywell did not bid. Years later, I found out that they saw that they > couldn't meet all the requirements, concluded that the RFP had been > rigged for IBM (not uncommon in those days), and forgot about it. They > failed to realize that the competition also could not meet all > requirements. They also failed to realize that folks spending > government money in those days could get into _serious_ trouble by > rigging a bid for IBM. It happened all the time. And in Federal Systems, we had better sense than to go after /every/ target -- we couldn't /begin/ to address every RFP that came through the door. For the record, I /think/ I was the person who wrote the internal memo recommending a no-bid on Lincoln Lab. In addition to doing all the Federal benchmarking group's tape conversions, I did a whole bunch of "skim the RFP and decide whether or not this is worth anybody else looking at." I can remember making no-bid recommendations on quite a few projects for technical reasons, all of which were supported by my management because we were already working everybody about seventy hours a week. Those were the days of the BIG government procurements, and when we had to decide between a Navships (nine systems), an APDS (possibly a couple of triple-processors, although we later got the workload running on one triple and /still/ lost the bid), Army Personnel Information Systems Command (PERSINSCOM), Air Force BASETOPS, USDA-GSA central computer resource for the entire government (RFP eventually died without our help but it was a real biggie), and a processor or two for LL or FNWC or other places, our priority from management was clear: go after the enormous bids first. Of the RFPs we saw at Honeywell Federal Systems, there was no strong correlation between probable size of system being procured and probable cost of benchmark. While it's true that the WWMCCS benchmark was the biggest FSO/FSD did, there were a lot of piss-ant little procurements that wanted scores of programs to be run, and if we were going to convert scores of programs and put a team of 3-5 people in Phoenix for a month tuning a system, why not apply the effort to a major procurement? I sat on the FIPS task group to develop a benchmarking standard, and tried to work up a survey that vendors and government users could answer to try to find out just what resources were being expended for benchmarks, and absolutely nobody would cooperate by answering any questions about the benchmarks they had done or the effort they'd applied. Back on the subject of rigged RFPs, about 20-30% of the RFPs we saw were heavily slanted for one machine or another, with an IBM slant in about half of those cases. The IBM people at least had the sense to provide their contacts with things that /might/ fly when oversight occurred, like "Must be able to run 500 existing applications ranging from 10,000 source lines of COBOL to 100,000 source lines of FORTRAN IV without source program changes." I recall the Burroughs sites -- which were brand-loyal to a fault -- insisting on specific methods of solving problems, specific file formats which had to be supported as "native" formats by the operating system, right down to the necessary granularity of the real-time clock. We once threw the US Bureau of Reclamation for a loop by successfully de-virtualizing a B-5500 program which they'd been running, and getting it to run (just barely) on a GCOS system. I got to know the technical contact very well over time, and he told us that everybody there was /amazed/ that we'd managed to do that. (Speaking for myself and the two months of work I put in on that damned program, the amazement wasn't all on the government side, either.) ###### From: eva_remove_this_ns@evans.pgh.pa.us (Arthur Evans Jr) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Mon, 09 Nov 1998 08:14:31 -0500 Organization: Ada Consulting Lines: 9 Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> NNTP-Posting-Host: ppp5.s8.pgh.net X-Newsreader: MT-NewsWatcher 2.3.1 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!newspump.monmouth.com!newspeer.monmouth.com!howland.erols.net!newsfeed.fast.net!news.pgh.net!ppp5.s8.pgh.net!user Barry Margolin wrote: >For that matter, how many portable OSes have been implemented by *anyone*? A British OS written in BCPL was quite successfully ported to sevreral platforrms, probably in the late '70s or early '80s. Art Evans Make the obvious fix to my address to reply. ###### From: Vince Scarafino Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Mon, 09 Nov 1998 11:13:04 -0500 Organization: Ford Motor Company Lines: 64 Message-ID: <36471490.3EAC70F5@pobox.srl.ford.com> References: <71huie$meo$2@irk.zetnet.co.uk> <36449E63.65F3@ix.netcom.com> <722d8h$qka@srvr1.engin.umich.edu> Reply-To: vscarafi@pobox.srl.ford.com NNTP-Posting-Host: pms045.pd9.ford.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 4.5 [en]C-c32f404p (Win95; U) X-Accept-Language: en, en-GB, de, fr, ja, ko, zh Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsfeed.direct.ca!newsxfer3.itd.umich.edu!jobone!dailyplanet.srl.ford.com!eccws1.dearborn.ford.com!news Sergej Roytman wrote: > > In article <36449E63.65F3@ix.netcom.com>, > John F. Couleur wrote: > >As a further thought, has anyone besides me noticed the similarity > >between MULTICS and Windows9x. When my young friends from Microsoft tell > >me about the wonderful things that they are going to do in Washington, I > >have to tell them that the idea is at least 35 years old. DLLs, OLE, > >virtual memory, ODBC, heirarchical file system, ease of use, all were > >part of the original MULTICS. > > First of all, thank you for the different perspective on the decisions > that affect the life of a computer system. I have always approached my > machines from a programmer's point of view, and I guess, have sort of > forgotten that not everyone thinks of programming as an art, or values > an elegant design for its own sake. Strange as it may sound, I needed > reminding that there is another way to view these things, and that it > looks not at what a clean architecture this machine has, but at how > many dollars it can generate. Not that I'm naive; I'm working at a > large automobile company as I'm doing grad school, so I must have > encountered this attitude before, but somehow it never really sank in. > > In my opinion, it's not a good move to impose restrictive structure on > a creative effort. Even though it makes a certain sort of sense to > commoditize everything, and assign a neat number to it, it seems like a > bunch of overgrown bean-counting to me, and I think it always ends up > missing something. I still like the feeling or spirit of some things, > and would be sad to see them disappear. But that is as I said, my > opinion. > > What I wanted to comment on was that last part, about Microsoft's "new" > features. I was reading a recent OS book, to update my knowledge, > a while ago (I do AI, so do not get much exposure to new low-level > stuff unless I look for it myself). It struck me that Microsoft is > just now blundering upon things that better, smarter programmers had > invented *thirty* *years* *ago!* There is much evidence in this. Microsoft is "reinventing" many things, and its implementation shows that they didn't bother to do any research to help understand what they are doing. The most blatant area that I see is the way they manage the namespace for dynamically linked objects. In reality, they don't bother to manage it! If you want to replace a dynamic link you have to reboot. I can find no concept of terminate reference. This is quite a bother when you are trying to develop applications that require changes in dynamically linked libraries. > Only now, programming no longer > involves squeezing the last drop of performance out of a slow, 32K > mainframe, and the money keeps coming in no matter what the code looks > like. Things become bloated out of all proportion. > > I read once that aeronautical engineers say that with a powerful enough > motor, you can make a barn door fly. I wonder if this isn't happening > in the software world now. I just hope that I'll continue to be able > to avoid using your young friends' products. It will be a sad day in > my little cubicle in Dilbert-land if they ever swap out my workstation > for an NT box. > > -- > Sergej Roytman ###### From: dpeschel@u.washington.edu (D. Peschel) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 9 Nov 1998 16:16:01 GMT Organization: University of Washington, Seattle Lines: 26 Message-ID: <7274g1$pem$1@nntp1.u.washington.edu> References: <71huie$meo$2@irk.zetnet.co.uk> NNTP-Posting-Host: saul6.u.washington.edu X-Trace: nntp1.u.washington.edu 910628161 26070 (None) 140.142.17.40 X-Complaints-To: help@cac.washington.edu NNTP-Posting-User: dpeschel Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newscore.univie.ac.at!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.direct.ca!news.u.washington.edu!dpeschel In article , Arthur Evans Jr wrote: >A British OS written in BCPL was quite successfully ported to >sevreral platforrms, probably in the late '70s or early '80s. Would that be TRIPOS? (or is it TripOS? I don't think so) The only things I know about it are: - Was used to create DOS on the Amiga (or some other parts of the OS) - May have nasty license requirements (apparently Amiga couldn't deliver some of the documentation they wanted) - Ported to a bunch of architectures - Supposedly a pretty clean design, though the BCPL runtime environment can interfere with the C runtime environment (probably also with Pascal and other environments too) - Named for an examination (at Cambridge or Oxford?) which in turn was named for the three-legged stool provided during the examination -- Is there another acronym? Was TRIPOS used as course material? -- Derek ###### From: atbowler@thinkage.on.ca (Alan Bowler) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 9 Nov 1998 18:16:18 GMT Organization: Thinkage Ltd. Lines: 28 Message-ID: <727bhi$om$1@nntp2.uunet.ca> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> NNTP-Posting-Host: 192.102.11.4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!newsfeed.wirehub.nl!howland.erols.net!torn!news.uunet.ca!atbowler In article ehrice@his.com (Edward Rice) writes: >In article <71cf04$5fk@jamesa.cstone.net>, >jadams@NO.SPAM.PLEASE (James W. Adams) wrote: > > > Later versions used increasing levels of > > microcode. I recall having to "miniboot" a DPS-8 using > > microcode on a 8 inch floppy disk which was read into > > a writable control store. > >I have no recollection of this -- any details? > >The Microprogrammed Controllers had to be booted, and conceivable one might >need to be fiddled to get the system to boot if interrupts weren't being >honored, somehow, but I disremember any floppy drive on any DPS-8 I used. The ELS-2 systems, DPS-8/47 and 8/49, were microcoded systems and the system did use an 8-inch floopy to store the microcode for the main CPU, and the dianostic software for the service processor. (The service processor loaded the microcode into the main CPU, and read internal state data data of the main processor). The ELS-2 was just a refinement on Toshiba's orginal ELS-1 design (DPS-8/20 and 8/44), so I suspect that those also used an 8-inch floppy. (Many of drawing that came with the DPS-8/49, still had Toshiba written on them). You don't hear much about the ELS-2 systems because they were sold mostly to smaller shops, but they were incredibly reliable. ###### From: dbarker@camosun.bc.nospam.ca (Deryk Barker) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Followup-To: alt.folklore.computers,alt.os.multics Date: 9 Nov 1998 21:36:19 GMT Organization: Camosun College, Victoria B.C. Lines: 28 Distribution: world Message-ID: <727n8j$qrs$1@news.islandnet.com> References: <71huie$meo$2@irk.zetnet.co.uk> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> NNTP-Posting-Host: 204.174.56.1 X-Newsreader: TIN [UNIX 1.3 950824BETA PL0] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!logbridge.uoregon.edu!news.islandnet.com!not-for-mail Barry Margolin (barmar@bbnplanet.com) wrote: [...] : vendor, not OS vendor. Unix's other radical feature was that it was : written by a third party, not an organization with a stake in the success : of a particular computer system design. How many portable OSes have been : written by hardware manufacturers? VAX/VMS was developed long after Unix : had proven that it could be done, but AFAIK they didn't design it to be : portable, either. And let us not forget, either, that the first version of Unix was written in PDP-11 assembler. Unless I am mistaken the rewrite in C (1973ish?) was in order to facilitate the port to the Unidata machine. : : For that matter, how many portable OSes have been implemented by *anyone*? : MacOS runs on 68K and PPC processors, but it makes heavy use of a 68K : emulator. I think NT runs on x86 and Alpha architecures. On top of the Mach (?) microkernel. As far as most of the OS is concerned there is a uniform underlying machine (the Hardware Abstraction Layer I believe). -- |Deryk Barker, Computer Science Dept. | Music does not have to be understood| |Camosun College, Victoria, BC, Canada| It has to be listened to. | |email: dbarker@camosun.bc.ca | | |phone: +1 250 370 4452 | Hermann Scherchen. | ###### From: dbarker@camosun.bc.nospam.ca (Deryk Barker) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Followup-To: alt.folklore.computers,alt.os.multics Date: 9 Nov 1998 21:44:17 GMT Organization: Camosun College, Victoria B.C. Lines: 37 Message-ID: <727nnh$qrs$2@news.islandnet.com> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> <723b0e$88j@newsops.execpc.com> NNTP-Posting-Host: 204.174.56.1 X-Newsreader: TIN [UNIX 1.3 950824BETA PL0] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!logbridge.uoregon.edu!news.islandnet.com!not-for-mail Peter H. Coffin (hellsop@execpc.com) wrote: : Barry Margolin wrote: : : > For that matter, how many portable OSes have been implemented by *anyone*? : : The most current OS/400 (v4r3) will still run programs compiled on the : System 38 tranparently. That's 10-15 year old object code over at least : one complete architecture change. The machine also has a history of : being able to run in System 36 mode, over that same architecture change. : : I'm not sure if that counts as a ported OS or not, and I do work for : IBM, so I'll stop now... Peanuts. When my wife was working for Honeywell, in the 1980s, one of the customers she had dealings with was British Telecom. BT, at one location, had what they called the "orange Leos". Now, for those who don't know this, the LEO was the world's first-ever commercially-oriented machine (1951). Even more amazingly, the Lyons Electronic Office was designed and built by the J Lyons company, best-known as manufacturers of cakes and for their nationwide chain of corner tea shops. Anyway, an "orange Leo" was an ICL 2900 mainframe (they came in orange cabinets), emulating an ICL 1900 mainframe, emulating a GEC System 4 mainframe emulating a LEO. 30+ year old executable code over 3 architecture changes.... -- |Deryk Barker, Computer Science Dept. | Music does not have to be understood| |Camosun College, Victoria, BC, Canada| It has to be listened to. | |email: dbarker@camosun.bc.ca | | |phone: +1 250 370 4452 | Hermann Scherchen. | ###### From: dbarker@camosun.bc.nospam.ca (Deryk Barker) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Followup-To: alt.folklore.computers,alt.os.multics Date: 9 Nov 1998 21:45:49 GMT Organization: Camosun College, Victoria B.C. Lines: 17 Message-ID: <727nqd$qrs$3@news.islandnet.com> References: <71huie$meo$2@irk.zetnet.co.uk> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> NNTP-Posting-Host: 204.174.56.1 X-Newsreader: TIN [UNIX 1.3 950824BETA PL0] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!logbridge.uoregon.edu!news.islandnet.com!not-for-mail Arthur Evans Jr (eva_remove_this_ns@evans.pgh.pa.us) wrote: : Barry Margolin wrote: : >For that matter, how many portable OSes have been implemented by *anyone*? : : A British OS written in BCPL was quite successfully ported to : sevreral platforrms, probably in the late '70s or early '80s. I believe this would have been Martin Richards' TRIPOS. Richards was the author of BCPL (and CPL before it) and I believe TRIPOS was the foundation for the Atari (or was it Amiga?) OS. Can anyone confirm this? -- |Deryk Barker, Computer Science Dept. | Music does not have to be understood| |Camosun College, Victoria, BC, Canada| It has to be listened to. | |email: dbarker@camosun.bc.ca | | |phone: +1 250 370 4452 | Hermann Scherchen. | ###### From: jones@cs.uiowa.edu (Douglas W. Jones,201H MLH,3193350740,3193382879) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 9 Nov 1998 22:16:49 GMT Organization: The University of Iowa Lines: 20 Distribution: world Message-ID: <727pkh$11j6$1@flood.weeg.uiowa.edu> References: <727n8j$qrs$1@news.islandnet.com> NNTP-Posting-Host: pyrite.cs.uiowa.edu Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newspeer.monmouth.com!chippy.visi.com!news-out.visi.com!uunet!in1.uu.net!news.uiowa.edu!not-for-mail From article <727n8j$qrs$1@news.islandnet.com>, by dbarker@camosun.bc.nospam.ca (Deryk Barker): > > And let us not forget, either, that the first version of Unix was > written in PDP-11 assembler. The first version was written in assembly language for an 18-bit PDP-9 (ancestor of the PDP-15, and descendant of the PDP-7) > Unless I am mistaken the rewrite in C (1973ish?) was in order to > facilitate the port to the Unidata machine. PDP-11, if memory serves me right. I believe the port to an Interdata machine was later. My first edition of K&R says that C was developed under UNIX for PDP-11 UNIX, and that, as of the first edition of K&R, it ran on the IBM 370, Honeywell 6000, and the Interdata 8/32. Doug Jones jones@cs.uiowa.edu ###### Message-ID: <3647B503.20EE4ABE@shore.net> From: John W Gintell Reply-To: gintell@shore.net Organization: JWG Software Systems X-Mailer: Mozilla 4.5 [en] (WinNT; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics References: <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 40 Date: Mon, 09 Nov 1998 22:37:39 -0500 NNTP-Posting-Host: 204.167.102.124 X-Complaints-To: abuse@shore.net X-Trace: news.shore.net 910668975 204.167.102.124 (Mon, 09 Nov 1998 22:36:15 EDT) NNTP-Posting-Date: Mon, 09 Nov 1998 22:36:15 EDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!cpk-news-hub1.bbnplanet.com!cam-news-feed2.bbnplanet.com!news.gtei.net!news.shore.net!not-for-mail Tom Van Vleck wrote: > > Richard Plinston wrote: > > AFAIK Multics died because it was horribly inefficient. > > Multics died because Honeywell management didn't see a way > to make money from it in the mid 80s. Maybe they wre right. > > Multics is a mainframe timesharing operating system. > In some sense the mainframe became obsolete when the minicomputer > came out in the early 70s, and we were too stubborn to notice. > When PCs came out in the late 70s and early 80s, the minicomputer > became obsolete as well. Each new smaller generation of machine > began with the idea that a simple operating system would be > more efficient and would be adequate for the new environemnt. > In each case, the complexity that was pruned off grew back. > (Some would say, in deformed and debased versions.) So now here it is 1998. Lets take a snapshot of the current status. Bull took over the Honeywell computer business and merged it with theirs - easy since they had identical products. Also because of their relationships with NEC they had a source of high-end hardware as well. Classical mainframes and midframes are still not dead. In fact, the mainframe business is thriving and is profitable for several vendors. Bull made a net profit in 96 and 97, and I believe they expect to do it again. Their combined product line is much simpler since they got rid of their fringe products and focused on their strengths. Also they used emulation for their mini computers- Level 6 emulation on a PowerPC running AIX works pretty well so they didn't have to continue that hardware product line. They do about 4 billion dollars of business a year. Not huge, but not insignificant. Do they have a long term future as the nature of IS continues to evolve? probably not? but that is true for lots of companies. Given the technical, economic, and political realities, I think they did the right thing. Note that since Honeywell jettisoned its computer business and a lot of its military business it has become quite profitable. Do I like the result? No, of course not; a great opportunity was lost with Multics. But a lot of great stuff has died or is dying a slow death. ###### Newsgroups: alt.folklore.computers,alt.os.multics From: bdb@GTS.Net (Hello Kittyhawk) Subject: Re: Multics Message-ID: Organization: G.T.S., Toronto, Ontario X-Newsreader: trn 4.0-test66 (4 June 1998) References: <71huie$meo$2@irk.zetnet.co.uk> <727nqd$qrs$3@news.islandnet.com> Date: Mon, 9 Nov 1998 23:43:34 GMT Lines: 26 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news-feed.inet.tele.dk!bofh.vszbr.cz!howland.erols.net!dca1-hub1.news.digex.net!digex!tor-nx1.netcom.ca!gts!bdb In article <727nqd$qrs$3@news.islandnet.com>, Deryk Barker wrote: |Arthur Evans Jr (eva_remove_this_ns@evans.pgh.pa.us) wrote: |: Barry Margolin wrote: |: >For that matter, how many portable OSes have been implemented by *anyone*? |: |: A British OS written in BCPL was quite successfully ported to |: sevreral platforrms, probably in the late '70s or early '80s. | |I believe this would have been Martin Richards' TRIPOS. Richards was |the author of BCPL (and CPL before it) and I believe TRIPOS was the |foundation for the Atari (or was it Amiga?) OS. Can anyone confirm |this? For part of the Amiga system. It was grafted in at the last minute because of some time slippage in the production of the original Amiga I/O code... too bad, since they had to rip it all out again later... -- ,u, Bruce Becker Toronto, Ontario 1 416 699 1868 a \i\ Internet: bdb@gts.org Uucp: ...!gts!bdb `/o/-e "We congratulate you on your free elections and your _\ >_ successful dengue-abatement campaign!" - W. Gibson ###### Newsgroups: alt.folklore.computers,alt.os.multics From: eric@fudge.uchicago.edu (Eric Fischer) Subject: Re: Multics X-Nntp-Posting-Host: fudge.uchicago.edu Message-ID: Sender: news@midway.uchicago.edu (News Administrator) X-Newsposter: Pnews 4.0-test51 (15 Jan 97) Organization: The University of Chicago References: <727n8j$qrs$1@news.islandnet.com> <727pkh$11j6$1@flood.weeg.uiowa.edu> Date: Tue, 10 Nov 1998 03:00:40 GMT Lines: 24 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!logbridge.uoregon.edu!enews.sgi.com!uwvax!uchinews!not-for-mail Douglas W. Jones,201H MLH,3193350740,3193382879 wrote: > > And let us not forget, either, that the first version of Unix was > > written in PDP-11 assembler. > > The first version was written in assembly language for an 18-bit > PDP-9 (ancestor of the PDP-15, and descendant of the PDP-7) I think you've got the details a little wrong. According to Dennis Ritchie's paper, "The Evolution of the Unix Time-sharing System," it was a PDP-7. > > Unless I am mistaken the rewrite in C (1973ish?) was in order to > > facilitate the port to the Unidata machine. > > PDP-11, if memory serves me right. I believe the port to > an Interdata machine was later. The first PDP-11 version was also in assembler. Then came a C version, but only for the PDP-11, and then, still later, the Interdata port. (Again according to a paper by Dennis Ritchie, this time from "The Development of the C Programming Language.") eric ###### From: Dennis Ritchie Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Tue, 10 Nov 1998 03:58:53 +0000 Organization: Bell Labs, Lucent Technologies Lines: 46 Message-ID: <3647B9FD.4516@bell-labs.com> References: <71huie$meo$2@irk.zetnet.co.uk> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> <727nqd$qrs$3@news.islandnet.com> Reply-To: dmr@bell-labs.com NNTP-Posting-Host: cebu.cs.bell-labs.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 3.01 (Win95; U) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news.idt.net!netnews.com!dca1-hub1.news.digex.net!digex!cyclone.i1.net!uunet!in4.uu.net!nntphub.cb.lucent.com!news Deryk Barker wrote (there's subquoting here and this is a remark on several articles in this thread): > > Arthur Evans Jr (eva_remove_this_ns@evans.pgh.pa.us) wrote: > : Barry Margolin wrote: > : >For that matter, how many portable OSes have been implemented by *anyone*? > : > : A British OS written in BCPL was quite successfully ported to > : sevreral platforrms, probably in the late '70s or early '80s. > > I believe this would have been Martin Richards' TRIPOS. Richards was > the author of BCPL (and CPL before it) and I believe TRIPOS was the > foundation for the Atari (or was it Amiga?) OS. Can anyone confirm > this? The paper about the portability of Unix (BLTJ 57, Jul-Aug 1978) by Steve Johnson and me had references to the following in regard to OS portability: Morris, Frank, Thacker, "Machine-independent operating systems", Information Processing 77, (1977) Two papers by Stoy and Strachey on OS6, in Comp. J. 1972. Thalman and Levrat, "SPIP, An way of writing portable operating systems," ACM Computing Symposium, 1977. L.S. Melen, "Thoth, A portable real-time executive", MS thesis, U. Waterloo, 1976. [there was considerable work on Thoth and its own offspring following this, usually connected with Dave Cheriton.] It's my impression that the TRIPOS work was just a bit later. OS6 is probably more directly relevant. I was about to give the URL for the BLTJ paper, then realized that I hadn't yet added it to the home page. I'll do this. Searching around the old files, I discovered the memo that Johnson and I wrote proposing this. Incidentally, Unix was first written for the PDP-7, not -9 in assembly language, rewritten (in assembler) for the PDP-11, rewritten in C (1973) for the PDP-11, then revamped for portability somewhat later. Dennis ###### From: dpeschel@u.washington.edu (D. Peschel) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 10 Nov 1998 07:37:44 GMT Organization: University of Washington, Seattle Lines: 21 Message-ID: <728qg8$ods$1@nntp1.u.washington.edu> References: <71huie$meo$2@irk.zetnet.co.uk> <727nqd$qrs$3@news.islandnet.com> <3647B9FD.4516@bell-labs.com> NNTP-Posting-Host: saul9.u.washington.edu X-Trace: nntp1.u.washington.edu 910683464 25020 (None) 140.142.17.39 X-Complaints-To: help@cac.washington.edu NNTP-Posting-User: dpeschel Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!howland.erols.net!newsfeed.direct.ca!news.u.washington.edu!dpeschel In article <3647B9FD.4516@bell-labs.com>, Dennis Ritchie wrote: >The paper about the portability of Unix (BLTJ 57, Jul-Aug 1978) >by Steve Johnson and me had references to the following in regard >to OS portability: [...] >Two papers by Stoy and Strachey on OS6, in Comp. J. 1972. Now this is a slightly wierd connection. I was just reading about the OS for the Xerox Alto, which was based somewhat on OS6, and tried to look like UNIX. (Though internally it was different: a simpler shell, a very redundant file system, not much memory management, and a wierd "world-swap" or "junta" idea that was very powerful but also quite ad-hoc. The OS also depended on lots of microcode.) I'm not really saying that Unix is related to OS6, but that maybe there's a primal state that all good OS's evolve from? -- Derek ###### Newsgroups: alt.folklore.computers,alt.os.multics From: Kit Powell Subject: Re: Multics X-Nntp-Posting-Host: bass.cse.bris.ac.uk Content-Type: text/plain; charset=us-ascii Message-ID: <3647FDCC.7E07A98F@bristol.ac.uk> Sender: usenet@fsa.bris.ac.uk (Usenet) Content-Transfer-Encoding: 7bit Organization: University of Bristol, England References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> <723b0e$88j@newsops.execpc.com> <727nnh$qrs$2@news.islandnet.com> Mime-Version: 1.0 Date: Tue, 10 Nov 1998 08:48:12 GMT X-Mailer: Mozilla 4.04 [en] (X11; I; IRIX 6.2 IP22) Lines: 27 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!nntp.news.xara.net!xara.net!server5.netnews.ja.net!bris.ac.uk!usenet Deryk Barker wrote: ... > > Now, for those who don't know this, the LEO was the world's first-ever > commercially-oriented machine (1951). Even more amazingly, the Lyons > Electronic Office was designed and built by the J Lyons company, > best-known as manufacturers of cakes and for their nationwide chain of > corner tea shops. > > Anyway, an "orange Leo" was an ICL 2900 mainframe (they came in orange > cabinets), emulating an ICL 1900 mainframe, emulating a GEC System 4 > mainframe emulating a LEO. > Steady on Deryk: that'll be English Electric System 4 please. English Electric merged with ICT in 1968 to form ICL. At a previous point the company had, splendidly, been called English Electric Leo Marconi. One of my uncles was in REME and joined Lyons after the war. He worked on Leo and his employer was taken over three times (I think he had left by the time Fujitsu grabbed ICL). -- Kit Powell, Networking and Digital Communications Manager Computing Service, University of Bristol, Bristol BS8 1UD, GB c.j.powell@bristol.ac.uk +44 117 9287864 (fax +44 117 9291576) ###### From: dpeschel@u.washington.edu (D. Peschel) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 10 Nov 1998 10:07:07 GMT Organization: University of Washington, Seattle Lines: 24 Message-ID: <72938b$toq$1@nntp1.u.washington.edu> References: <71huie$meo$2@irk.zetnet.co.uk> <727nqd$qrs$3@news.islandnet.com> <3647B9FD.4516@bell-labs.com> NNTP-Posting-Host: saul7.u.washington.edu X-Trace: nntp1.u.washington.edu 910692427 30490 (None) 140.142.17.38 X-Complaints-To: help@cac.washington.edu NNTP-Posting-User: dpeschel Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!howland.erols.net!newsfeed.cwix.com!192.220.250.21!netnews1.nw.verio.net!netnews.nwnet.net!news.u.washington.edu!dpeschel [My server complained about alt.os.multics, so I removed that group.] In article <3647B9FD.4516@bell-labs.com>, Dennis Ritchie wrote: >The paper about the portability of Unix (BLTJ 57, Jul-Aug 1978) >by Steve Johnson and me had references to the following in regard >to OS portability: [...] >Two papers by Stoy and Strachey on OS6, in Comp. J. 1972. Now this is a slightly wierd connection. I was just reading about the OS for the Xerox Alto, which was based somewhat on OS6, and tried to look like UNIX. (Though internally it was different: a simpler shell, a very redundant file system, not much memory management, and a wierd "world-swap" or "junta" idea that was very powerful but also quite ad-hoc. The OS also depended on lots of microcode.) I'm not really saying that Unix is related to the Alto OS, but that maybe there's a primal origin (or design) that all good OS's evolve from? As for the bad OS's. how *they* happen is a major mystery of computer science. :) -- Derek ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Tue, 10 Nov 98 12:14:33 GMT Organization: UltraNet Communications, Inc. Lines: 38 Message-ID: <729bpn$kmg$4@ligarius.ultra.net> References: <71huie$meo$2@irk.zetnet.co.uk> <727nqd$qrs$3@news.islandnet.com> <3647B9FD.4516@bell-labs.com> <72938b$toq$1@nntp1.u.washington.edu> NNTP-Posting-Host: d7.dial-16.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 10 Nov 1998 12:32:55 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!212.63.192.161.MISMATCH!newshub.bart.net!news.tele2.nl!newsfeed1.swip.net!swipnet!newspump.monmouth.com!newspeer.monmouth.com!news.ultranet.com!d7 In article <72938b$toq$1@nntp1.u.washington.edu>, dpeschel@u.washington.edu (D. Peschel) wrote: >In article <3647B9FD.4516@bell-labs.com>, >Dennis Ritchie wrote: > >>The paper about the portability of Unix (BLTJ 57, Jul-Aug 1978) >>by Steve Johnson and me had references to the following in regard >>to OS portability: >[...] >>Two papers by Stoy and Strachey on OS6, in Comp. J. 1972. > >Now this is a slightly wierd connection. I was just reading >about the OS for the Xerox Alto, which was based somewhat on >OS6, and tried to look like UNIX. > >(Though internally it was different: a simpler shell, a very >redundant file system, not much memory management, and a wierd >"world-swap" or "junta" idea that was very powerful but also >quite ad-hoc. The OS also depended on lots of microcode.) > >I'm not really saying that Unix is related to the Alto OS, >but that maybe there's a primal origin (or design) that all >good OS's evolve from? You can only do so much with hardware. If one is writing a timesharing OS, the design is limited by certain things like maximum throughput, anti-thrashing, etc. >As for the bad OS's. how *they* happen >is a major mystery of computer science. :) No, it's not. :-) /BAH Subtract a hundred and four for e-mail. ###### From: "Samael" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Tue, 10 Nov 1998 13:08:52 -0000 Lines: 11 Message-ID: <36483ae4.0@127.0.0.1> References: <71huie$meo$2@irk.zetnet.co.uk> <727nqd$qrs$3@news.islandnet.com> <3647B9FD.4516@bell-labs.com> <72938b$toq$1@nntp1.u.washington.edu> NNTP-Posting-Host: 195.147.15.86 X-Newsreader: Microsoft Outlook Express 4.72.3155.0 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3155.0 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!diablo.theplanet.net!btnet-peer!btnet!nntp.news.xara.net!xara.net!news.itg.net.uk!127.0.0.1!192.168.0.1 D. Peschel wrote in message <72938b$toq$1@nntp1.u.washington.edu>... >I'm not really saying that Unix is related to the Alto OS, but that maybe >there's a primal origin (or design) that all good OS's evolve from? As for >the bad OS's. how *they* happen is a major mystery of computer science. :) It's more of a mystery of marketing. Samael ###### From: dpeschel@u.washington.edu (D. Peschel) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 10 Nov 1998 13:18:45 GMT Organization: University of Washington, Seattle Lines: 60 Message-ID: <729efl$7h4$1@nntp1.u.washington.edu> References: <71huie$meo$2@irk.zetnet.co.uk> <3647B9FD.4516@bell-labs.com> <72938b$toq$1@nntp1.u.washington.edu> <729bpn$kmg$4@ligarius.ultra.net> NNTP-Posting-Host: saul9.u.washington.edu X-Trace: nntp1.u.washington.edu 910703925 7716 (None) 140.142.17.38 X-Complaints-To: help@cac.washington.edu NNTP-Posting-User: dpeschel Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!howland.erols.net!newsfeed.direct.ca!news.u.washington.edu!dpeschel In article <729bpn$kmg$4@ligarius.ultra.net>, wrote: >In article <72938b$toq$1@nntp1.u.washington.edu>, > dpeschel@u.washington.edu (D. Peschel) wrote: >>I'm not really saying that Unix is related to the Alto OS, >>but that maybe there's a primal origin (or design) that all >>good OS's evolve from? > >You can only do so much with hardware. If one is writing >a timesharing OS, the design is limited by certain things >like maximum throughput, anti-thrashing, etc. That's true but it doesn't really explain the structure of the software. I think it's interesting that powerful, well-designed OS's tend to be made a certain way of certain parts. By your criterion, I could show you a very efficient but internally ugly, unexpandable, clunky OS. Lynn Wheeler posts here sometimes; he talks about efficiency concerns exactly as you describe because they are/were a major part of his successes. I would not say that the OS's he usually discusses (early IBM mainframe OS's, starting with OS/360 and ending at early CMS) have a very pretty structure. For the user interface especially, they tend to use ad-hoc solutions. I don't think it's easy to tweak those parts of the OS. I'm talking about the OS as an abstraction built for the user to use and derive power from, and I'm noticing that the clean abstractions tend to share certain common traits. (For example, they have a few important elements which they then combine to the best possible effect.) The Alto OS is hardly a timesharing OS! The Alto was designed *not* to be a timeshared computer. (The designers were happy it wasn't faster at night.) Many of the elements of a "real" OS are missing, but the ones that are there are well-thought out. >>As for the bad OS's. how *they* happen >>is a major mystery of computer science. :) > >No, it's not. :-) Of course you're right. But I'm serious. Again, I'm talking about features (not marketing). And it's not just Micro- soft. Programmers have an amazing tendency to come up with far worse designs than anyone can imagine... haven't you ever gotten that feeling, looking at a piece of software and thinking, "How could they *possibly* have thought this was a good idea?"? I need to find a copy of the chart that shows how to change between the many modes of the SOS editor. I've seen it but don't currently have a copy. I've never used SOS (but if I ever do I'm going to put a LARGE copy of the chart by the computer first -- I may have a sacrifice first for good measure). And it must have taken *ahem* talent to come up with something that contorted. It would really have been easier to design a good program instead. At least you guys redeemed yourselves with EDT. :) Now if you want to talk about why bad OS's *sell*, that's not a mystery. (Though it is a tragedy.) -- Derek ###### From: jones@cs.uiowa.edu (Douglas W. Jones,201H MLH,3193350740,3193382879) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 10 Nov 1998 14:51:10 GMT Organization: The University of Iowa Lines: 21 Message-ID: <729jsu$abg$1@flood.weeg.uiowa.edu> References: <72938b$toq$1@nntp1.u.washington.edu> NNTP-Posting-Host: pyrite.cs.uiowa.edu Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news-feed.fnsi.net!nntp.msen.com!news.altair.com!NewsNG.Chicago.Qual.Net!news.uiowa.edu!not-for-mail In article <3647B9FD.4516@bell-labs.com>, Dennis Ritchie wrote: >The paper about the portability of Unix (BLTJ 57, Jul-Aug 1978) >by Steve Johnson and me had references to the following in regard >to OS portability: ... >Two papers by Stoy and Strachey on OS6, in Comp. J. 1972. The work by Stoy and Strachey on OS6 is some of the most overlooked but seminal stuff I've ever come across! They describe, in fairly complete detail, how to implement polymorphic objects. Their goal is device independent I/O, and they solve the problem in BCPL. If you use modern terminology to describe what they did, they created the abstract class file, and then they created, for each device, a concrete subclass that implemented the abstract interface. Aside from somewhat clumsy BCPL syntax, the code they wrote is exactly the code you'd expect from a clean object oriented solution to this problem in a decent object oriented language! Doug Jones jones@cs.uiowa.edu ###### From: alex*@*rockvax.rockefeller.edu (Alexandre Pechtchanski) Newsgroups: alt.folklore.computers Subject: Re: Multics Organization: Rockefeller University Hospital (GCRC), New York Message-ID: <36487903.97497300@Rockyd> References: <72938b$toq$1@nntp1.u.washington.edu> <729jsu$abg$1@flood.weeg.uiowa.edu> X-Newsreader: Forte Agent 1.5/32.451 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 27 Date: Tue, 10 Nov 1998 17:35:26 GMT NNTP-Posting-Host: 129.85.24.56 X-Trace: rockyd.rockefeller.edu 910719397 129.85.24.56 (Tue, 10 Nov 1998 12:36:37 EDT) NNTP-Posting-Date: Tue, 10 Nov 1998 12:36:37 EDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news-feed.inet.tele.dk!bofh.vszbr.cz!howland.erols.net!newsfeed.cwix.com!209.150.97.11!feeder.qis.net!newsfeed.nyu.edu!rockyd.rockefeller.edu!not-for-mail On 10 Nov 1998 14:51:10 GMT, jones@cs.uiowa.edu (Douglas W. Jones,201H MLH,3193350740,3193382879) wrote: >In article <3647B9FD.4516@bell-labs.com>, >Dennis Ritchie wrote: > >>The paper about the portability of Unix (BLTJ 57, Jul-Aug 1978) >>by Steve Johnson and me had references to the following in regard >>to OS portability: > ... >>Two papers by Stoy and Strachey on OS6, in Comp. J. 1972. > >The work by Stoy and Strachey on OS6 is some of the most overlooked but >seminal stuff I've ever come across! They describe, in fairly complete >detail, how to implement polymorphic objects. Their goal is device >independent I/O, and they solve the problem in BCPL. If you use modern >terminology to describe what they did, they created the abstract class >file, and then they created, for each device, a concrete subclass that >implemented the abstract interface. Aside from somewhat clumsy BCPL >syntax, the code they wrote is exactly the code you'd expect from a clean >object oriented solution to this problem in a decent object oriented >language! Which proves again what the whipper-snappers tend to ignore: one doesn't need object-oriented language to write object-oriented code! [ When replying, remove *'s from address ] Alexandre Pechtchanski, Systems Manager, RUH, NY ###### From: jeff@jakfield.xu-netx.com (Jeff York) Newsgroups: alt.folklore.computers Subject: Re: Multics Reply-To: jeff@jakfield.xu-netx.com Message-ID: <364879d0.20064871@news.u-net.com> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> X-Newsreader: Forte Agent 1.5/32.452 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Tue, 10 Nov 1998 18:23:29 GMT NNTP-Posting-Host: 195.102.196.240 X-Complaints-To: news@u-net.net X-Trace: newsr2.u-net.net 910722488 195.102.196.240 (Tue, 10 Nov 1998 18:28:08 BST) NNTP-Posting-Date: Tue, 10 Nov 1998 18:28:08 BST Organization: (Posted via) U-NET Internet Ltd. Lines: 16 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!dispose.news.demon.net!demon!ayres.ftech.net!news.ftech.net!peer.news.zetnet.net!peer.news.bb.u-net.net!u-net!newsr2.u-net.net.POSTED!not-for-mail jadams@NO.SPAM.PLEASE (James W. Adams) wrote: >G(E)COS= General (Electric) Comprehensive Operation Supervisor, IIRC. >I can't recall whether it was "operation" or "operating" but it >was definitely "supervisor." In my Honeywell days, GCOS was generally understood to stand for "God's Chosen Operating System".. :-) -- Jeff. Ironbridge, Shrops, U.K. jeff@jakfield.xu-netx.com (remove the x..x round u-net for return address) ... Thimk ###### From: westin*nospam@graphics.cornell.edu (Stephen H. Westin) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 10 Nov 1998 09:09:13 -0500 Organization: Program of Computer Graphics -- Cornell University Lines: 42 Sender: westin@blynken.graphics.cornell.edu Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> NNTP-Posting-Host: blynken.graphics.cornell.edu X-Newsreader: Gnus v5.6.44/Emacs 20.2 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news.syr.edu!newsstand.cit.cornell.edu!news.graphics.cornell.edu!news karger@watson.ibm.com (Paul A. Karger) writes: > On the question of Multics portability: > > At the time Multics was designed, the idea of writing an operating > system in a higher level language was very radical, and making it > portable was totally unheard of. Furthermore, there were no other > processors at the time that could have run Multics. The 360/67 was > the only other CPU with paging and segmentation, and the 360 had > been explicitly rejected for Multics. Multics also was pushing the > state of the art in so many ways, that to add yet another radical > requirement - portability - would probably have doomed it to total > failure. As it was, it took a LOT longer to get it working than any > of the original designers had expected. I think the notion of developing hardware and the OS together, so that each exploited the other, was seen as *positive* in those days. Early hardware designers were just happy to get anything running, and software efficiency wasn't seen as a goal. Shoot, *software* wasn't seen as a goal. Beginning with, say, the Burroughs B5000 and continuing through System/360, Multics, and others, the strong notion was that the machine, OS, compilers, etc. would all be designed and developed together. Certainly Multics would have been very different had it been implemented on out-of-the-box hardware. I would say that the last system developed in this way was probably VAX/VMS, or perhaps the Lisp Machine. After the VAX, the target became building the most cost-effective Unix box. Which sort of automatically made everyone else inefficient, as they had to run (if at all) on hardware designed for Unix. Just as everyone decided that C was the world's most efficient compiled language, as it was run on machines designed to optimize it. So, for better or worse, we have C and Unix, except as Microsoft replaces it with something yet cruder. -- -Stephen H. Westin Any information or opinions in this message are mine: they do not represent the position of Cornell University or any of its sponsors. ###### Message-ID: <3648BCB4.F483E01B@shore.net> From: John W Gintell Reply-To: gintell@shore.net Organization: JWG Software Systems X-Mailer: Mozilla 4.5 [en] (WinNT; U) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics References: <71huie$meo$2@irk.zetnet.co.uk> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 32 Date: Tue, 10 Nov 1998 17:22:44 -0500 NNTP-Posting-Host: 204.167.102.156 X-Complaints-To: abuse@shore.net X-Trace: news.shore.net 910736477 204.167.102.156 (Tue, 10 Nov 1998 17:21:17 EDT) NNTP-Posting-Date: Tue, 10 Nov 1998 17:21:17 EDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!newsfeed.wirehub.nl!howland.erols.net!cpk-news-hub1.bbnplanet.com!news.gtei.net!feed1.news.rcn.net!rcn!newsfeed.xcom.net!news.shore.net!not-for-mail "Stephen H. Westin" wrote: > > karger@watson.ibm.com (Paul A. Karger) writes: > > > On the question of Multics portability: .......... > I think the notion of developing hardware and the OS together, so that > each exploited the other, was seen as *positive* in those days. Early > hardware designers were just happy to get anything running, and > software efficiency wasn't seen as a goal. Shoot, *software* wasn't > seen as a goal. Beginning with, say, the Burroughs B5000 and > continuing through System/360, Multics, and others, the strong notion > was that the machine, OS, compilers, etc. would all be designed and > developed together. Certainly Multics would have been very different > had it been implemented on out-of-the-box hardware. I would say that > the last system developed in this way was probably VAX/VMS, or perhaps > the Lisp Machine. ....... > After the VAX, the target became building the most cost-effective Unix > box. Which sort of automatically made everyone else inefficient, as > they had to run (if at all) on hardware designed for Unix. Just as > everyone decided that C was the world's most efficient compiled > language, as it was run on machines designed to optimize it. > Well, the Macintosh isn't exactly new, nor is it exactly successful (although Apple did $5.5 billion last year and made $313M, and there are a lot of running Macs out there). This machine is a case of hardware and software designed together. Remember there is a lot more about hardware than the CPU instruction set. This is a contributing reason to its technical success and its much weaker marketing success. Apple has retained this relationship for technical and profitability reasons. ###### From: ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 10 Nov 1998 18:38:34 GMT Organization: The National Capital FreeNet Lines: 8 Message-ID: <72a17a$gfj@freenet-news.carleton.ca> References: <72938b$toq$1@nntp1.u.washington.edu> <729jsu$abg$1@flood.weeg.uiowa.edu> <36487903.97497300@Rockyd> Reply-To: ab528@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) NNTP-Posting-Host: freenet5.carleton.ca X-Given-Sender: ab528@freenet5.carleton.ca (Heinz W. Wiggeshoff) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news-feed.inet.tele.dk!bofh.vszbr.cz!newshub.northeast.verio.net!news.new-york.net!nntp.frontiernet.net!node17.frontiernet.net!xcski.com!freenet-news.carleton.ca!FreeNet.Carleton.CA!ab528 Alexandre Pechtchanski (alex*@*rockvax.rockefeller.edu) writes: > > Which proves again what the whipper-snappers tend to ignore: one doesn't need > object-oriented language to write object-oriented code! As demonstrated by the Borland C++ for OS/2 _assembler_ !!! ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Tue, 10 Nov 1998 19:14:25 GMT Organization: . Lines: 35 Message-ID: <36487bd7.1065201@news.uunet.be> References: <71huie$meo$2@irk.zetnet.co.uk> <36449E63.65F3@ix.netcom.com> <722d8h$qka@srvr1.engin.umich.edu> <36471490.3EAC70F5@pobox.srl.ford.com> NNTP-Posting-Host: pool02b-194-7-145-55.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news-feed.inet.tele.dk!bofh.vszbr.cz!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach Vince Scarafino on Mon, 09 Nov 1998 11:13:04 -0500 to alt.folklore.computers: > There is much evidence in this. Microsoft is "reinventing" many things, > and its implementation shows that they didn't bother to do any research > to help understand what they are doing. I kinda agree here, maybe not for 100%, but sorta. ;-) I (personally) think it's more likely a lack of experienced coders (straight from school they cost less, and I deliberately didn't say 'programmers'), than a lack of research. If they attracted more people that are experienced in other systems, I even think they wouldn't need a lot of research. > The most blatant area that I > see is the way they manage the namespace for dynamically linked > objects. In reality, they don't bother to manage it! If you want to > replace a dynamic link you have to reboot. I can find no concept of > terminate reference. But here, I don't understand what you mean exactly. Unload (terminate) all programs that depend on a dll (or ocx), replace it, restart programs: that's what I do several times a day when developing them, but that's not what I call rebooting. If you're referring to dynamically loaded device drivers, I think it's exactly the same thing - though I'm not fully sure since I don't write them, and seldom have to replace them (for which I rely on installation programs). If it's a dynamic 'thing' the system (or the shell) depends on, unloading dependent programs means shutting down - if that is what you meant, you're right. ###### From: Barry Margolin Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics References: <71tln8$qmm$1@mdnews.btv.ibm.com> <723b0e$88j@newsops.execpc.com> Organization: GTE Internetworking, Cambridge, MA X-Newsreader: trn 4.0-test60 (5 October 1997) Originator: barmar@bbnplanet.com (Barry Margolin) Lines: 18 Message-ID: Date: Tue, 10 Nov 1998 19:27:50 GMT NNTP-Posting-Host: 4.2.32.97 X-Trace: burlma1-snr1.gtei.net 910726070 4.2.32.97 (Tue, 10 Nov 1998 19:27:50 GMT) NNTP-Posting-Date: Tue, 10 Nov 1998 19:27:50 GMT Distribution: world Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!cpk-news-hub1.bbnplanet.com!news.gtei.net!burlma1-snr1.gtei.net!not-for-mail In article <723b0e$88j@newsops.execpc.com>, Peter H. Coffin wrote: >The most current OS/400 (v4r3) will still run programs compiled on the >System 38 tranparently. That's 10-15 year old object code over at least >one complete architecture change. The machine also has a history of >being able to run in System 36 mode, over that same architecture change. > >I'm not sure if that counts as a ported OS or not, and I do work for >IBM, so I'll stop now... No, emulation doesn't count. Emulators are used specifically to allow a new machine to run *nonportable* programs from an old machine. -- Barry Margolin, barmar@bbnplanet.com GTE Internetworking, Powered by BBN, Burlington, MA *** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups. Don't bother cc'ing followups to me. ###### From: jeff@jakfield.xu-netx.com (Jeff York) Newsgroups: alt.folklore.computers Subject: Re: Multics Reply-To: jeff@jakfield.xu-netx.com Message-ID: <364a968a.27419126@news.u-net.com> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> X-Newsreader: Forte Agent 1.5/32.452 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Tue, 10 Nov 1998 19:40:12 GMT NNTP-Posting-Host: 195.102.196.223 X-Complaints-To: news@u-net.net X-Trace: newsr2.u-net.net 910727090 195.102.196.223 (Tue, 10 Nov 1998 19:44:50 BST) NNTP-Posting-Date: Tue, 10 Nov 1998 19:44:50 BST Organization: (Posted via) U-NET Internet Ltd. Lines: 16 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!ayres.ftech.net!news.ftech.net!dispose.news.demon.net!demon!peer.news.zetnet.net!peer.news.bb.u-net.net!u-net!newsr2.u-net.net.POSTED!not-for-mail jadams@NO.SPAM.PLEASE (James W. Adams) wrote: >G(E)COS= General (Electric) Comprehensive Operation Supervisor, IIRC. >I can't recall whether it was "operation" or "operating" but it >was definitely "supervisor." In my Honeywell days, GCOS was generally understood to stand for "God's Chosen Operating System".. :-) -- Jeff. Ironbridge, Shrops, U.K. jeff@jakfield.xu-netx.com (remove the x..x round u-net for return address) ... Thimk ###### From: Dennis Ritchie Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Wed, 11 Nov 1998 05:40:20 +0000 Organization: Bell Labs, Lucent Technologies Lines: 32 Message-ID: <36492344.644@bell-labs.com> References: <71huie$meo$2@irk.zetnet.co.uk> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> <727nqd$qrs$3@news.islandnet.com> <3647B9FD.4516@bell-labs.com> Reply-To: dmr@bell-labs.com NNTP-Posting-Host: cebu.cs.bell-labs.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 3.01 (Win95; U) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!newspump.monmouth.com!newspeer.monmouth.com!news.maxwell.syr.edu!wn4feed!worldnet.att.net!207.24.196.41!nntphub.cb.lucent.com!news I couldn't track down the online copy of Johnson's and my BSTJ paper about Unix portability, but I did find the 1976 memo proposing the project: http://www.cs.bell-labs.com/~dmr/firstport.html The obvious backup from this, the main homepage, links to more historical stuff. Incidentally, I don't understand some of the claims in subsequent articles about influence (e.g. Unix on the Alto). There certainly was some common ancestry (even via BCPL) and a desire for interactive computing. Beyond that? In the other direction, certainly, from the early 80s and onward. Also, although there may be a case to be made for C or C-like languages influencing (parts of) CPU design, I don't really see the one for Unix having much influence on broader aspects of hardware architecture (for better or worse). Over the last few decades there has clearly been a tension between the well-engineered, coherent, somewhat closed single solution product (the Mac may be the best example) and the one put together from commodity parts of various kinds--examples range from PC hardware to the Unix software-tool approach. The difficulty is finding the current sweet spot and get something that partakes of both. It's fairly clear who's figured this out over the last few years. Dennis ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Wed, 11 Nov 98 12:10:30 GMT Organization: UltraNet Communications, Inc. Lines: 55 Message-ID: <72bvue$k4b$1@strato.ultra.net> References: <71huie$meo$2@irk.zetnet.co.uk> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> NNTP-Posting-Host: d10.dial-15.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 11 Nov 1998 12:29:02 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!212.63.192.161.MISMATCH!newshub.bart.net!news.tele2.nl!newsfeed1.swip.net!swipnet!newspump.monmouth.com!newspeer.monmouth.com!news.ultranet.com!d10 [1 newsgroup snipped] In article , westin*nospam@graphics.cornell.edu (Stephen H. Westin) wrote: >karger@watson.ibm.com (Paul A. Karger) writes: > >> On the question of Multics portability: >> >> At the time Multics was designed, the idea of writing an operating >> system in a higher level language was very radical, and making it >> portable was totally unheard of. Furthermore, there were no other >> processors at the time that could have run Multics. The 360/67 was >> the only other CPU with paging and segmentation, and the 360 had >> been explicitly rejected for Multics. Multics also was pushing the >> state of the art in so many ways, that to add yet another radical >> requirement - portability - would probably have doomed it to total >> failure. As it was, it took a LOT longer to get it working than any >> of the original designers had expected. > >I think the notion of developing hardware and the OS together, so that >each exploited the other, was seen as *positive* in those days. It was called revenue. Hardware just wasn't very interesting without software (something that people just didn't seem to understand then or now). > Early >hardware designers were just happy to get anything running, and >software efficiency wasn't seen as a goal. And that is bullshit. > Shoot, *software* wasn't >seen as a goal. Beginning with, say, the Burroughs B5000 and >continuing through System/360, Multics, and others, the strong notion >was that the machine, OS, compilers, etc. would all be designed and >developed together. This certainly isn't how we (TOPS10 timesharing OS running on PDP-10) worked. > Certainly Multics would have been very different >had it been implemented on out-of-the-box hardware. I would say that >the last system developed in this way was probably VAX/VMS, or perhaps >the Lisp Machine. Have people really forgotten how software was developed? There wasn't any such thing as out-of-the-box hardware when OSs were developed. (The closest might have been Dennis' effort.) /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Wed, 11 Nov 98 12:13:32 GMT Organization: UltraNet Communications, Inc. Lines: 25 Message-ID: <72c043$k4b$2@strato.ultra.net> References: <71huie$meo$2@irk.zetnet.co.uk> <36449E63.65F3@ix.netcom.com> <722d8h$qka@srvr1.engin.umich.edu> <36471490.3EAC70F5@pobox.srl.ford.com> <36487bd7.1065201@news.uunet.be> NNTP-Posting-Host: d10.dial-15.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 11 Nov 1998 12:32:03 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!feed1.news.luth.se!luth.se!cam-news-hub1.bbnplanet.com!news.gtei.net!newsfeed.xcom.net!news.ultranet.com!d10 In article <36487bd7.1065201@news.uunet.be>, lucvdv@null.net (Luc Van der Veken) wrote: >Also sprach Vince Scarafino on Mon, >09 Nov 1998 11:13:04 -0500 to alt.folklore.computers: > >> There is much evidence in this. Microsoft is "reinventing" many things, >> and its implementation shows that they didn't bother to do any research >> to help understand what they are doing. > >I kinda agree here, maybe not for 100%, but sorta. ;-) >I (personally) think it's more likely a lack of experienced >coders (straight from school they cost less, and I deliberately >didn't say 'programmers'), than a lack of research. > >If they attracted more people that are experienced in other >systems, I even think they wouldn't need a lot of research. Ahem...if? Why do you think things are getting reinvented. The ideas are getting carried in by people who had already done it before. /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Wed, 11 Nov 98 12:26:40 GMT Organization: UltraNet Communications, Inc. Lines: 105 Message-ID: <72c0sp$k4b$3@strato.ultra.net> References: <71huie$meo$2@irk.zetnet.co.uk> <3647B9FD.4516@bell-labs.com> <72938b$toq$1@nntp1.u.washington.edu> <729bpn$kmg$4@ligarius.ultra.net> <729efl$7h4$1@nntp1.u.washington.edu> NNTP-Posting-Host: d10.dial-15.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 11 Nov 1998 12:45:13 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!194.165.93.117.MISMATCH!newshub.bart.net!newsfeed.wirehub.nl!howland.erols.net!feed1.news.rcn.net!rcn!news.ultranet.com!d10 In article <729efl$7h4$1@nntp1.u.washington.edu>, dpeschel@u.washington.edu (D. Peschel) wrote: >In article <729bpn$kmg$4@ligarius.ultra.net>, wrote: >>In article <72938b$toq$1@nntp1.u.washington.edu>, >> dpeschel@u.washington.edu (D. Peschel) wrote: > >>>I'm not really saying that Unix is related to the Alto OS, >>>but that maybe there's a primal origin (or design) that all >>>good OS's evolve from? >> >>You can only do so much with hardware. If one is writing >>a timesharing OS, the design is limited by certain things >>like maximum throughput, anti-thrashing, etc. > >That's true but it doesn't really explain the structure of the software. >I think it's interesting that powerful, well-designed OS's tend to be made >a certain way of certain parts. By your criterion, I could show you a very >efficient but internally ugly, unexpandable, clunky OS. Lynn Wheeler posts >here sometimes; he talks about efficiency concerns exactly as you describe >because they are/were a major part of his successes. Lynn Wheeler is very experienced in certain areas but not all and most definitely not in my area. > >I would not say that the OS's he usually discusses (early IBM mainframe OS's, >starting with OS/360 and ending at early CMS) have a very pretty structure. And they're not timesharing. (Even though IBM thought so.) >For the user interface especially, they tend to use ad-hoc solutions. I don't >think it's easy to tweak those parts of the OS. > >I'm talking about the OS as an abstraction built for the user to use and >derive power from, and I'm noticing that the clean abstractions tend to share >certain common traits. (For example, they have a few important elements >which they then combine to the best possible effect.) > >The Alto OS is hardly a timesharing OS! The Alto was designed *not* to be a >timeshared computer. (The designers were happy it wasn't faster at night.) >Many of the elements of a "real" OS are missing, but the ones that are there >are well-thought out. > >>>As for the bad OS's. how *they* happen >>>is a major mystery of computer science. :) >> >>No, it's not. :-) > >Of course you're right. But I'm serious. Oh, I was being very serious. A programmer's style tells a lot of stories. > >Again, I'm talking about features (not marketing). I haven't even begun to talk about marketing. > And it's not just Microsoft. Programmers have an amazing >tendency to come up with far worse designs than anyone can >imagine... haven't you ever gotten that feeling, looking at >a piece of software and thinking, "How could they *possibly* >have thought this was a good idea?"? Actually, no. The question that usually popped up was "Why did they think this was a good idea?" Once you got that answer you knew the basis (or assumptions) that design had to deal with. To be a skillful bug fixer, one has to fix the code in the same style as it was written otherwise all you get is bandages upon bandages and the bug never really gets fixed. > >I need to find a copy of the chart that shows how to change >between the many modes of the SOS editor. I've seen it but >don't currently have a copy. I've never used SOS (but if I >ever do I'm going to put a LARGE copy of the >chart by the computer first -- I may have a sacrifice first >for good measure). And it must have taken *ahem* talent to >come up with something that contorted. I don't know why you're thread drifting into this subject. I don't remember who wrote SOS. I think it was a college project somewhere. >It would really have been easier to design a good program instead. > >At least you guys redeemed yourselves with EDT. :) You really ought to be sure of your facts before you talk :-). Not all software that was shipped was developed in-house. /BAH Subtract a hundred and four for e-mail. ###### From: Ken Haley Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Wed, 11 Nov 1998 12:52:40 +0000 Organization: Syntegra Lines: 46 Message-ID: <36498898.24D6@syntegra.bt.co.uk> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> <723b0e$88j@newsops.execpc.com> <727nnh$qrs$2@news.islandnet.com> NNTP-Posting-Host: 193.113.75.37 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 2.0 (Win16; I) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!212.63.192.161.MISMATCH!newshub.bart.net!news.tele2.nl!newsfeed1.swip.net!swipnet!news-feed.inet.tele.dk!bofh.vszbr.cz!news-lond.gip.net!news.gsl.net!gip.net!btnet-peer!btnet-feed2!btnet!bt!not-for-mail Deryk Barker wrote: > > [snip] > > BT, at one location, had what they called the "orange Leos". > > Now, for those who don't know this, the LEO was the world's first-ever > commercially-oriented machine (1951). Even more amazingly, the Lyons > Electronic Office was designed and built by the J Lyons company, > best-known as manufacturers of cakes and for their nationwide chain of > corner tea shops. > > Anyway, an "orange Leo" was an ICL 2900 mainframe (they came in orange > cabinets), emulating an ICL 1900 mainframe, emulating a GEC System 4 > mainframe emulating a LEO. > > 30+ year old executable code over 3 architecture changes.... Orange - Never ! it was "Hot Tango" Emulation was at microcode level - all the "S" series 2900 series and the 2960 of the "P" series were "soft" machines in that the microcode was loaded from disk. If you took your nice shiney 2900 and loaded DME1900, the microcode loaded made the machine, including peripherals,appear as a 1900 series machine. If you loaded DME System4, it looked just like a system 4 (DME370 was never released, and initially denied, but apparently the microcode writers got it running as an excercise) DME1900 continued to run long after the last real 1900 hardware was decommissioned, but faced competition with the CME environments - which basically loaded both New Range (2900) microcode and either 1900 or S4 microcode and timesliced the microcode engine which made the nice shiney machine appear as both types of machine (and incidentailly, made interrupt processing interesting) It was primarily a workload movement tool, as you could get a new machine, make it look like the old one (only much faster) then get the new environment working, develop applications and slowly move work across to the new environment. Eventually CME* also appeared, which allowed a VM running under VME, to run a different instruction set (supported by microcode) - there was a 2 bit (?) field in the SSR of the process to select the microcode slot used by the VM I haven't been able to trace a reference to find out whether the Leo environment was actually a "DME Leo" approach at microcode level, or an emulator running under DME System4 ###### From: westin*nospam@graphics.cornell.edu (Stephen H. Westin) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 11 Nov 1998 09:39:01 -0500 Organization: Program of Computer Graphics -- Cornell University Lines: 23 Sender: westin@blynken.graphics.cornell.edu Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> <727nqd$qrs$3@news.islandnet.com> <3647B9FD.4516@bell-labs.com> <36492344.644@bell-labs.com> NNTP-Posting-Host: blynken.graphics.cornell.edu X-Newsreader: Gnus v5.6.44/Emacs 20.2 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news-peer.gip.net!news.gsl.net!gip.net!portc01.blue.aol.com!newsstand.cit.cornell.edu!news.graphics.cornell.edu!news Dennis Ritchie writes: > Also, although there may be a case to be made for C or C-like > languages influencing (parts of) CPU design, I don't really > see the one for Unix having much influence on broader aspects > of hardware architecture (for better or worse). One example is addressing architecture: Unix pretty much presumes a single-level contiguous address space, so modern processors lack support for Multics-style segmentation. This comes back to bite when we try to share data and code between dissimilar processes; though position-independent code is usually possible, there is often a performance penalty, and position-independent data structures can be really awkward. -- -Stephen H. Westin Any information or opinions in this message are mine: they do not represent the position of Cornell University or any of its sponsors. ###### From: Barry Margolin Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics References: <71huie$meo$2@irk.zetnet.co.uk> <727nqd$qrs$3@news.islandnet.com> <3647B9FD.4516@bell-labs.com> <36492344.644@bell-labs.com> Organization: GTE Internetworking, Cambridge, MA X-Newsreader: trn 4.0-test60 (5 October 1997) Originator: barmar@bbnplanet.com (Barry Margolin) Lines: 18 Message-ID: Date: Wed, 11 Nov 1998 17:46:46 GMT NNTP-Posting-Host: 4.2.32.97 X-Trace: burlma1-snr1.gtei.net 910806406 4.2.32.97 (Wed, 11 Nov 1998 17:46:46 GMT) NNTP-Posting-Date: Wed, 11 Nov 1998 17:46:46 GMT Distribution: world Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newscore.univie.ac.at!btnet-peer!btnet!newsfeed.cwix.com!4.1.16.34!cpk-news-hub1.bbnplanet.com!news.gtei.net!burlma1-snr1.gtei.net!not-for-mail In article <36492344.644@bell-labs.com>, Dennis Ritchie wrote: >Also, although there may be a case to be made for C or C-like >languages influencing (parts of) CPU design, I don't really >see the one for Unix having much influence on broader aspects >of hardware architecture (for better or worse). I think the main influence is that C and Unix don't tend to make use of fancy architectural features, so hardware designers have been motivated towards simpler designs. This dovetailed well with the move towards RISC processors, but it's not clear how much direct causal relationship there was. -- Barry Margolin, barmar@bbnplanet.com GTE Internetworking, Powered by BBN, Burlington, MA *** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups. Don't bother cc'ing followups to me. ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Wed, 11 Nov 1998 19:14:49 GMT Organization: . Lines: 28 Message-ID: <3649c9ad.792409@news.uunet.be> References: <71huie$meo$2@irk.zetnet.co.uk> <36449E63.65F3@ix.netcom.com> <722d8h$qka@srvr1.engin.umich.edu> <36471490.3EAC70F5@pobox.srl.ford.com> <36487bd7.1065201@news.uunet.be> <72c043$k4b$2@strato.ultra.net> NNTP-Posting-Host: pool02b-194-7-145-83.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!btnet-peer!btnet!diablo.theplanet.net!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach jmfbahciv@aol.com on Wed, 11 Nov 98 12:13:32 GMT to alt.folklore.computers: > In article <36487bd7.1065201@news.uunet.be>, > lucvdv@null.net (Luc Van der Veken) wrote: > >If they attracted more people that are experienced in other > >systems, I even think they wouldn't need a lot of research. > > Ahem...if? Why do you think things are getting reinvented. The > ideas are getting carried in by people who had already done it > before. The way I see it, a few people on top bring in the ideas (which makes your statement correct), but most of the actual coding is done lower down the tree - by less experienced workers. On itself, this isn't even unexpected. Of course, I did say "I think" somewhere in that other post - I don't work for MS, so I don't know too much about who's working there. It's just that if their coders are made of the same material you get to talk to on the phone (I never called their support, but I had the opportunity to 'feel' out, by talking to some people you get on the phone at MSDN's subscription services, in how much detail they know what they're talking about. Which isn't really unexpected either - their job is to manage subscriptions, not to know every detail of NT server.) ###### From: genew@vip.net (Gene Wirchenko) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Wed, 11 Nov 1998 21:48:22 GMT Organization: Posted via Supernews, http://www.supernews.com - Discussions start here! Lines: 41 Message-ID: <3649dd08.6725378@news.vip.net> References: <71huie$meo$2@irk.zetnet.co.uk> <36449E63.65F3@ix.netcom.com> <722d8h$qka@srvr1.engin.umich.edu> <36471490.3EAC70F5@pobox.srl.ford.com> <36487bd7.1065201@news.uunet.be> <72c043$k4b$2@strato.ultra.net> Reply-To: genew@vip.net NNTP-Posting-Host: 204.209.212.24 X-Trace: 910820807 A01OARAUVD418CCD1C usenet87.supernews.com X-Complaints-To: newsabuse@supernews.com X-Newsreader: Forte Free Agent 1.1/32.230 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsfeed.cwix.com!209.150.97.11!feeder.qis.net!newsfeed-east.supernews.com!supernews.com!Supernews69!not-for-mail jmfbahciv@aol.com wrote: >In article <36487bd7.1065201@news.uunet.be>, > lucvdv@null.net (Luc Van der Veken) wrote: >>Also sprach Vince Scarafino on Mon, >>09 Nov 1998 11:13:04 -0500 to alt.folklore.computers: >> >>> There is much evidence in this. Microsoft is "reinventing" many things, >>> and its implementation shows that they didn't bother to do any research >>> to help understand what they are doing. >> >>I kinda agree here, maybe not for 100%, but sorta. ;-) >>I (personally) think it's more likely a lack of experienced >>coders (straight from school they cost less, and I deliberately >>didn't say 'programmers'), than a lack of research. >> >>If they attracted more people that are experienced in other >>systems, I even think they wouldn't need a lot of research. > >Ahem...if? Why do you think things are getting reinvented. The >ideas are getting carried in by people who had already done it >before. Another explanation is that they worked it out themselves. I independently discovered finite state automata over fifteen years ago. I ran across the term many times and puzzled over it until finally one text was good enough to define it. Two thoughts occurred in quick succession. 1) Oh, that's what they are. 2) I worked that out fifteen years ago. A number of scientific discoveries were independently come to at about the same time by more than one person or group. Sincerely, Gene Wirchenko Computerese Irregular Verb Conjugation: I have preferences. You have biases. He/She has prejudices. ###### From: "Harry V. Quackenboss" Newsgroups: alt.folklore.computers,alt.os.multics References: <71huie$meo$2@irk.zetnet.co.uk> <727nqd$qrs$3@news.islandnet.com> <3647B9FD.4516@bell-labs.com> <36492344.644@bell-labs.com> Subject: Re: Multics Date: Thu, 12 Nov 1998 03:23:30 -0800 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Newsreader: Microsoft Outlook Express 4.72.3110.5 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 Lines: 22 Message-ID: <364ac535$0$12781@nntp1.ba.best.com> NNTP-Posting-Host: 209.157.70.90 X-Trace: nntp1.ba.best.com 910869813 12781 209.157.70.90 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news1.best.com!news3.best.com!nntp1.ba.best.com!not-for-mail One example of direct influence of C & Unix on a CPU design is the first MIPS Computer Systems chip, the R2000 (as distinguished from the Stanford MIPS machine) The early MIPS team sent a draft overview of the architecture, including despriptions of theinstruction set, addressing, cache, etc. to John Mashey (Unix PWB) who was the Unix engineering group manager at Convergent Technologies. To their suprise Mash sent back a multi-page critique, ending with a comment that it would be an interesting architecture upon which to run Unix. Shortly after John joined Mips, Inc., and had a big influence on the memory management and instruction set. The design was constrained, however, by what you could put on a single CMOS chip in 1985, which included a direct-map cache which was all they could afford to fit in the gate budget. Barry Margolin wrote in message ... >In article <36492344.644@bell-labs.com>, >Dennis Ritchie wrote: >>Also, although there may be a case to be made for C or C-like >>languages influencing (parts of) CPU design, I don't really >>see the one for Unix having much influence on broader aspects >>of hardware architecture (for better or worse). ###### From: "Harry V. Quackenboss" Newsgroups: alt.folklore.computers,alt.os.multics References: <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> <36449E63.65F3@ix.netcom.com> Subject: Re: Multics Date: Thu, 12 Nov 1998 04:35:16 -0800 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Newsreader: Microsoft Outlook Express 4.72.3110.5 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 Lines: 101 Message-ID: <364ad605$0$12761@nntp1.ba.best.com> NNTP-Posting-Host: 209.157.70.90 X-Trace: nntp1.ba.best.com 910874117 12761 209.157.70.90 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news1.best.com!news3.best.com!nntp1.ba.best.com!not-for-mail With due respect to John Couleur, I disagree with several of his points: John Couleur wrote: >The NSA architecture >was a complete secret outside of Honeywell. Inside of Honeywell, the NSA >was viewed as an architecture that corrected major flaws of the 645 - >small address, page sized segments, limited protection for need to know, >etc. CP-6 used the virtual memory features of NSA and GCOS was moving to >using the 34 bit address. Reverting to the MULTICS architecture meant >taking a giant step backward architecturally for GCOS. > While NSA certainly was innovative, it, and its cabability machine architecture siblings (Intel 432, Tandem Rainbow, etc.) proved unweildly complicated to use, and commercially unviable. GCOS never used most of the features, and was never made secure to the extent of NSA's lofty goals. Further to assert that the Multics architecture would have meant a "giant step backward" is debatable. First, as noted GCOS never used most of the complex features (I recall discussions over coffee about the byzantine ramifications of the "climb" instruction), and second, some important features of Multics, such as dynamic linking, could not be easily implemented with NSA. >Honeywell was selling >over $500 million per year of "equivalent" GCOS systems. Very few of >them were new systems. We were selling add-ons and upgrades for old >systems. And, the market for sales to old customers was growing by 20% >per year. It looked like we had a highly profitable money machine. Neither GCOS nor Multics was "highly profitable". Up until Jim Barrett's arrival the internal financial scorekeeping rules in LISD were rigged to overstate profitability across the board. This was done by reporting factory transfer prices and computing division operating margins without taking into consideration R&D for un-released products. The result was the manufacturing operation looked highly profitable, but Honeywell, Inc. wasn't. >At the time, the only >way that we could make MULTICS sales was to give away additional >processors and memories to get the performance/cost up to acceptable >levels. Obviously, none of the other studies had considered the >economics of the decision. In order to give MULTICS every opportunity to succeed, MULTICS >was put under a "czar" who was given authority to set prices and margins >down to production cost. In effect, with this arrangement GCOS' high >profitability would subsidize MULTICS and avoid putting MULTICS on a >profit or die basis. John is right that a big problem with Multics was that growing the business required securing new installations. Selling a $2M entry-level machine, where the customer had to port all their software, versus selling add-on equipment to a GCOS customer was a tough proposition. Where John's characterization is wrong is that the discounts for Multics were comparable to the discounts for competitive-bid GCOS systems, and based on all the hardware gross margin reports I saw so was the gross profit (profit without selling expense). Examples: at GM, the discounts for the initial installation and add-ons for Mulitcs were in the same ballpark as the GCOS systems at GM parts division, where IBM was the competition. Also, in all the UK univiersity competitive bids where we competed with DEC,CDC, and IBM, everybody's discounts got brutal. One other point about GCOS. What history has shown is that while the GCOS was perceived by its stakeholders as profitable, it was steadily losing market share. The company was not investing in new account sales nor in the features needed to compete for new accounts. In reality, it was slowly dying. . >It wasn't until the microcomputer came along that computing for >intellectual support became economically viable. While measurement of cost-savings and productivity was and still is hard to measure, the above statement is not true. Ford, for example, spent extensive time analyzing productivity and the impact of interactive computing, and they had a convincing story about several decision support applications (some on Multics, others on GCOS and DEC) that leveraged productivity improved car quality and customer satisfaction. Also, my father's market research department @ GM used outside time-sharing services to analyze data in ways that could not be done manually or with batch processing. >As a further thought, has anyone besides me noticed the similarity >between MULTICS and Windows9x. When my young friends from Microsoft tell >me about the wonderful things that they are going to do in Washington, I >have to tell them that the idea is at least 35 years old. DLLs, OLE, >virtual memory, ODBC, heirarchical file system, ease of use, all were >part of the original MULTICS. When do we get dynamic linking? How about being able to change the IP address without rebooting? Dynamic reconfiguration? Seriously, the best thing about Multics is how its students and developers have taken what they learned to the benefit of those of us who use Windows, SQL, Java, Unix, et al. In closing, I want to thank John for his post, and hope he stays engaged. ###### From: "Peter Mc Shane" Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 12 Nov 1998 14:27:25 GMT Organization: Indigo Lines: 23 Message-ID: <01be0e49$40a96f80$7b373737@peter> References: <71huie$meo$2@irk.zetnet.co.uk> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> NNTP-Posting-Host: ts07-002.dublin.indigo.ie X-Newsreader: Microsoft Internet News 4.70.1161 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!cpk-news-hub1.bbnplanet.com!cam-news-feed2.bbnplanet.com!news.gtei.net!news.indigo.ie!not-for-mail CP/M was quite portable as well. It was supported on 8080/Z80, 80X6, 68K and Z8000 processors as far as I am aware. Peter Seebach wrote in article ... > In article , > Barry Margolin wrote: > >For that matter, how many portable OSes have been implemented by *anyone*? > > Mach and Be come to mind. Also EPOC32, the more recent Psion system. > I'd count Linux as a separate run from Unix, because they reinvented all > the portability stuff from scratch, whereas NetBSD, et al., inherited it. > > -s > -- > Copyright 1998, All rights reserved. Peter Seebach / seebs@plethora.net > C/Unix wizard, Pro-commerce radical, Spam fighter. Boycott Spamazon! > Seeking interesting programming projects. Not interested in commuting. > Visit my new ISP --- More Net, Less Spam! > ###### From: Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Thu, 12 Nov 1998 18:16:26 -0500 Organization: Kettering University (formerly GMI E&MI) - Flint MI Lines: 34 Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> <01be0e49$40a96f80$7b373737@peter> NNTP-Posting-Host: nova.kettering.edu Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <01be0e49$40a96f80$7b373737@peter> Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!isdnet!newsfeed.cwix.com!24.128.1.125!chnws03.mediaone.net!24.131.1.12!denws01.mw.mediaone.net!news.gmi.edu!nova.kettering.edu!lee1089 On 12 Nov 1998, Peter Mc Shane wrote: > > CP/M was quite portable as well. > It was supported on 8080/Z80, 80X6, 68K and Z8000 processors as far as > I am aware. > > Peter Seebach wrote in article > ... > > In article , > > Barry Margolin wrote: > > >For that matter, how many portable OSes have been implemented by > *anyone*? > > > > Mach and Be come to mind. Also EPOC32, the more recent Psion system. > > I'd count Linux as a separate run from Unix, because they reinvented all > > the portability stuff from scratch, whereas NetBSD, et al., inherited it. The Mac OS is almost ported to PPC. Sure, most of the components on a Mac are similar between PPC and 68k versions. And the 68k emulator has kept Apple from totally rewriting everything in PPC code. And the port was very costly and time consuming so it was *ported* but not really *portable*. Also, NeXTStep/OpenStep was ported from 68k to 80x86. ____________________________________________________________________________ | "A little nonsense now and then, | "If it walks out of the fridge, let Is relished by the wisest men." | it go" -- John Dougherty --W.W. | "If it loves you it will come back." | -- Ian Davis __________________________________|_________________________________________ 340 days and counting... ###### From: benh@lsl.co.uk (Ben Hutchings) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Followup-To: alt.folklore.computers,alt.os.multics Date: 12 Nov 1998 20:59:20 GMT Organization: Laser-Scan Ltd. Lines: 48 Message-ID: <72fi78$20v@relay.lsl.co.uk> References: <71huie$meo$2@irk.zetnet.co.uk> <7274g1$pem$1@nntp1.u.washington.edu> NNTP-Posting-Host: 192.9.200.20 X-Newsreader: TIN [version 1.2 PL2] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!194.165.93.117.MISMATCH!newshub.bart.net!newsfeed.wirehub.nl!diablo.theplanet.net!join.news.pipex.net!pipex!news.lsl.co.uk!benh D. Peschel (dpeschel@u.washington.edu) wrote: : In article , : Arthur Evans Jr wrote: : >A British OS written in BCPL was quite successfully ported to : >sevreral platforrms, probably in the late '70s or early '80s. : Would that be TRIPOS? (or is it TripOS? I don't think so) I think so, yes. : The only things I know about it are: : - Was used to create DOS on the Amiga (or some other parts of the OS) The DOS library, file-system, console handler, shell, and a few other bits and pieces. : - May have nasty license requirements (apparently Amiga couldn't deliver : some of the documentation they wanted) The official documentation for AmigaDOS has always been very poor, but I've not heard that explanation given. Since the state of documentation has not improved despite many changes to AmigaDOS, I don't think that is the reason. : - Ported to a bunch of architectures That was the idea. TRIPOS is an abbreviation for TRIvial Portable Operating System. : - Supposedly a pretty clean design, though the BCPL runtime environment : can interfere with the C runtime environment (probably also with Pascal : and other environments too) The fact that BCPL assumes word addressing while C assumes byte addressing does not help. The BCPL compiler for a byte-addressing machine must left-shift pointers before using them. This is not only inefficient but also a real pain when integrating code written in other languages with a BCPL system like AmigaDOS. : - Named for an examination (at Cambridge or Oxford?) which in turn was : named for the three-legged stool provided during the examination -- : Is there another acronym? Was TRIPOS used as course material? A 'Tripos' can also mean the group of courses taught by a particular department that one can choose from. I understand that TRIPOS is still used at course material at Cambridge. ###### From: benh@lsl.co.uk (Ben Hutchings) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Followup-To: alt.folklore.computers,alt.os.multics Date: 12 Nov 1998 21:05:15 GMT Organization: Laser-Scan Ltd. Lines: 27 Message-ID: <72fiib$20v@relay.lsl.co.uk> References: <71huie$meo$2@irk.zetnet.co.uk> <727nqd$qrs$3@news.islandnet.com> NNTP-Posting-Host: 192.9.200.20 X-Newsreader: TIN [version 1.2 PL2] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newscore.univie.ac.at!newsfeed.ecrc.net!join.news.pipex.net!pipex!news.lsl.co.uk!benh Hello Kittyhawk (bdb@GTS.Net) wrote: : In article <727nqd$qrs$3@news.islandnet.com>, : Deryk Barker wrote: : |Arthur Evans Jr (eva_remove_this_ns@evans.pgh.pa.us) wrote: : |: Barry Margolin wrote: : |: >For that matter, how many portable OSes have been implemented by *anyone*? : |: : |: A British OS written in BCPL was quite successfully ported to : |: sevreral platforrms, probably in the late '70s or early '80s. : | : |I believe this would have been Martin Richards' TRIPOS. Richards was : |the author of BCPL (and CPL before it) and I believe TRIPOS was the : |foundation for the Atari (or was it Amiga?) OS. Can anyone confirm : |this? : For part of the Amiga system. It was grafted in at : the last minute because of some time slippage in the : production of the original Amiga I/O code... too bad, : since they had to rip it all out again later... No, that's not quite correct. It wasn't 'slippage' in the 'I/O code'. Another company whose name I don't know was contracted to produce the high-level parts of the operating system, to be called CAOS, on top of Carl Sassenrath's kernel (the executive, or 'exec'). It sound some- what Unix-ish. You can find the full story at . ###### From: "Samael" Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Fri, 13 Nov 1998 09:29:55 -0000 Lines: 35 Message-ID: <364bfc06.0@127.0.0.1> References: <71huie$meo$2@irk.zetnet.co.uk> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> <01be0e49$40a96f80$7b373737@peter> NNTP-Posting-Host: 195.147.15.144 X-Newsreader: Microsoft Outlook Express 4.72.3155.0 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3155.0 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!212.63.192.161.MISMATCH!newshub.bart.net!news.tele2.nl!newsfeed1.swip.net!swipnet!news.maxwell.syr.edu!nntp.news.xara.net!xara.net!news.itg.net.uk!127.0.0.1!192.168.0.1 lee1089@kettering.edu wrote in message ... >On 12 Nov 1998, Peter Mc Shane wrote: > >> >> CP/M was quite portable as well. >> It was supported on 8080/Z80, 80X6, 68K and Z8000 processors as far as >> I am aware. >> >> Peter Seebach wrote in article >> ... >> > In article , >> > Barry Margolin wrote: >> > >For that matter, how many portable OSes have been implemented by >> *anyone*? >> > >> > Mach and Be come to mind. Also EPOC32, the more recent Psion system. >> > I'd count Linux as a separate run from Unix, because they reinvented all >> > the portability stuff from scratch, whereas NetBSD, et al., inherited it. > >The Mac OS is almost ported to PPC. Sure, most of the components on a Mac >are similar between PPC and 68k versions. And the 68k emulator has kept >Apple from totally rewriting everything in PPC code. And the port was >very costly and time consuming so it was *ported* but not really >*portable*. Also, NeXTStep/OpenStep was ported from 68k to 80x86. NT is available for Intel and Alpha (and theoretically prtable to other systems as well, as you should only need to rewrite the HAL layer. Samael ###### From: gsgatlin@unity.ncsu.edu (Gary Scott Gatling) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Followup-To: alt.folklore.computers,alt.os.multics Date: 13 Nov 1998 12:25:43 GMT Organization: North Carolina State University Lines: 37 Message-ID: <72h8g7$29f$1@uni00nw.unity.ncsu.edu> References: <71huie$meo$2@irk.zetnet.co.uk> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> <727nqd$qrs$3@news.islandnet.com> <3647B9FD.4516@bell-labs.com> NNTP-Posting-Host: cc06du.unity.ncsu.edu X-Newsreader: TIN [UNIX 1.3 950824BETA PL0] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!howland.erols.net!news.maxwell.syr.edu!cpk-news-hub1.bbnplanet.com!news.gtei.net!gatech!nntp-xfer.ncsu.edu!not-for-mail > Deryk Barker wrote (there's subquoting here and this > > I believe this would have been Martin Richards' TRIPOS. Richards was > the author of BCPL (and CPL before it) and I believe TRIPOS was the > foundation for the Atari (or was it Amiga?) OS. Can anyone confirm > this? The Amiga's AmigaDOS was a port of TRIPOS. The original CAOS wasn't finished, so a company called Metacomco was hired to convert TRIPOS for use on the Amiga. When I say AmigaDOS I mean the the dos.library, console handler, and the command line interface (CLI). The exec kernal and other parts of the sysem like intuiion were not TRIPOS. AmigaDOS was written in BCPL and 68000 assembler and I think all the Amiga's CLI commands for the early releases were written in BCPL. Later all of it was coded in C, around the 2.0 release...(Thankfully) Incedentally, AmigaDOS is really weird with all the BCPL data structures hidden wihin it. I have recenly been trying to program my way through some stuff involving the dos.library on my Amiga. (BCPL strings to C strings, BCPL pointers to C pointers, and back...) :) I believe Metacomco's version of TRIPOS for the 68000 was called MCC. They made changes to their MCC to make it function with the rest of the AmigaOS. I wonder if Professor Richards will ever finish his TRIPOS for Linux? Looks interesting. -- ****************************************************** ** Gary Gatling (gsgatlin@eos.ncsu.edu) ** ** ICQ UIN 13514523 http://www4.ncsu.edu/~gsgatlin/ ** ** Justify my text? I'm sorry but it has no excuse. ** ****************************************************** ###### From: Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Fri, 13 Nov 1998 14:28:20 -0500 Organization: Kettering University (formerly GMI E&MI) - Flint MI Lines: 46 Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> <01be0e49$40a96f80$7b373737@peter> <364bfc06.0@127.0.0.1> NNTP-Posting-Host: nova.kettering.edu Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <364bfc06.0@127.0.0.1> Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!cpk-news-hub1.bbnplanet.com!news.gtei.net!newsfeed.cwix.com!24.130.1.14!lsnws01.we.mediaone.net!24.131.1.12!denws01.mw.mediaone.net!news.gmi.edu!nova.kettering.edu!lee1089 On Fri, 13 Nov 1998, Samael wrote: > > lee1089@kettering.edu wrote in message ... > >On 12 Nov 1998, Peter Mc Shane wrote: > > > >> > >> CP/M was quite portable as well. > >> It was supported on 8080/Z80, 80X6, 68K and Z8000 processors as far as > >> I am aware. > >> > >> Peter Seebach wrote in article > >> ... > >> > In article , > >> > Barry Margolin wrote: > >> > >For that matter, how many portable OSes have been implemented by > >> *anyone*? > >> > > >> > Mach and Be come to mind. Also EPOC32, the more recent Psion system. > >> > I'd count Linux as a separate run from Unix, because they reinvented > all > >> > the portability stuff from scratch, whereas NetBSD, et al., inherited > it. > > > >The Mac OS is almost ported to PPC. Sure, most of the components on a Mac > >are similar between PPC and 68k versions. And the 68k emulator has kept > >Apple from totally rewriting everything in PPC code. And the port was > >very costly and time consuming so it was *ported* but not really > >*portable*. Also, NeXTStep/OpenStep was ported from 68k to 80x86. > > > NT is available for Intel and Alpha (and theoretically prtable to other > systems as well, as you should only need to rewrite the HAL layer. VMS was also ported to Alpha. ____________________________________________________________________________ | "A little nonsense now and then, | "If it walks out of the fridge, let Is relished by the wisest men." | it go" -- John Dougherty --W.W. | "If it loves you it will come back." | -- Ian Davis __________________________________|_________________________________________ 336 days and counting... ###### Newsgroups: alt.folklore.computers References: <71huie$meo$2@irk.zetnet.co.uk> <01be0e49$40a96f80$7b373737@peter> <364bfc06.0@127.0.0.1> Reply-To: sarr@umich.edu Organization: University of Michigan Subject: Re: Multics From: sarr@engin.umich.edu (Sarr J. Blumson) Lines: 14 Message-ID: <2zX22.6511$fS.21378109@news.itd.umich.edu> Date: Fri, 13 Nov 1998 14:42:06 GMT NNTP-Posting-Host: 141.211.43.166 X-Trace: news.itd.umich.edu 910968126 141.211.43.166 (Fri, 13 Nov 1998 09:42:06 EDT) NNTP-Posting-Date: Fri, 13 Nov 1998 09:42:06 EDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!newspump.monmouth.com!newspeer.monmouth.com!newsxfer.itd.umich.edu!news.itd.umich.edu!sarr In article <364bfc06.0@127.0.0.1>, Samael wrote: > >NT is available for Intel and Alpha (and theoretically prtable to other >systems as well, as you should only need to rewrite the HAL layer. Dave Cutler has been quoted as saying that porting NT to a machine that isn't little-endian would be impossible. -- -------- Sarr Blumson sarr@umich.edu voice: +1 734 764 0253 home: +1 734 665 9591 ITD, University of Michigan http://www-personal.umich.edu/~sarr/ 519 W William, Ann Arbor, MI 48103-4943 ###### From: karger@watson.ibm.com (Paul A. Karger) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 13 Nov 1998 15:35:05 GMT Organization: IBM T.J. Watson Research Center Lines: 32 Message-ID: <72hjj9$f3g$1@mdnews.btv.ibm.com> References: <71huie$meo$2@irk.zetnet.co.uk> <7274g1$pem$1@nntp1.u.washington.edu> <72fi78$20v@relay.lsl.co.uk> NNTP-Posting-Host: watpub1.watson.ibm.com X-Newsreader: xrn 8.01 Originator: karger@watson.ibm.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news.idt.net!nyd.news.ans.net!news-w.ans.net!newsfeeds.ans.net!news.chips.ibm.com!newsfeed.btv.ibm.com!mdnews.btv.ibm.com!karger |> |> That was the idea. TRIPOS is an abbreviation for TRIvial Portable |> Operating System. |> |> |> : - Named for an examination (at Cambridge or Oxford?) which in turn was |> : named for the three-legged stool provided during the examination -- |> : Is there another acronym? Was TRIPOS used as course material? |> |> A 'Tripos' can also mean the group of courses taught by a particular |> department that one can choose from. I understand that TRIPOS is |> still used at course material at Cambridge. Yes - Tripos is the name of the exams taken at the conclusion of a course of study at Cambridge and is taken from the name of the three-legged stools that the students sat on when taking the examinations. Note that a course of study comprises what a US university would think of as many different courses. Tripos exams are closest to the comprehensive exams that are given in US universities in the early stages of a graduate degree program to assess how well you have assimilated the undergraduate material. TRIPOS was written to be the operating system run on the many processors of the Cambridge processor bank. Those processors were not all the same, so TRIPOS had to be portable to begin with. See "The Cambridge Distributed Computing System" by Needham and Herbert, published by Addison-Wesley in 1982. ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 13 Nov 1998 20:26:29 GMT Lines: 25 Message-ID: <72i4ll$59l$2@irk.zetnet.co.uk> References: <72fi78$20v@relay.lsl.co.uk> NNTP-Posting-Host: man-001.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 910988789 5429 194.247.41.1 (13 Nov 1998 20:26:29 GMT) NNTP-Posting-Date: 13 Nov 1998 20:26:29 GMT X-Everything: Net-Tamer V 1.08X Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!ayres.ftech.net!news.ftech.net!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-11-12 benh@lsl.co.uk(BenHutchings) said: :The fact that BCPL assumes word addressing while C assumes byte :addressing does not help. The BCPL compiler for a byte-addressing :machine must left-shift pointers before using them. This is not :only inefficient but also a real pain when integrating code written :in other languages with a BCPL system like AmigaDOS. Not necessarily inefficient; the 386 is a great BCPL machine, since it can auto-shift its pointers. I believe the ARM can do the same trick. But I see your point about integration with other languages, except those which also assume word addressing. I personally liked the Zed approach, which kept the concept of a word (and in fact defaulted to an unsigned word) but added typing such that you can have signed words, pointers to Zed types and functions returning Zed types. That way you keep the "everything's a word" simplicity of BCPL, but you don't have to muck about with addressing. And in any case, at root, the only purpose to C's typing system seems to have been to say how much space to reserve on the stack (I've got an FPU and I'm gonna use it); there doesn't appear to have been quite the Wirthian level of ideological commitment to data typing at Bell Labs... ;> -- Communa (lisard@zetnet.co.uk) -- you know soft spoken changes nothing ###### From: Dave Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Sat, 14 Nov 1998 01:25:06 +0000 Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <7274g1$pem$1@nntp1.u.washington.edu> <72fi78$20v@relay.lsl.co.uk> <72hjj9$f3g$1@mdnews.btv.ibm.com> NNTP-Posting-Host: sourplum.demon.co.uk X-NNTP-Posting-Host: sourplum.demon.co.uk:158.152.63.37 X-Trace: news.demon.co.uk 911007579 nnrp-09:22876 NO-IDENT sourplum.demon.co.uk:158.152.63.37 X-Complaints-To: abuse@demon.net MIME-Version: 1.0 X-Newsreader: Turnpike (32) Trial Version 3.05 Lines: 21 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.cwix.com!194.72.7.126!btnet-peer!btnet!dispose.news.demon.net!demon!news.demon.co.uk!demon!sourplum.demon.co.uk!dave In article <72hjj9$f3g$1@mdnews.btv.ibm.com>, Paul A. Karger writes >TRIPOS was written to be the operating system run on the many processors of the >Cambridge processor bank. Those processors were not all the same, so TRIPOS >had to be portable to begin with. See "The Cambridge Distributed Computing >System" by Needham and Herbert, published by Addison-Wesley in 1982. > IIRC Martin Richards said (in one of the CS lectures @ Cambridge in the early/mid eighties[1]) that Tripos was first conceived whilst he was at CERN as an OS to run on the many data collection machines that they had. Of course for several years porting Tripos to Architecture X (where X was the latest & greatest hw available that year) was a common 3rd year project for the CS students - the last I recall being to the 16032 second processor for a BBC micro....... [1] I'll see if I can find the notes from that one -- Dave ###### From: Vince Scarafino Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Mon, 16 Nov 1998 12:50:10 -0500 Organization: Ford Motor Company Lines: 24 Message-ID: <365065D2.4ADED3DA@pobox.srl.ford.com> References: <727n8j$qrs$1@news.islandnet.com> <727pkh$11j6$1@flood.weeg.uiowa.edu> Reply-To: vscarafi@pobox.srl.ford.com NNTP-Posting-Host: pms045.pd9.ford.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 4.5 [en]C-c32f404p (Win95; U) X-Accept-Language: en, en-GB, de, fr, ja, ko, zh Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.cwix.com!141.211.144.13!newsxfer3.itd.umich.edu!jobone!dailyplanet.srl.ford.com!eccws1.dearborn.ford.com!news I thought the first version of UNIX was on a GCOS system. "Douglas W. Jones,201H MLH,3193350740,3193382879" wrote: > > From article <727n8j$qrs$1@news.islandnet.com>, > by dbarker@camosun.bc.nospam.ca (Deryk Barker): > > > > And let us not forget, either, that the first version of Unix was > > written in PDP-11 assembler. > > The first version was written in assembly language for an 18-bit > PDP-9 (ancestor of the PDP-15, and descendant of the PDP-7) > > > Unless I am mistaken the rewrite in C (1973ish?) was in order to > > facilitate the port to the Unidata machine. > > PDP-11, if memory serves me right. I believe the port to > an Interdata machine was later. My first edition of K&R > says that C was developed under UNIX for PDP-11 UNIX, and > that, as of the first edition of K&R, it ran on the > IBM 370, Honeywell 6000, and the Interdata 8/32. > > Doug Jones > jones@cs.uiowa.edu ###### From: Vince Scarafino Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Mon, 16 Nov 1998 13:09:42 -0500 Organization: Ford Motor Company Lines: 42 Message-ID: <36506A66.816BEC36@pobox.srl.ford.com> References: <71huie$meo$2@irk.zetnet.co.uk> <36449E63.65F3@ix.netcom.com> <722d8h$qka@srvr1.engin.umich.edu> <36471490.3EAC70F5@pobox.srl.ford.com> <36487bd7.1065201@news.uunet.be> Reply-To: vscarafi@pobox.srl.ford.com NNTP-Posting-Host: pms045.pd9.ford.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 4.5 [en]C-c32f404p (Win95; U) X-Accept-Language: en, en-GB, de, fr, ja, ko, zh Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.cwix.com!141.211.144.13!newsxfer3.itd.umich.edu!jobone!dailyplanet.srl.ford.com!eccws1.dearborn.ford.com!news Luc Van der Veken wrote: > > Also sprach Vince Scarafino on Mon, > 09 Nov 1998 11:13:04 -0500 to alt.folklore.computers: > > > > The most blatant area that I > > see is the way they manage the namespace for dynamically linked > > objects. In reality, they don't bother to manage it! If you want to > > replace a dynamic link you have to reboot. I can find no concept of > > terminate reference. > > But here, I don't understand what you mean exactly. > Unload (terminate) all programs that depend on a dll (or ocx), > replace it, restart programs: that's what I do several times a > day when developing them, but that's not what I call rebooting. > > If you're referring to dynamically loaded device drivers, I think > it's exactly the same thing - though I'm not fully sure since I > don't write them, and seldom have to replace them (for which I > rely on installation programs). > > If it's a dynamic 'thing' the system (or the shell) depends on, > unloading dependent programs means shutting down - if that is > what you meant, you're right. Multics had the concept of "initiated segments" that was usually the first item in a process' search rules. Once a segment was referenced, it was normally used by any code that referenced the same external symbol. This worked very well for most environments, but made it difficult to test new code that was meant as a direct replacement for current production routines. Multics had the capability to terminate the segment's reference (un-initiate it) so that the next reference to that external symbol would not find it in "initiated segments" and therefore go through the other search rules to find the desired code. Thus, Multics allowed most of its development to be done without requiring a shutdown--something I see as a requirement for "Enterprise Class" software. I know of no way to do this with Microsoft operating systems. ###### From: jadams@NO.SPAM.PLEASE (James W. Adams) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 16 Nov 1998 14:13:25 -0500 Organization: UNIX Internals, Charlottesville, VA Lines: 28 Message-ID: <72ptgl$sru@jamesa.cstone.net> References: <727n8j$qrs$1@news.islandnet.com> <727pkh$11j6$1@flood.weeg.uiowa.edu> <365065D2.4ADED3DA@pobox.srl.ford.com> NNTP-Posting-Host: jamesa.cstone.net X-Trace: Skuzzy.cstone.net 911243916 30606 206.205.42.225 (16 Nov 1998 19:18:36 GMT) X-Complaints-To: usenet@news.cstone.net NNTP-Posting-Date: 16 Nov 1998 19:18:36 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!howland.erols.net!dca1-hub1.news.digex.net!dca1-feed4.news.digex.net!digex!news1.cstone.net!news.COMET.NET!not-for-mail In article <365065D2.4ADED3DA@pobox.srl.ford.com>, Vince Scarafino wrote: >I thought the first version of UNIX was on a GCOS system. It was cross-assembled on a GE-635 running GECOS for a DEC PDP-7 until it became self-sufficient on the '7. It never ran on GECOS or GCOS. The first UNIX port was to a PDP-11/20, still in assembler, in 1971. The initial impetus for development was reportedly to run Thompson's "Space Travel" simulation at lower cost and with smoother "motion" than could be obtained on the 645 Multics box which was loaded by other users. The prototype for what was to be the UNIX filesystem had already been developed and modeled under Multics, and evidently formed the basis for initial development for the PDP-7. Thus, early work seems to have been driven more by research in improved development environments, not merely the desire to run the simulation, which could already be cross-compiled/assembled. -- James W. Adams -- jamesa @ cstone.net <-- remove spaces "I became obsessed with angels and ballerinas, things of grace and beauty, otherworldly." Charlottesville, VA 22903 --C. Love ###### From: jadams@NO.SPAM.PLEASE (James W. Adams) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: 16 Nov 1998 14:38:31 -0500 Organization: UNIX Internals, Charlottesville, VA Lines: 20 Message-ID: <72puvn$st4@jamesa.cstone.net> References: <71huie$meo$2@irk.zetnet.co.uk> <36471490.3EAC70F5@pobox.srl.ford.com> <36487bd7.1065201@news.uunet.be> <36506A66.816BEC36@pobox.srl.ford.com> NNTP-Posting-Host: jamesa.cstone.net X-Trace: Skuzzy.cstone.net 911245381 30751 206.205.42.225 (16 Nov 1998 19:43:01 GMT) X-Complaints-To: usenet@news.cstone.net NNTP-Posting-Date: 16 Nov 1998 19:43:01 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!howland.erols.net!newsfeed.fast.net!ptdnetP!newsgate.ptd.net!news1.cstone.net!news.COMET.NET!not-for-mail In article <36506A66.816BEC36@pobox.srl.ford.com>, Vince Scarafino wrote: >Thus, Multics allowed most of its development to be done without >requiring a shutdown--something I see as a requirement for "Enterprise >Class" software. I know of no way to do this with Microsoft operating >systems. To be fair, though, the OS on most "Enterprise Class" systems in the hands of end user clients serves as little more than an environment to run a DBMS and a transaction monitor. Unless one is a OS vendor, OS development is at best overhead, not an end product. -- James W. Adams -- jamesa @ cstone.net <-- remove spaces "I became obsessed with angels and ballerinas, things of grace and beauty, otherworldly." Charlottesville, VA 22903 --C. Love ###### From: eva_remove_this_ns@evans.pgh.pa.us (Arthur Evans Jr) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Mon, 16 Nov 1998 15:24:00 -0500 Organization: Ada Consulting Lines: 19 Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <727nqd$qrs$3@news.islandnet.com> <72fiib$20v@relay.lsl.co.uk> NNTP-Posting-Host: ppp39.s8.pgh.net X-Newsreader: MT-NewsWatcher 2.3.1 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!howland.erols.net!newsfeed.fast.net!news.pgh.net!ppp39.s8.pgh.net!user I said > A British OS written in BCPL was quite successfully ported to > sevreral platforrms, probably in the late '70s or early '80s. and Deryk Barker wrote: > I believe this would have been Martin Richards' TRIPOS. Richards was > the author of BCPL (and CPL before it) Yes it was TRIPOS (thanks; I had forgotten that name), and yes Martin Richards was involved (though I'm not sure that he was the author). However, Martin did _not_ write CPL; rather, Christopher Strachey did. Martin was Christopher's graduate student, and after completing his graduate work, he came to MIT to implement CPL on several computers. As a first step he implemented a bootstrap version of CPL, called BCPL, and never got back to booting CPL. BCPL, of course, is the ancestor of C. Art Evans ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers,alt.os.multics,microsoft.public.win32.programmer.kernel Subject: Re: Multics Followup-To: microsoft.public.win32.programmer.kernel Date: Mon, 16 Nov 1998 22:16:06 GMT Organization: . Lines: 74 Message-ID: <3654a3c1.8021003@news.uunet.be> References: <71huie$meo$2@irk.zetnet.co.uk> <36449E63.65F3@ix.netcom.com> <722d8h$qka@srvr1.engin.umich.edu> <36471490.3EAC70F5@pobox.srl.ford.com> <36487bd7.1065201@news.uunet.be> <36506A66.816BEC36@pobox.srl.ford.com> NNTP-Posting-Host: pool02b-194-7-146-129.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newscore.univie.ac.at!btnet-peer!btnet!diablo.theplanet.net!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Vince Scarafino wrote in alt.folklore.computers: [I hope he doesn't mind me pushing this toward win32.kernel - if it isn't on topic there, it isn't on topic anywhere, and maybe it could lead to some interesting discussion. Followups set to m.p.w32.p.k, btw] > ... Multics had the capability to terminate > the segment's reference (un-initiate it) so that the next reference to > that external symbol would not find it in "initiated segments" and > therefore go through the other search rules to find the desired code. > Thus, Multics allowed most of its development to be done without > requiring a shutdown--something I see as a requirement for "Enterprise > Class" software. I know of no way to do this with Microsoft operating > systems. Thanks for clearing that up - but maybe you better hadn't told me. I'm starting to miss it in NT already... You other win32-ers, how would you feel about being able (up to a certain point) to replace a dll without having to shut down the applications that are using it (let alone the complete system)? I may be oversimplifying it (due to a lack of in-depth knowledge), but as far as I understand how a journaling file system works, it doesn't seem like a job that can't be overcome in NT (though it's on a much lower level than where I feel at ease). The restrictions placed on the new version would be severe, but during debugging or for minor modifications it should be possible, considering that: - all entry points must stay at the same addresses (a jump table at the start of the image would ease that), or there must be a way to tell callers the 'GetProcAddress'-ed pointers have gone bad. - all static data must stay at the same addresses - there must be a way to tell the new version that it's already been initialized (sounds easy - just don't call the init code). The way I see it, replacing the old version would require some cooperation between different parts of the system (file system, thread scheduler): - write the new version to disk while the old one is still being used (meaning both versions must be able to co-exist on disk briefly - this part seems trickier for a non-journaling FS) - as soon as the new copy is complete, hold new calls to entry points in the old image (preempt those threads), wait for calls in progress to complete (hoping they're not waiting for some event that's supposed to get signaled by one of the threads that got held) - invalidate the old image in memory and on disk, make the new image valid - map the new image to memory and start processing the calls that were held. - wait for your screen to turn blue ;-) Probably my fantasy is carrying me beyond what's practically possible, but it sure sounds nice. Maybe something for NT6 ?? Next item on my wish list: replace exe's while they are running. Oops - sorry - they already invented that in VC6 (at least as long as you run them in the debugger ;) [and I hope Vince isn't laughing too hard by now :)] -- My undergraduate experience convinced me that I was not smart enough to be a physicist, and that computers were quite neat. Dennis M. Ritchie ###### From: Vince Scarafino Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Tue, 17 Nov 1998 11:03:10 -0500 Organization: Ford Motor Company Lines: 27 Message-ID: <36519E3E.46A881D1@pobox.srl.ford.com> References: <71huie$meo$2@irk.zetnet.co.uk> <36449E63.65F3@ix.netcom.com> <722d8h$qka@srvr1.engin.umich.edu> <36471490.3EAC70F5@pobox.srl.ford.com> <36487bd7.1065201@news.uunet.be> Reply-To: vscarafi@pobox.srl.ford.com NNTP-Posting-Host: pms045.pd9.ford.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 4.5 [en]C-c32f404p (Win95; U) X-Accept-Language: en, en-GB, de, fr, ja, ko, zh To: Luc Van der Veken Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.cwix.com!141.211.144.13!newsxfer3.itd.umich.edu!jobone!dailyplanet.srl.ford.com!eccws1.dearborn.ford.com!news Luc Van der Veken wrote: > > Also sprach Vince Scarafino on Mon, > 09 Nov 1998 11:13:04 -0500 to alt.folklore.computers: > > > There is much evidence in this. Microsoft is "reinventing" many things, > > and its implementation shows that they didn't bother to do any research > > to help understand what they are doing. > > I kinda agree here, maybe not for 100%, but sorta. ;-) > I (personally) think it's more likely a lack of experienced > coders (straight from school they cost less, and I deliberately > didn't say 'programmers'), than a lack of research. > > If they attracted more people that are experienced in other > systems, I even think they wouldn't need a lot of research. > I'd like to believe you, but it appears to me that the shortcoming in the dynamic linking subsystem is a fundamental design problem. The namespace for dynamically linked objects doesn't appear to be managed. Indeed, there doesn't seem to even be the concept of a namespace applied to these objects. Can't blame coders for that. If you're telling me there are no senior designers, only coders, than I guess I understand your logic. ###### From: Mike Swaim Subject: Re: Multics Newsgroups: alt.folklore.computers,alt.os.multics References: <71huie$meo$2@irk.zetnet.co.uk> <36449E63.65F3@ix.netcom.com> <722d8h$qka@srvr1.engin.umich.edu> <36471490.3EAC70F5@pobox.srl.ford.com> <36487bd7.1065201@news.uunet.be> <36519E3E.46A881D1@pobox.srl.ford.com> User-Agent: tin/pre-1.4-980226 (UNIX) (FreeBSD/2.2.6-RELEASE (i386)) Lines: 16 Message-ID: Date: Tue, 17 Nov 1998 18:01:59 GMT NNTP-Posting-Host: 209.127.0.130 X-Trace: news1.giganews.com 911325719 209.127.0.130 (Tue, 17 Nov 1998 12:01:59 CDT) NNTP-Posting-Date: Tue, 17 Nov 1998 12:01:59 CDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.cwix.com!207.207.0.26!nntp.giganews.com!news1.giganews.com.POSTED!gemini.c-com.net!swaim In alt.folklore.computers Vince Scarafino wrote: : I'd like to believe you, but it appears to me that the shortcoming in : the dynamic linking subsystem is a fundamental design problem. The : namespace for dynamically linked objects doesn't appear to be managed. : Indeed, there doesn't seem to even be the concept of a namespace applied : to these objects. It's just very poorly documented. The rules NT use to identify DLLs are described in about a half dozen knowledge base articles, some of which are available via MSDN online. The differences between NT and Win9x are even more poorly documented. -- Mike Swaim, Avatar of Chaos: Disclaimer:I sometimes lie. Home: swaim@c-com.net Alum: swaim@alumni.rice.edu Quote: "Boingie"^4 Y,W&D ###### From: dbarker@camosun.bc.nospam.ca (Deryk Barker) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Followup-To: alt.folklore.computers,alt.os.multics Date: 18 Nov 1998 01:47:05 GMT Organization: Camosun College, Victoria B.C. Lines: 13 Message-ID: <72t8up$l05$3@news.islandnet.com> References: <71huie$meo$2@irk.zetnet.co.uk> <727nqd$qrs$3@news.islandnet.com> <72fiib$20v@relay.lsl.co.uk> NNTP-Posting-Host: 204.174.56.1 X-Newsreader: TIN [UNIX 1.3 950824BETA PL0] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newshub.northeast.verio.net!news1.best.com!news.islandnet.com!not-for-mail Arthur Evans Jr (eva_remove_this_ns@evans.pgh.pa.us) wrote: [...] : However, Martin did _not_ write CPL; rather, Christopher Strachey did. Ah! The denotational semantics guy? You learn something new every day.... -- |Deryk Barker, Computer Science Dept. | Music does not have to be understood| |Camosun College, Victoria, BC, Canada| It has to be listened to. | |email: dbarker@camosun.bc.ca | | |phone: +1 250 370 4452 | Hermann Scherchen. | ###### Message-ID: <36538129.E68C20FF@shore.net> From: John W Gintell X-Mailer: Mozilla 4.5 (Macintosh; U; PPC) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics References: <71huie$meo$2@irk.zetnet.co.uk> <36449E63.65F3@ix.netcom.com> <722d8h$qka@srvr1.engin.umich.edu> <36471490.3EAC70F5@pobox.srl.ford.com> <36487bd7.1065201@news.uunet.be> <36519E3E.46A881D1@pobox.srl.ford.com> Content-Type: text/plain; charset=us-ascii; x-mac-type="54455854"; x-mac-creator="4D4F5353" Content-Transfer-Encoding: 7bit Lines: 21 Date: Wed, 18 Nov 1998 21:26:07 -0500 NNTP-Posting-Host: 204.167.102.48 X-Complaints-To: abuse@shore.net X-Trace: news.shore.net 911442180 204.167.102.48 (Wed, 18 Nov 1998 21:23:00 EDT) NNTP-Posting-Date: Wed, 18 Nov 1998 21:23:00 EDT Organization: Shore.Net/Eco Software, Inc; (info@shore.net) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news.shore.net!not-for-mail Mike Swaim wrote: > In alt.folklore.computers Vince Scarafino wrote: > : I'd like to believe you, but it appears to me that the shortcoming in > : the dynamic linking subsystem is a fundamental design problem. The > : namespace for dynamically linked objects doesn't appear to be managed. > : Indeed, there doesn't seem to even be the concept of a namespace applied > : to these objects. > > It's just very poorly documented. The rules NT use to identify DLLs > are described in about a half dozen knowledge base articles, some of which > are available via MSDN online. > The differences between NT and Win9x are even more poorly documented. It actually amazes me how poorly mainstream products from mainstream companies are documented. In our miniscule Multics group we stressed documentation in many ways. From the first Multics documents in the MSPM, to including the tech writer organization in all aspects of the development process, to having common "source" for help files and printed documents so that they would be consistent, and many more themes helped. ###### From: Allyn Lai Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Thu, 19 Nov 1998 08:23:12 +0000 Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <36449E63.65F3@ix.netcom.com> <722d8h$qka@srvr1.engin.umich.edu> <36471490.3EAC70F5@pobox.srl.ford.com> <36487bd7.1065201@news.uunet.be> <36519E3E.46A881D1@pobox.srl.ford.com> <36538129.E68C20FF@shore.net> NNTP-Posting-Host: allyn.demon.co.uk X-NNTP-Posting-Host: allyn.demon.co.uk:158.152.189.215 X-Trace: news.demon.co.uk 911504126 nnrp-07:12832 NO-IDENT allyn.demon.co.uk:158.152.189.215 X-Complaints-To: abuse@demon.net MIME-Version: 1.0 X-Newsreader: Turnpike (32) Version 4.00 Lines: 47 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!news-feed1.eu.concert.net!btnet-peer!btnet!dispose.news.demon.net!demon!news.demon.co.uk!demon!allyn.demon.co.uk!allyn In article <36538129.E68C20FF@shore.net>, John W Gintell writes >Mike Swaim wrote: > >> In alt.folklore.computers Vince Scarafino wrote: >> : I'd like to believe you, but it appears to me that the shortcoming in >> : the dynamic linking subsystem is a fundamental design problem. The >> : namespace for dynamically linked objects doesn't appear to be managed. >> : Indeed, there doesn't seem to even be the concept of a namespace applied >> : to these objects. >> >> It's just very poorly documented. The rules NT use to identify DLLs >> are described in about a half dozen knowledge base articles, some of which >> are available via MSDN online. >> The differences between NT and Win9x are even more poorly documented. > >It actually amazes me how poorly mainstream products from mainstream companies >are documented. In our miniscule Multics group we stressed documentation in many >ways. From the first Multics documents in the MSPM, to including the tech writer >organization in all aspects of the development process, to having common >"source" >for help files and printed documents so that they would be consistent, and many >more themes helped. > The lack of documentation used to mystify me too. But then I realized that training wasn't such a big industry in the old days. Companies like Microsoft and Novell realized that there was a lot money to be made through training and certification programs. Clear, concise documentation goes against this philosophy. Maybe that's too cynical because there is a lot to be said in favor of professional training. Good training can be faster, more efficient and give more consistent results than having people read through manuals. The software companies probably feel that if you are a serious, large corporate user you'll send your users to a training course. Therefore, they don't see the point in spending a lot of time on documentation. Besides, if you had better documentation you wouldn't have all these fun conferences to go to. It's all a big scam....(-; Cynically yours, Allyn -- Allyn Lai ###### From: jnickelsen@acm.org (Juergen Nickelsen) Newsgroups: alt.folklore.computers,alt.os.multics Subject: Re: Multics Date: Sat, 21 Nov 1998 01:18:40 +0100 Organization: [Posted via] Interactive Networx Lines: 11 Message-ID: <1dinwoc.1hkdtrgz13o3iN@n65-146.berlin.snafu.de> References: <71a66l$bfe$1@flood.weeg.uiowa.edu> <71cf04$5fk@jamesa.cstone.net> <71d3e0$b1h@freenet-news.carleton.ca> <71huie$meo$2@irk.zetnet.co.uk> <3298305.66060.3537@kcbbs.gen.nz> <364092B2.8277B176@nvi.to> <71stjv$fio@jamesa.cstone.net> <71tln8$qmm$1@mdnews.btv.ibm.com> NNTP-Posting-Host: n65-146.berlin.snafu.de X-Newsreader: MacSOUP 2.3.3 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!news.maxwell.syr.edu!newsfeed.ecrc.net!195.21.255.252.MISMATCH!unlisys!news.snafu.de!jnickelsen Barry Margolin wrote: > I think NT runs on x86 and Alpha architecures. Not to support NT, but: I remember it has been available for PowerPC and MIPS platforms. I think this was with NT 3.51. (According to some source it was ported to the MIPS R4000 because the 32-bit Intel CPUs were no longer able to express Bill Gates's wealth in one machine word. :-) -- Juergen Nickelsen ###### From: Bruce Cook Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 21 Nov 1998 01:34:45 -0800 Organization: Synonet Corporation - The Bicycle Factory Lines: 34 Message-ID: References: <71huie$meo$2@irk.zetnet.co.uk> <36538129.E68C20FF@shore.net> <733rdt$h9r$1@ligarius.ultra.net> NNTP-Posting-Host: per2-77.wantree.com.au X-Newsreader: Gnus v5.3/Emacs 19.34 Cache-Post-Path: cletus!unknown@donal.tara X-Cache: nntpcache 2.3.3 (see http://www.nntpcache.org/) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news.tele2.nl!newsfeed1.swip.net!swipnet!news.maxwell.syr.edu!news.mel.connect.com.au!news.per.connect.com.au!news.waia.asn.au!news.wantree.com.au!not-for-mail sarr@engin.umich.edu (Sarr J. Blumson) writes: > True, but my recollection is that the _only_ way to get the TOPS-10 > internal documentation was to take the courses. Interesting... I've always found DEC documentation to be excellent but then I never did any real work with the '10. Their RSX and VMS documentation was accurate, showed all side effects, explained what all the possible errors meant in the context of the operation, cross referenced other relevent functions & documentation and provided excellent examples. I never did an RSX course, but was doing systems programming on '11s for about 12 years. (even wrote an ACP {the strange coloured rat from memory} :) When we got O/S updates I'd read all the updated manual sections (having read all the manuals before) and was able to use new directives or update drivers immediately without having to muck about working out why things behaved strangely. When we moved over to developing on NT, I found the documentation utter crap - inaccurate, incomplete, simplistic, non-existant the few examples often didn't even compile. Coding by trial-and-error. -- ...BRU Bruce Cook, Synonet Corp. E-Mail: bruce@bicycle.synonet.com Phone: +61 417 967 468 Fax: +61 8 9227 7390 ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 21 Nov 1998 02:19:13 GMT Lines: 48 Message-ID: <7357v1$uiv$2@irk.zetnet.co.uk> References: <733rdt$h9r$1@ligarius.ultra.net> NNTP-Posting-Host: man-107.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 911614753 31327 194.247.40.137 (21 Nov 1998 02:19:13 GMT) NNTP-Posting-Date: 21 Nov 1998 02:19:13 GMT X-Everything: Net-Tamer V 1.08X Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!news-feed1.eu.concert.net!btnet-peer!btnet!news.freedom2surf.net!ayres.ftech.net!news.ftech.net!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-11-20 jmfbahciv@aol.com said: (here, have a decaf) :Were you around back then? Training was a part of the package :that was sold in our business. :That last line is bullshit. How does one teach the stuff if clear, :concise documentation doesn't exist? Biblically. ("The blind leading the blind"...) :In my day, the manuals were used as reference material; one could :not possibly know everything about a system. Updates to the :manuals was also the way to document changes to new releases of :software w/o expending resources on an extra manual of the :before/now changes. They still are, now, but I'll warrant that the major thing to have changed since earlier times is that the expectation is that users will no longer program their computers. Hence MS documentation is lousy, because it is uneconomical to make it good; 90% of the time, the stuff that doesn't get covered doesn't get done either. The common stuff will work out of the box, and anything beyond that is probably not something MS are too keen on themselves... That's another reason for encouragng the spread of Open Source software, incidentally. It doesn't matter how lousy the documentation is for the Linux kernel; anyone who can read C can work out how it works, and more importantly, why it isn't. With MS, when the internals of the particular knackered bit being presented as a "black box", the documentation of how to get it to work has to be that much better to be of any use. It isn't. The "no documentation because of CNE/MSCE" line is bollocks, of course, because the two things are complementary (or should be). Unfortunately, these days the long-term part of the equation is too often absent. It's understandable why, though, given the pace of change in the industry. Who wants to sit around documenting something that'll be obsolete by the time the proofs have come back? Another issue, though. Perhaps the documentation is not actually worse, it just feels worse because it isn't printed out in a 6ft wall of rainforest...? Only now can I read a screen as easily as a printed page. -- Communa (lisard@zetnet.co.uk) -- you know soft spoken changes nothing ###### From: fleetvis@ricochet.net (phil) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sat, 21 Nov 1998 10:35:47 GMT Organization: Posted via Supernews, http://www.supernews.com - Discussions start here! Lines: 61 Message-ID: <365691f6.19845816@news.ricochet.net> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> NNTP-Posting-Host: 204.179.134.61 X-Trace: 911644375 F7W8ZAFJM863DCCB3C usenet78.supernews.com X-Complaints-To: newsabuse@supernews.com X-Newsreader: Forte Free Agent 1.11/32.235 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.berkeley.edu!Supernews73!supernews.com!Supernews69!not-for-mail On 21 Nov 1998 02:19:13 GMT, lisard@zetnet.co.uk wrote: >On 1998-11-20 jmfbahciv@aol.com said: >(here, have a decaf) > :Were you around back then? Training was a part of the package > :that was sold in our business. > :That last line is bullshit. How does one teach the stuff if clear, > :concise documentation doesn't exist? >Biblically. ("The blind leading the blind"...) > :In my day, the manuals were used as reference material; one could > :not possibly know everything about a system. Updates to the > :manuals was also the way to document changes to new releases of > :software w/o expending resources on an extra manual of the > :before/now changes. >They still are, now, but I'll warrant that the major thing to have >changed since earlier times is that the expectation is that users will >no longer program their computers. Hence MS documentation is lousy, >because it is uneconomical to make it good; 90% of the time, the stuff >that doesn't get covered doesn't get done either. The common stuff will >work out of the box, and anything beyond that is probably not something >MS are too keen on themselves... No this hasn't changed. Users are not programmers. This has been accepted in the PC world for a long time. It was also accepted in the IBM world. The stuff is equipment and it works out of the box, be it software or hardware. >That's another reason for encouragng the spread of Open Source >software, incidentally. It doesn't matter how lousy the documentation is >for the Linux kernel; anyone who can read C can work out how it works, >and more importantly, why it isn't. With MS, when the internals of the >particular knackered bit being presented as a "black box", the >documentation of how to get it to work has to be that much better to be >of any use. It isn't. Waste of time. 99% of users are not programmers. Of the 1% that are, 95% will not modify an existing function within a library because it will cause compatibility problems and quite rightly so. >The "no documentation because of CNE/MSCE" line is bollocks, of course, >because the two things are complementary (or should be). Unfortunately, >these days the long-term part of the equation is too often absent. It's >understandable why, though, given the pace of change in the industry. >Who wants to sit around documenting something that'll be obsolete by the >time the proofs have come back? Yes, this is total bollocks. It's a mind game persuading people to attend bizarre and irrelevant courses. I should know, my company is convinced that this is important and pays people silly amounts of money to present courses to otheres and then pays the same people silly amounts of money for us to attend them. Could this be a scam? I don't think so. :) >Another issue, though. Perhaps the documentation is not actually worse, >it just feels worse because it isn't printed out in a 6ft wall of >rainforest...? Only now can I read a screen as easily as a printed page. >Communa (lisard@zetnet.co.uk) -- you know soft spoken changes nothing phil. ###### From: huge@nospam.demon.co.uk (Hugh Davies) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 21 Nov 1998 11:37:34 GMT Organization: Piglet's Pickles and Preserves Message-ID: <7368lu$h03@axalotl.demon.co.uk> References: <733rdt$h9r$1@ligarius.ultra.net> Reply-To: huge@nospam.demon.co.uk NNTP-Posting-Host: axalotl.demon.co.uk X-NNTP-Posting-Host: axalotl.demon.co.uk:158.152.24.143 X-Trace: news.demon.co.uk 911649676 nnrp-09:14411 NO-IDENT axalotl.demon.co.uk:158.152.24.143 X-Complaints-To: abuse@demon.net x-no-archive: yes Lines: 23 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!dispose.news.demon.net!demon!news.demon.co.uk!demon!axalotl.demon.co.uk!axalotl!usenet In article <733rdt$h9r$1@ligarius.ultra.net>, jmfbahciv@aol.com writes: >>Companies like Microsoft and Novell realized that there was a lot money >>to be made through training and certification programs. Clear, concise >>documentation goes against this philosophy. > >That last line is bullshit. How does one teach the stuff if clear, >concise documentation doesn't exist? Forget to switch your brain on this morning? Oh, hang on, you're an AOL subscriber. You don't have one. Hint: The only way to get "clear, concise documentation" is to pay extra for it. Either as training, or as a 2000 page book with a "free" CD-ROM. At least in the World According to Bill Gates it is. -- "The road to Paradise is through Intercourse." The uk.transport FAQ; http://www.axalotl.demon.co.uk/transport/FAQ.html [Substitute "axalotl" for "nospam" to email me] ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sat, 21 Nov 98 13:25:08 GMT Organization: UltraNet Communications, Inc. Lines: 107 Message-ID: <736g5a$i5u$6@ligarius.ultra.net> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> NNTP-Posting-Host: d11.dial-11.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 21 Nov 1998 13:45:14 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newscore.univie.ac.at!howland.erols.net!feed1.news.rcn.net!rcn!news.ultranet.com!d11 In article <7357v1$uiv$2@irk.zetnet.co.uk>, lisard@zetnet.co.uk wrote: > > >On 1998-11-20 jmfbahciv@aol.com said: > >(here, have a decaf) Thanks. I need something to chase down this Valium. oooohhmmmmmm..... > > :Were you around back then? Training was a part of the package > :that was sold in our business. > > :That last line is bullshit. How does one teach the stuff if clear, > :concise documentation doesn't exist? > >Biblically. ("The blind leading the blind"...) Yea, well, I think their eyes are wide open. > > :In my day, the manuals were used as reference material; one could > :not possibly know everything about a system. Updates to the > :manuals was also the way to document changes to new releases of > :software w/o expending resources on an extra manual of the > :before/now changes. > >They still are, now, but I'll warrant that the major thing to have >changed since earlier times is that the expectation is that users will >no longer program their computers. Hence MS documentation is lousy, >because it is uneconomical to make it good; 90% of the time, the stuff >that doesn't get covered doesn't get done either. The common stuff will >work out of the box, and anything beyond that is probably not something >MS are too keen on themselves... Since their common stuff doesn't work out of the box, I've figured out that the reason there isn't more of a hue and cry is that this not-working behaviour has just been accepted by the uneducated masses as "that's how it works" and live with it. If they ever had an exposure to software that worked well (in your and my definition), there would be a great increase in market pressure for improvement; this action has been muddied by the alleged monopoly. And one doesn't have to know programming or be a programmer in order to use a computer system. Horrors! Have I just committed a sin? :-) People in the olden days didn't have to know how to program in order to use the system; it was just more efficient if you knew a trick or two. > >That's another reason for encouragng the spread of Open Source >software, incidentally. I am assuming that by "Open Source" you mean that the sources are shipped with the code? I still haven't decided if that's a good way to go. We did ship sources and the side effects of that action was ultimately our downfall. Due to bad judgement, mismanagement, and politics, the side effects of shipping sources were used to "prove" that our product was no good. > It doesn't matter how lousy the documentation is >for the Linux kernel; anyone who can read C can work out how it works, >and more importantly, why it isn't. That's true as far as it goes. What isn't documented in the sources is the intent of the implementation. This was discussed in our RFP docs (Request For P???) from the field, architectural specs, functional specs, and the notes of the brainstorming meetings. > With MS, when the internals of the >particular knackered bit being presented as a "black box", the >documentation of how to get it to work has to be that much better to be >of any use. It isn't. > >The "no documentation because of CNE/MSCE" line is bollocks, CNE/MSCE? > of course, >because the two things are complementary (or should be). Unfortunately, >these days the long-term part of the equation is too often absent. It's >understandable why, though, given the pace of change in the industry. >Who wants to sit around documenting something that'll be obsolete by the >time the proofs have come back? I don't think that's why the stuff isn't documented. > >Another issue, though. Perhaps the documentation is not actually worse, >it just feels worse because it isn't printed out in a 6ft wall of >rainforest...? Only now can I read a screen as easily as a printed page. IMO, what's missing is the "I want to do this thingie, what's the command?" One can't scan on-line documentation because of the structure of the keyword heirarchy. One tends to get lost in the trees (of course, that's assuming the information is there in the first place; it isn't w.r.t. Msoft stuff). /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sat, 21 Nov 98 13:28:26 GMT Organization: UltraNet Communications, Inc. Lines: 35 Message-ID: <736gbf$i5u$7@ligarius.ultra.net> References: <733rdt$h9r$1@ligarius.ultra.net> <7368lu$h03@axalotl.demon.co.uk> NNTP-Posting-Host: d11.dial-11.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 21 Nov 1998 13:48:31 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.maxwell.syr.edu!newsfeed.xcom.net!news.ultranet.com!d11 In article <7368lu$h03@axalotl.demon.co.uk>, huge@nospam.demon.co.uk (Hugh Davies) wrote: >In article <733rdt$h9r$1@ligarius.ultra.net>, jmfbahciv@aol.com writes: > >>>Companies like Microsoft and Novell realized that there was a lot money >>>to be made through training and certification programs. Clear, concise >>>documentation goes against this philosophy. >> >>That last line is bullshit. How does one teach the stuff if clear, >>concise documentation doesn't exist? > >Forget to switch your brain on this morning? That's always possible. > Oh, hang on, you're an AOL subscriber. You don't have one. Oops. You just lost 50 points on my "good-guy" list. I'll have to be more careful with your posts [sad emoticon here]. > >Hint: The only way to get "clear, concise documentation" is to pay extra >for it. Either as training, or as a 2000 page book with a "free" CD-ROM. > >At least in the World According to Bill Gates it is. > Hint: I wasn't talking about BGates' business; I was talking about my business. The original poster hit a sore point and I blasted. Would you like to blast back now? :-) /BAH Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sat, 21 Nov 98 13:37:47 GMT Organization: UltraNet Communications, Inc. Lines: 55 Message-ID: <736gt0$i5u$8@ligarius.ultra.net> References: <71huie$meo$2@irk.zetnet.co.uk> <36538129.E68C20FF@shore.net> <733rdt$h9r$1@ligarius.ultra.net> NNTP-Posting-Host: d11.dial-11.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 21 Nov 1998 13:57:52 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsfeed.xcom.net!news.ultranet.com!d11 In article , sarr@engin.umich.edu (Sarr J. Blumson) wrote: >In article <733rdt$h9r$1@ligarius.ultra.net>, wrote: >>In article , >>> >>>The lack of documentation used to mystify me too. But then I realized >>>that training wasn't such a big industry in the old days. >> >>Were you around back then? Training was a part of the package >>that was sold in our business. > >True, but my recollection is that the _only_ way to get the TOPS-10 >internal documentation was to take the courses. No. With TOPS-10, the only internal documentation was stuff that we didn't want customers to see. We were the only ones to put our functional specs in the Notebooks (although that was a constant fight when we acquired professional editors). We had quite a number of internal problems which involved an unrealistic pricing policy. That snuffed more available documentation than anything else. Wait a minute...are you talking about the Monitor Internals course? There was set of slides that were transferred to paper for the students that described 5.07 monitor tables. That particular piece of documentation did not make it to the Notebooks (but not for lack of trying). If you wish, I can try to remember the reasons it didn't get there. > >>>Companies like Microsoft and Novell realized that there was a lot money >>>to be made through training and certification programs. Clear, concise >>>documentation goes against this philosophy. >> >>That last line is bullshit. How does one teach the stuff if clear, >>concise documentation doesn't exist? > >Because the instructor provides much of the functionality >that's missing if you don't have things like good indexes >and cross referencing. > And do you think that these companies would trust the individual instructor enough to not insist that the course materials be put on paper? I don't think so. Getting the course material in some form other than a person's memory is just good business. Consider the case if that person gets hit by a truck. /BAH Subtract a hundred and four for e-mail. ###### From: huge@nospam.demon.co.uk (Hugh Davies) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 21 Nov 1998 14:13:11 GMT Organization: Piglet's Pickles and Preserves Message-ID: <736hpn$h4o@axalotl.demon.co.uk> References: <736gbf$i5u$7@ligarius.ultra.net> Reply-To: huge@nospam.demon.co.uk NNTP-Posting-Host: axalotl.demon.co.uk X-NNTP-Posting-Host: axalotl.demon.co.uk:158.152.24.143 X-Trace: news.demon.co.uk 911664088 nnrp-11:1966 NO-IDENT axalotl.demon.co.uk:158.152.24.143 X-Complaints-To: abuse@demon.net x-no-archive: yes Lines: 46 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!isdnet!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!dispose.news.demon.net!demon!news.demon.co.uk!demon!axalotl.demon.co.uk!axalotl!usenet In article <736gbf$i5u$7@ligarius.ultra.net>, jmfbahciv@aol.com writes: >In article <7368lu$h03@axalotl.demon.co.uk>, > huge@nospam.demon.co.uk (Hugh Davies) wrote: >>In article <733rdt$h9r$1@ligarius.ultra.net>, jmfbahciv@aol.com writes: >> >>>>Companies like Microsoft and Novell realized that there was a lot money >>>>to be made through training and certification programs. Clear, concise >>>>documentation goes against this philosophy. >>> >>>That last line is bullshit. How does one teach the stuff if clear, >>>concise documentation doesn't exist? >> >>Forget to switch your brain on this morning? > >That's always possible. > >> Oh, hang on, you're an AOL subscriber. You don't have one. > >Oops. You just lost 50 points on my "good-guy" list. Shucks. I expect I'll live. >I'll have >to be more careful with your posts [sad emoticon here]. > >> >>Hint: The only way to get "clear, concise documentation" is to pay extra >>for it. Either as training, or as a 2000 page book with a "free" CD-ROM. >> >>At least in the World According to Bill Gates it is. >> > >Hint: I wasn't talking about BGates' business; I was talking >about my business. I guess the words "Companies like Microsoft and Novell" went whistling right over your pointy little head. How sad. -- "The road to Paradise is through Intercourse." The uk.transport FAQ; http://www.axalotl.demon.co.uk/transport/FAQ.html [Substitute "axalotl" for "nospam" to email me] ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sat, 21 Nov 1998 15:22:08 GMT Organization: . Message-ID: <3656d249.334310@news.uunet.be> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> NNTP-Posting-Host: pool02b-194-7-150-28.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Lines: 23 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newsfeed.uk.ibm.net!ibm.net!dispose.news.demon.net!demon!newsfeed.ecrc.net!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach lisard@zetnet.co.uk on 21 Nov 1998 02:19:13 GMT to alt.folklore.computers: > Another issue, though. Perhaps the documentation is not actually worse, > it just feels worse because it isn't printed out in a 6ft wall of > rainforest...? Only now can I read a screen as easily as a printed page. That might be a major point. With dead tree documentation, you can easily have 5 or 6 binders open on your desk, consulting 2 or 3 of them in parallel without losing track of where you were. Now try that on a monitor. Another possible point: didn't systems get more complex over the years? Managing the docs (especially cross-links etc.) would become an increasingly more complex job as well. How many pages of text information can you store on a CD? My estimate is, uncompressed, at least 250-300,000. That makes an MSDN library (1 install CD, 2 data CD's) more than just a single wall - yet a lot of people (in a lot of different newsgroups) say there is more than just a /feeling/ of incompleteness when you're looking for some specific things (and some of them quote actual examples). ###### From: huge@nospam.demon.co.uk (Hugh Davies) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 21 Nov 1998 16:05:57 GMT Organization: Piglet's Pickles and Preserves Message-ID: <736od5$h6j@axalotl.demon.co.uk> References: <3656d249.334310@news.uunet.be> Reply-To: huge@nospam.demon.co.uk NNTP-Posting-Host: axalotl.demon.co.uk X-NNTP-Posting-Host: axalotl.demon.co.uk:158.152.24.143 X-Trace: news.demon.co.uk 911668418 nnrp-10:24646 NO-IDENT axalotl.demon.co.uk:158.152.24.143 X-Complaints-To: abuse@demon.net x-no-archive: yes Lines: 15 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!dispose.news.demon.net!demon!news.demon.co.uk!demon!axalotl.demon.co.uk!axalotl!usenet In article <3656d249.334310@news.uunet.be>, lucvdv@null.net (Luc Van der Veken) writes: >Another possible point: didn't systems get more complex over the >years? No. The wheel just got re-invented over and over and over and over again. -- "The road to Paradise is through Intercourse." The uk.transport FAQ; http://www.axalotl.demon.co.uk/transport/FAQ.html [Substitute "axalotl" for "nospam" to email me] ###### From: "Donald Tees" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sat, 21 Nov 1998 10:48:22 -0500 Organization: IGS - Information Gateway Services Lines: 11 Message-ID: <7372j8$k7c$1@news.igs.net> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <3656d249.334310@news.uunet.be> NNTP-Posting-Host: ttye07.kw.igs.net X-Trace: news.igs.net 911674792 20716 206.248.37.135 (21 Nov 1998 18:59:52 GMT) X-Complaints-To: abuse@igs.net NNTP-Posting-Date: 21 Nov 1998 18:59:52 GMT X-Newsreader: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!cyclone.bc.net!torn!nntp.igs.net!news.igs.net!not-for-mail Luc Van der Veken wrote in message <3656d249.334310@news.uunet.be>... >That might be a major point. With dead tree documentation, you >can easily have 5 or 6 binders open on your desk, consulting 2 or >3 of them in parallel without losing track of where you were. >Now try that on a monitor. I have four monitors ... ###### From: jadams@NO.SPAM.PLEASE (James W. Adams) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 21 Nov 1998 16:46:49 -0500 Organization: UNIX Internals, Charlottesville, VA Lines: 20 Message-ID: <737cc9$e8s@jamesa.cstone.net> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <365691f6.19845816@news.ricochet.net> NNTP-Posting-Host: jamesa.cstone.net X-Trace: Skuzzy.cstone.net 911685093 14659 206.205.42.225 (21 Nov 1998 21:51:33 GMT) X-Complaints-To: usenet@news.cstone.net NNTP-Posting-Date: 21 Nov 1998 21:51:33 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newscore.univie.ac.at!howland.erols.net!newsfeed.fast.net!ptdnetP!newsgate.ptd.net!news1.cstone.net!news.COMET.NET!not-for-mail In article <365691f6.19845816@news.ricochet.net>, phil wrote: >On 21 Nov 1998 02:19:13 GMT, lisard@zetnet.co.uk wrote: > >>On 1998-11-20 jmfbahciv@aol.com said: >>Biblically. ("The blind leading the blind"...) > >No this hasn't changed. Users are not programmers. This has been All hail the Ignorant, for They shall inherit the Earth. If users won't program, why are they being sold general-purpose computers? -- James W. Adams -- jamesa @ cstone.net <-- remove spaces "I became obsessed with angels and ballerinas, things of grace and beauty, otherworldly." Charlottesville, VA 22903 --C. Love ###### Sender: marc@dumbcat.snafu.org Newsgroups: alt.folklore.computers Subject: Re: Multics References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <736g5a$i5u$6@ligarius.ultra.net> From: Marco S Hyman Date: 21 Nov 1998 17:18:40 -0800 Message-ID: Organization: S.N.A.F.U. (www.snafu.org) X-Newsreader: Gnus v5.5/Emacs 20.3 Lines: 41 NNTP-Posting-Host: 204.94.187.130 X-Trace: nntp1.ba.best.com 911697521 12773 marc@204.94.187.130 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!news.idt.net!news.voicenet.com!news2.best.com!news3.best.com!nntp1.ba.best.com!not-for-mail jmfbahciv@aol.com writes: > Since their common stuff doesn't work out of the box, I've figured > out that the reason there isn't more of a hue and cry is that this > not-working behaviour has just been accepted by the uneducated masses > as "that's how it works" and live with it. If they ever had an So true. Have you ever noticed the strange looks you get from one of the "uneducated masses" if you happen to bitch about the lack of quality in their favorite application. I've come to use this line: as a trained professional I recognize crap when I see it. That usually gets me a shocked look or two! > And one doesn't have to know programming or be a programmer in > order to use a computer system. Horrors! Have I just committed > a sin? :-) People in the olden days didn't have to know how > to program in order to use the system; it was just more efficient > if you knew a trick or two. Minor disagreement. It was better if joe user became an expert in the application s/he was using, not necessarily the ins and outs of programming in general. In the USAF, for example, there was a job title something along the lines of "system monitor". A "system monitor" was the person who knew how to translate a users request for "the monthly foobar report" into operator instructions to "run program zyz999 with these control cards using tape 12345 as input and put the output on 5-part paper. Note: the documentation for these systems were for the "systems monitor" and not for the end user. Or maybe the end user just had different doc to which I never paid any attention :-) The users that had the most fun (and frustration) and got the most out of the system were those that managed to learn how to use the generic data extract and report utility to meet their custom reporting needs. I'd guess that today the equivalent would be a user who's become a SQL expert. Does that make the user a programmer? // marc ###### From: arargh@arargh.com (Arargh!) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sat, 21 Nov 1998 23:33:28 GMT Organization: Arargh!! Lines: 15 Message-ID: <36574d7a.52718663@news.mcs.net> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <3656d249.334310@news.uunet.be> <7372j8$k7c$1@news.igs.net> Reply-To: arargh@arargh.com NNTP-Posting-Host: jwright.pr.mcs.net Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.452 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!cpk-news-hub1.bbnplanet.com!chicago-news-feed1.bbnplanet.com!news.gtei.net!news.mcs.net!ddsw1!news.mcs.net!not-for-mail On Sat, 21 Nov 1998 10:48:22 -0500, "Donald Tees" wrote: >Luc Van der Veken wrote in message <3656d249.334310@news.uunet.be>... >>That might be a major point. With dead tree documentation, you >>can easily have 5 or 6 binders open on your desk, consulting 2 or >>3 of them in parallel without losing track of where you were. >>Now try that on a monitor. > > >I have four monitors ... gee, I only have three ..... but I'm working on more. > > ###### From: arargh@arargh.com (Arargh!) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sat, 21 Nov 1998 23:33:35 GMT Organization: Arargh!! Lines: 13 Message-ID: <36584dba.52782585@news.mcs.net> References: <3656d249.334310@news.uunet.be> <736od5$h6j@axalotl.demon.co.uk> Reply-To: arargh@arargh.com NNTP-Posting-Host: jwright.pr.mcs.net Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.452 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!news.new-york.net!schenectady.netmonger.net!news.mcs.net!ddsw1!news.mcs.net!not-for-mail On 21 Nov 1998 16:05:57 GMT, huge@nospam.demon.co.uk (Hugh Davies) wrote: >In article <3656d249.334310@news.uunet.be>, lucvdv@null.net (Luc Van der Veken) writes: > >>Another possible point: didn't systems get more complex over the >>years? > >No. > >The wheel just got re-invented over and over and over and over again. but, what color should it be? arargh ###### From: "Donald Tees" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sun, 22 Nov 1998 03:45:07 -0500 Organization: IGS - Information Gateway Services Lines: 12 Message-ID: <738il2$i4q$1@news.igs.net> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <365691f6.19845816@news.ricochet.net> <737cc9$e8s@jamesa.cstone.net> NNTP-Posting-Host: ttye12.kw.igs.net X-Trace: news.igs.net 911724002 18586 206.248.37.146 (22 Nov 1998 08:40:02 GMT) X-Complaints-To: abuse@igs.net NNTP-Posting-Date: 22 Nov 1998 08:40:02 GMT X-Newsreader: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!news.maxwell.syr.edu!sunqbc.risq.qc.ca!torn!nntp.igs.net!news.igs.net!not-for-mail James W. Adams wrote in message <737cc9$e8s@jamesa.cstone.net>... >All hail the Ignorant, for They shall inherit the Earth. > >If users won't program, why are they being sold general-purpose computers? Probably because they know a programmer that can write a program that lasts more than a couple years. I take it that you wrote every byte of software on all your computers? ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sun, 22 Nov 98 11:12:06 GMT Organization: UltraNet Communications, Inc. Lines: 60 Message-ID: <738so4$3e0$2@ligarius.ultra.net> References: <736gbf$i5u$7@ligarius.ultra.net> <736hpn$h4o@axalotl.demon.co.uk> NNTP-Posting-Host: d8.dial-14.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 22 Nov 1998 11:32:20 GMT X-Newsreader: News Xpress Version 1.0 Beta #4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!newspeer1.nac.net!netnews.com!feed1.news.rcn.net!rcn!news.ultranet.com!d8 In article <736hpn$h4o@axalotl.demon.co.uk>, huge@nospam.demon.co.uk (Hugh Davies) wrote: >In article <736gbf$i5u$7@ligarius.ultra.net>, jmfbahciv@aol.com writes: >>In article <7368lu$h03@axalotl.demon.co.uk>, >> huge@nospam.demon.co.uk (Hugh Davies) wrote: >>>In article <733rdt$h9r$1@ligarius.ultra.net>, jmfbahciv@aol.com writes: >>> >>>>>Companies like Microsoft and Novell realized that there was a lot money >>>>>to be made through training and certification programs. Clear, concise >>>>>documentation goes against this philosophy. >>>> >>>>That last line is bullshit. How does one teach the stuff if clear, >>>>concise documentation doesn't exist? >>> >>>Forget to switch your brain on this morning? >> >>That's always possible. >> >>> Oh, hang on, you're an AOL subscriber. You don't have one. >> >>Oops. You just lost 50 points on my "good-guy" list. > >Shucks. I expect I'll live. I suspect you will....but you won't have any fun :-). > > >>I'll have >>to be more careful with your posts [sad emoticon here]. >> >>> >>>Hint: The only way to get "clear, concise documentation" is to pay extra >>>for it. Either as training, or as a 2000 page book with a "free" CD-ROM. >>> >>>At least in the World According to Bill Gates it is. >>> >> >>Hint: I wasn't talking about BGates' business; I was talking >>about my business. > >I guess the words "Companies like Microsoft and Novell" went whistling >right over your pointy little head. How sad. Read the post before the little companies were dragged as a red herring. He was talking about main frame companies in the beginning of the post. However, it is possible that I misread the intent of the post. Now, do you really want to get nasty? Rearranging my cranial geometry is a good start. /BAH > > Subtract a hundred and four for e-mail. ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sun, 22 Nov 98 11:20:19 GMT Organization: UltraNet Communications, Inc. Lines: 68 Message-ID: <738t7h$3e0$3@ligarius.ultra.net> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <736g5a$i5u$6@ligarius.ultra.net> NNTP-Posting-Host: d8.dial-14.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 22 Nov 1998 11:40:33 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-nyc.telia.net!newspump.monmouth.com!newspeer.monmouth.com!news.ultranet.com!d8 In article , Marco S Hyman wrote: >jmfbahciv@aol.com writes: > >> Since their common stuff doesn't work out of the box, I've figured >> out that the reason there isn't more of a hue and cry is that this >> not-working behaviour has just been accepted by the uneducated masses >> as "that's how it works" and live with it. If they ever had an > >So true. Have you ever noticed the strange looks you get from >one of the "uneducated masses" if you happen to bitch about the >lack of quality in their favorite application. I've come to use >this line: as a trained professional I recognize crap when I see >it. That usually gets me a shocked look or two! Yup. I had figured that the reason nothing worked for me was a side effect of my magnetic personality :-). Then I started to do an informal survey and discovered just how bad the stuff worked. I was just a tad disgruntled because we could have addressed that as part of our OS 20-25 years ago. But we didn't have the fore-sight. > >> And one doesn't have to know programming or be a programmer in >> order to use a computer system. Horrors! Have I just committed >> a sin? :-) People in the olden days didn't have to know how >> to program in order to use the system; it was just more efficient >> if you knew a trick or two. > >Minor disagreement. It was better if joe user became an expert >in the application s/he was using, not necessarily the ins and >outs of programming in general. But that's not programming. I agree that there's a gray area. > >In the USAF, for example, there was a job title something along >the lines of "system monitor". A "system monitor" was the person >who knew how to translate a users request for "the monthly foobar >report" into operator instructions to "run program zyz999 with >these control cards using tape 12345 as input and put the output >on 5-part paper. Maybe that's why menus got invented? > >Note: the documentation for these systems were for the "systems >monitor" and not for the end user. Or maybe the end user just >had different doc to which I never paid any attention :-) IME, they did. Each level had their own docs to describe the functions of the job. Those docs were around mainly for training or temping when the main character was sick or on vacation. Mind you, those docs could be just somebodys doodles on a napkin, but they did exist. > >The users that had the most fun (and frustration) and got the most >out of the system were those that managed to learn how to use the >generic data extract and report utility to meet their custom reporting >needs. I'd guess that today the equivalent would be a user who's >become a SQL expert. Does that make the user a programmer? It's a start. It was also [what is now known as] a "career path". A lot of those people got real curious about what was going on behind the scenes. Those are the ones who usually worked themselves into a programming job (I did :-)). /BAH Subtract a hundred and four for e-mail. ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sun, 22 Nov 1998 12:03:15 GMT Organization: . Lines: 28 Message-ID: <3659f9f9.2389085@news.uunet.be> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <365691f6.19845816@news.ricochet.net> <737cc9$e8s@jamesa.cstone.net> NNTP-Posting-Host: pool02b-194-7-150-109.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!howland.erols.net!rill.news.pipex.net!pipex!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach jadams@NO.SPAM.PLEASE (James W. Adams) on 21 Nov 1998 16:46:49 -0500 to alt.folklore.computers: > In article <365691f6.19845816@news.ricochet.net>, > phil wrote: > >On 21 Nov 1998 02:19:13 GMT, lisard@zetnet.co.uk wrote: > > > >>On 1998-11-20 jmfbahciv@aol.com said: > >>Biblically. ("The blind leading the blind"...) > > > >No this hasn't changed. Users are not programmers. This has been > > All hail the Ignorant, for They shall inherit the Earth. > > If users won't program, why are they being sold general-purpose computers? But of course. That's why every bank teller, every accountant's clerk and every doctor that wants to 'computerize' his patients records have to follow those programming courses, before they're allowed to use a computer. Companies that sell complete systems with the software for applications like the above pre-installed, and train their customers in the use of _just_ that application software, are a danger to society, and should be outlawed. All hail the ignorant indeed. ###### Date: 22 Nov 98 13:29:44 -0800 From: "Charlie Gibbs" Subject: Re: Multics References: <3656d249.334310@news.uunet.be> <736od5$h6j@axalotl.demon.co.uk> Message-ID: <807.630T2205T8095174@sky.bus.com> Newsgroups: alt.folklore.computers Lines: 24 X-Newsreader: THOR 2.5a (Amiga;TCP/IP) NNTP-Posting-Host: news.skybus.com X-Trace: 22 Nov 1998 14:27:40 -0800, news.skybus.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!cyclone.news.idirect.com!island.idirect.com!news1.tor.metronet.ca!news1.cal.metronet.ca!news1.van.metronet.ca!kneehigh.imag.net!news.westel.com!news.skybus.com!204.244.247.103 In article <736od5$h6j@axalotl.demon.co.uk> huge@nospam.demon.co.uk (Hugh Davies) writes: >In article <3656d249.334310@news.uunet.be>, lucvdv@null.net >(Luc Van der Veken) writes: > >>Another possible point: didn't systems get more complex over the >>years? > >No. That's debatable. Your average user now has more computer power on his or her desk than it took to put a man on the moon, but still has trouble writing a letter. Doing tech support for that sort of user is a lot harder under the latest whiz-bang Windoze environment than it is with one of those primitive command lines. >The wheel just got re-invented over and over and over and over again. "Necessity is the mother of the re-invented wheel." -- cgibbs@sky.bus.com (Charlie Gibbs) Remove the first period after the "at" sign to reply. ###### Date: 22 Nov 98 13:34:53 -0800 From: "Charlie Gibbs" Subject: Re: Multics References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <365691f6.19845816@news.ricochet.net> <737cc9$e8s@jamesa.cstone.net> <3659f9f9.2389085@news.uunet.be> Message-ID: <766.630T2695T8146524@sky.bus.com> Newsgroups: alt.folklore.computers Lines: 21 X-Newsreader: THOR 2.5a (Amiga;TCP/IP) NNTP-Posting-Host: news.skybus.com X-Trace: 22 Nov 1998 14:27:45 -0800, news.skybus.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newshub.northeast.verio.net!cyclone.news.idirect.com!island.idirect.com!news1.tor.metronet.ca!news1.cal.metronet.ca!news1.van.metronet.ca!kneehigh.imag.net!news.westel.com!news.skybus.com!204.244.247.103 In article <3659f9f9.2389085@news.uunet.be> lucvdv@null.net (Luc Van der Veken) writes: >Companies that sell complete systems with the software for >applications like the above pre-installed, and train their >customers in the use of _just_ that application software, >are a danger to society, and should be outlawed. Hear, hear! Remember when the word "turnkey" was used to describe such systems? Oddly, I haven't heard that term much lately. Stan Kelly-Bootle, in "The Devil's DP Dictionary", pointed out that the word "turnkey" is often pronounced with a silent "n". >All hail the ignorant indeed. Ignorance _is_ bliss - if you're the one exploiting the ignorant. -- cgibbs@sky.bus.com (Charlie Gibbs) Remove the first period after the "at" sign to reply. ###### From: "Donald Tees" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sun, 22 Nov 1998 23:08:32 -0500 Organization: IGS - Information Gateway Services Lines: 22 Message-ID: <73ampk$im3$1@news.igs.net> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <365691f6.19845816@news.ricochet.net> <737cc9$e8s@jamesa.cstone.net> <3659f9f9.2389085@news.uunet.be> <766.630T2695T8146524@sky.bus.com> NNTP-Posting-Host: ttye0b.kw.igs.net X-Trace: news.igs.net 911793780 19139 206.248.37.139 (23 Nov 1998 04:03:00 GMT) X-Complaints-To: abuse@igs.net NNTP-Posting-Date: 23 Nov 1998 04:03:00 GMT X-Newsreader: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!news.maxwell.syr.edu!sunqbc.risq.qc.ca!torn!nntp.igs.net!news.igs.net!not-for-mail Charlie Gibbs wrote in message <766.630T2695T8146524@sky.bus.com>... >In article <3659f9f9.2389085@news.uunet.be> lucvdv@null.net >(Luc Van der Veken) writes: > >>Companies that sell complete systems with the software for >>applications like the above pre-installed, and train their >>customers in the use of _just_ that application software, >>are a danger to society, and should be outlawed. > >Hear, hear! Remember when the word "turnkey" was used to >describe such systems? Oddly, I haven't heard that term >much lately. Stan Kelly-Bootle, in "The Devil's DP Dictionary", >pointed out that the word "turnkey" is often pronounced with a >silent "n". When it reaches the state that only programmers can use a computer, then it is time for all programmers to go back to school and learn to do it right. ###### Newsgroups: alt.folklore.computers References: <71huie$meo$2@irk.zetnet.co.uk> <733rdt$h9r$1@ligarius.ultra.net> <736gt0$i5u$8@ligarius.ultra.net> Reply-To: sarr@umich.edu Organization: University of Michigan Subject: Re: Multics From: sarr@engin.umich.edu (Sarr J. Blumson) Lines: 23 Message-ID: <8a562.964$CY1.4247777@news.itd.umich.edu> Date: Mon, 23 Nov 1998 04:05:24 GMT NNTP-Posting-Host: 141.212.2.33 X-Trace: news.itd.umich.edu 911793924 141.212.2.33 (Sun, 22 Nov 1998 23:05:24 EDT) NNTP-Posting-Date: Sun, 22 Nov 1998 23:05:24 EDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!howland.erols.net!newsxfer.itd.umich.edu!news.itd.umich.edu!sarr In article <736gt0$i5u$8@ligarius.ultra.net>, wrote: >In article , > sarr@engin.umich.edu (Sarr J. Blumson) wrote: >>True, but my recollection is that the _only_ way to get the TOPS-10 >>internal documentation was to take the courses. > >Wait a minute...are you talking about the Monitor Internals >course? There was set of slides that were transferred to >paper for the students that described 5.07 monitor tables. >That particular piece of documentation did not make it >to the Notebooks (but not for lack of trying). If you wish, >I can try to remember the reasons it didn't get there. Yes. And to be fair I had gone on to other things way before 5.07. The last time I needed the monitor internals was when I was trying to carry ADP's (actually still Cyphernetics') hacks from 4.72 to 5.02. -- -------- Sarr Blumson sarr@umich.edu voice: +1 734 764 0253 home: +1 734 665 9591 ITD, University of Michigan http://www-personal.umich.edu/~sarr/ 519 W William, Ann Arbor, MI 48103-4943 ###### From: jmfbahciv@aol.com Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Mon, 23 Nov 98 12:53:10 GMT Organization: UltraNet Communications, Inc. Lines: 28 Message-ID: <73bn1u$fcl$1@strato.ultra.net> References: <71huie$meo$2@irk.zetnet.co.uk> <733rdt$h9r$1@ligarius.ultra.net> <736gt0$i5u$8@ligarius.ultra.net> <8a562.964$CY1.4247777@news.itd.umich.edu> NNTP-Posting-Host: d1.dial-11.mbo.ma.ultra.net X-Complaints-To: abuse@ultra.net X-Ultra-Time: 23 Nov 1998 13:13:34 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!isdnet!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsfeed.cwix.com!209.244.253.199!newsfeed.xcom.net!news.ultranet.com!d1 In article <8a562.964$CY1.4247777@news.itd.umich.edu>, sarr@engin.umich.edu (Sarr J. Blumson) wrote: >In article <736gt0$i5u$8@ligarius.ultra.net>, wrote: >>In article , >> sarr@engin.umich.edu (Sarr J. Blumson) wrote: >>>True, but my recollection is that the _only_ way to get the TOPS-10 >>>internal documentation was to take the courses. >> >>Wait a minute...are you talking about the Monitor Internals >>course? There was set of slides that were transferred to >>paper for the students that described 5.07 monitor tables. >>That particular piece of documentation did not make it >>to the Notebooks (but not for lack of trying). If you wish, >>I can try to remember the reasons it didn't get there. > >Yes. And to be fair I had gone on to other things way before 5.07. The >last time I needed the monitor internals was when I was trying to carry >ADP's (actually still Cyphernetics') hacks from 4.72 to 5.02. > Ah, now I understand. There was a huge diffference between those years and the later years. At least, you had the sources :-). /BAH Subtract a hundred and four for e-mail. ###### From: jcmorris@mwunix.mitre.org (Joe Morris) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 23 Nov 1998 14:52:05 GMT Organization: The MITRE Corporation Lines: 13 Message-ID: <73bsql$r91@top.mitre.org> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <365691f6.19845816@news.ricochet.net> <737cc9$e8s@jamesa.cstone.net> <3659f9f9.2389085@news.uunet.be> <766.630T2695T8146524@sky.bus.com> NNTP-Posting-Host: mwunix.mitre.org Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news-feed.inet.tele.dk!bofh.vszbr.cz!newshub.northeast.verio.net!news-pen-3.sprintlink.net!news-in-east1.sprintlink.net!news.sprintlink.net!coop.net!world!blanket.mitre.org!news.mitre.org!mwunix!jcmorris "Charlie Gibbs" writes: >Hear, hear! Remember when the word "turnkey" was used to >describe such systems? Oddly, I haven't heard that term >much lately. Stan Kelly-Bootle, in "The Devil's DP Dictionary", >pointed out that the word "turnkey" is often pronounced with a >silent "n". Where is Kelly-Bootle these days? I've not seen anything from him since I lost my subscription to _DEC Professional_ (at least that's where I think his column was published...) Joe Morris ###### Newsgroups: alt.folklore.computers From: eric@fudge.uchicago.edu (Eric Fischer) Subject: Re: Multics X-Nntp-Posting-Host: fudge.uchicago.edu Message-ID: Sender: news@midway.uchicago.edu (News Administrator) X-Newsposter: Pnews 4.0-test51 (15 Jan 97) Organization: The University of Chicago References: <733rdt$h9r$1@ligarius.ultra.net> <3659f9f9.2389085@news.uunet.be> <766.630T2695T8146524@sky.bus.com> <73bsql$r91@top.mitre.org> Date: Mon, 23 Nov 1998 15:53:08 GMT Lines: 10 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news-feed.inet.tele.dk!bofh.vszbr.cz!howland.erols.net!vixen.cso.uiuc.edu!uchinews!not-for-mail Joe Morris wrote: > Where is Kelly-Bootle these days? I've not seen anything from him since > I lost my subscription to _DEC Professional_ (at least that's where I > think his column was published...) He still has a monthly column on the back page of Unix Review's Performance Computing. eric ###### From: Mike Swaim Subject: Re: Multics Newsgroups: alt.folklore.computers References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> User-Agent: tin/pre-1.4-980226 (UNIX) (FreeBSD/2.2.6-RELEASE (i386)) Lines: 19 Message-ID: Date: Mon, 23 Nov 1998 16:14:46 GMT NNTP-Posting-Host: 209.127.0.130 X-Trace: news2.giganews.com 911837686 209.127.0.130 (Mon, 23 Nov 1998 10:14:46 CDT) NNTP-Posting-Date: Mon, 23 Nov 1998 10:14:46 CDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.cwix.com!207.207.0.26!nntp.giganews.com!news2.giganews.com.POSTED!gemini.c-com.net!swaim lisard@zetnet.co.uk wrote: : Another issue, though. Perhaps the documentation is not actually worse, : it just feels worse because it isn't printed out in a 6ft wall of : rainforest...? Only now can I read a screen as easily as a printed page. On the other hand, the MSDN CDs (now DVD) were darn useful for looking up stuff. And they also include a decent number of white papers and sample apps (with source). A big part of the problem is that they're supporting two platforms with widely different internals (Win9x and NT), and they're always creating new APIs. (For instance, we're up to DirectX 6, with DirectX 7 on the way next year.) -- Mike Swaim, Avatar of Chaos: Disclaimer:I sometimes lie. Home: swaim@c-com.net Alum: swaim@alumni.rice.edu Quote: "Boingie"^4 Y,W&D ###### From: alex*@*rockvax.rockefeller.edu (Alexandre Pechtchanski) Newsgroups: alt.folklore.computers Subject: Re: Multics Organization: Rockefeller University Hospital (GCRC), New York Message-ID: <3659a660.264769714@Rockyd> References: <736gbf$i5u$7@ligarius.ultra.net> <736hpn$h4o@axalotl.demon.co.uk> X-Newsreader: Forte Agent 1.5/32.451 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 10 Date: Mon, 23 Nov 1998 18:18:53 GMT NNTP-Posting-Host: 129.85.24.56 X-Trace: rockyd.rockefeller.edu 911845136 129.85.24.56 (Mon, 23 Nov 1998 13:18:56 EDT) NNTP-Posting-Date: Mon, 23 Nov 1998 13:18:56 EDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news.tele2.nl!newsfeed1.swip.net!swipnet!newsfeed.nyu.edu!rockyd.rockefeller.edu!not-for-mail On 21 Nov 1998 14:13:11 GMT, huge@nospam.demon.co.uk (Hugh Davies) wrote: [ Courtesy cc'ed through e-mail to the quoted author ] [ lots of dumb vulgarities addressed to the person who forgot more about computers than Hugh knows now snipped ] PLONK! [ When replying, remove *'s from address ] Alexandre Pechtchanski, Systems Manager, RUH, NY ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Mon, 23 Nov 1998 20:26:06 GMT Organization: . Lines: 24 Message-ID: <3659aa35.675311@news.uunet.be> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <365691f6.19845816@news.ricochet.net> <737cc9$e8s@jamesa.cstone.net> <3659f9f9.2389085@news.uunet.be> <766.630T2695T8146524@sky.bus.com> <73ampk$im3$1@news.igs.net> NNTP-Posting-Host: pool02b-194-7-146-160.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!diablo.theplanet.net!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach "Donald Tees" on Sun, 22 Nov 1998 23:08:32 -0500 to alt.folklore.computers: > When it reaches the state that only programmers can use > a computer, then it is time for all programmers to go back > to school and learn to do it right. Actually, it seems to be getting better over the years. I do remember days when I personally believed a computer would be completely useless to someone not capable of programming, and when there were more programmers that seemed to join me in that opinion than there are today. I also remember being young (that's under 20) in those days, and working more in assembler than in higher languages (what, I started out on a KIM-1 in pure hex!). And that the banks (at least the small offices I went to) didn't have computers or terminals yet. Double age has brought along double wisdom, in that aspect (too bad it's not in every aspect ;-) -- My undergraduate experience convinced me that I was not smart enough to be a physicist, and that computers were quite neat. Dennis M. Ritchie ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Mon, 23 Nov 1998 20:26:11 GMT Organization: . Lines: 41 Message-ID: <365bc192.6656191@news.uunet.be> References: <3656d249.334310@news.uunet.be> <736od5$h6j@axalotl.demon.co.uk> NNTP-Posting-Host: pool02b-194-7-146-160.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!diablo.theplanet.net!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach huge@nospam.demon.co.uk (Hugh Davies) on 21 Nov 1998 16:05:57 GMT to alt.folklore.computers: > The wheel just got re-invented over and over and over and over again. And just like in real life, the result is that it's no longer a solid circle of rock or wood, but there are different models up to a spoke wheel with an alu rim and low profile high speed tires [brands not named, but it's like with good wine: from a good year is better] (1). Prettier than they were in the dark ages, but more work to keep them clean, and you have to watch out not to run into any nails on the road (or did you think all those megabytes are _only_ the result of bad programming habits?) I don't have the necessary docs to seek it out (nor the time it would take to count them), but I'm a bit curious about the number (and complexity) of API calls in different OSes. ======== (1) While I'm using analogies anyway: trucks like unix have more solid and less good looking wheels than a limmo, of course. They can also carry more load, at the expense of not being as luxurious. And they have the reputation of lasting many more miles. But I don't think when you call fopen(), the library passes this through to the kernel as a call with 9 arguments. (2)(3) (2) 7, of which 1 is a pointer to a structure with 3 members, so I count 9 values - most of them optional, but the code to handle them (in all their possible values and combinations) has to be there. (3) Maybe I should have picked a call that's less of a workhorse. Unix just might be able to handle a lot of data there too. -- "Managing senior programmers is like herding cats." --Dave Platt ###### From: Ingvar Mattsson Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 24 Nov 1998 11:41:35 +0100 Organization: Telia Internet Services Lines: 22 Message-ID: References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <365691f6.19845816@news.ricochet.net> <737cc9$e8s@jamesa.cstone.net> <3659f9f9.2389085@news.uunet.be> <766.630T2695T8146524@sky.bus.com> <73bsql$r91@top.mitre.org> NNTP-Posting-Host: d2o61.telia.com X-Newsreader: Gnus v5.5/Emacs 20.2 X-NNTP-Posting-Host: ns.idasys.se X-Complaints-To: abuse@internet.telia.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newshub.northeast.verio.net!news-feed.inet.tele.dk!bofh.vszbr.cz!masternews.telia.net!newspost.telia.com!d2o61.telia.com!ns.idasys.se jcmorris@mwunix.mitre.org (Joe Morris) writes: > "Charlie Gibbs" writes: > > >Hear, hear! Remember when the word "turnkey" was used to > >describe such systems? Oddly, I haven't heard that term > >much lately. Stan Kelly-Bootle, in "The Devil's DP Dictionary", > >pointed out that the word "turnkey" is often pronounced with a > >silent "n". > > Where is Kelly-Bootle these days? I've not seen anything from him since > I lost my subscription to _DEC Professional_ (at least that's where I > think his column was published...) He writes a column in Unix Revie^w^wPerformance Computing, named Devil's Advocate. //Ingvar (bats .333) -- Sysadmin, disgruntled, unpolite. I don't speak for my employer nor do they speak for me. Accept this and life will be easier. ingvar@idasys.se ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 24 Nov 1998 20:05:09 GMT Lines: 28 Message-ID: <73f3hl$f8l$2@irk.zetnet.co.uk> References: <766.630T2695T8146524@sky.bus.com> NNTP-Posting-Host: man-025.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 911937909 15637 194.247.41.31 (24 Nov 1998 20:05:09 GMT) NNTP-Posting-Date: 24 Nov 1998 20:05:09 GMT X-Everything: Net-Tamer V 1.08X Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newscore.univie.ac.at!btnet-peer!btnet-feed1!btnet!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-11-22 cgibbs@sky.bus.com said: :(Luc Van der Veken) writes: :>Companies that sell complete systems with the software for :>applications like the above pre-installed, and train their :>customers in the use of _just_ that application software, :>are a danger to society, and should be outlawed. :Hear, hear! Remember when the word "turnkey" was used to :describe such systems? Oddly, I haven't heard that term :much lately. Stan Kelly-Bootle, in "The Devil's DP Dictionary", :pointed out that the word "turnkey" is often pronounced with a :silent "n". I must chime in here again, having made a comment that originally looked like "people should know how to program before using a computer". I didn't actually mean to go that far, but I think that just as people should have a basic idea of how a car works before they buy said car, they should have an idea of how a computer works before buying that too. And the reason for that position, for me, is because a little knowledge may be a dangerous thing, but a little more will probably stop you from getting soundly ripped off. I think MS are trading on lack of knowledge to rip people off. (There. I've said it explicitly; it saves anyone trying to read it in. ;> ) -- Communa (lisard@zetnet.co.uk) -- you know soft spoken changes nothing ###### From: Mike Swaim Subject: Re: Multics Newsgroups: alt.folklore.computers References: <766.630T2695T8146524@sky.bus.com> <73f3hl$f8l$2@irk.zetnet.co.uk> User-Agent: tin/pre-1.4-980226 (UNIX) (FreeBSD/2.2.6-RELEASE (i386)) Lines: 18 Message-ID: Date: Tue, 24 Nov 1998 20:58:17 GMT NNTP-Posting-Host: 209.127.0.130 X-Trace: news2.giganews.com 911941097 209.127.0.130 (Tue, 24 Nov 1998 14:58:17 CDT) NNTP-Posting-Date: Tue, 24 Nov 1998 14:58:17 CDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.cwix.com!207.207.0.26!nntp.giganews.com!news2.giganews.com.POSTED!gemini.c-com.net!swaim lisard@zetnet.co.uk wrote: : I must chime in here again, having made a comment that originally looked : like "people should know how to program before using a computer". I : didn't actually mean to go that far, but I think that just as people : should have a basic idea of how a car works before they buy said car, : they should have an idea of how a computer works before buying that too. I know several otherwise intelligent people who've badly damaged cars because they didn't know about things like changing the oil. (Actually, one guy didn't even know that cars needed any other fluid than gas.) I've also seen/heard about people with just enough computer knowledge to be dangerous. I'd say that people should know enough to do their job, basic maintenance, and to leave the rest alone. -- Mike Swaim, Avatar of Chaos: Disclaimer:I sometimes lie. Home: swaim@c-com.net Alum: swaim@alumni.rice.edu Quote: "Boingie"^4 Y,W&D ###### From: p98mccabe@aol.com (P98McCabe) Newsgroups: alt.folklore.computers Subject: Re: Multics Lines: 26 NNTP-Posting-Host: ladder03.news.aol.com X-Admin: news@aol.com Date: 25 Nov 1998 16:16:57 GMT Organization: AOL http://www.aol.com References: Message-ID: <19981125111657.07068.00000427@ng147.aol.com> Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news.tele2.nl!newsfeed1.swip.net!swipnet!news.idt.net!news-peer.gip.net!news.gsl.net!gip.net!portc01.blue.aol.com!audrey03.news.aol.com!not-for-mail swaim@gemini.c-com.net said: >I've also seen/heard about people with just enough computer knowledge >to be dangerous. I'd say that people should know enough to do their >job, basic maintenance, and to leave the rest alone. I guess I'm somebody with "just enough" to be dangerous - but where do you draw the line? I know a whole lot about my Apple II and Vic-20 computers - having torn them apart, repaired them, modified them, programmed them, crashed them, and generally played with them since around 1979 (In the case of the Apple II). Windows scares the SH** out of me. As the local "Guru", I'm often the one called to resucitate them when they no longer respond appropriately to a given stimulus. Usually, I can make them work - even though I don't "know anything" about them (Heck, I don't even use one!) There's something to be said for a good set of heuristics and "flying by the seat of your pants." p98mccabe@aol.com Micheal H. McCabe P.S. Another 'Clueless AOL user.' ###### From: Mike Swaim Subject: Re: Multics Newsgroups: alt.folklore.computers References: <19981125111657.07068.00000427@ng147.aol.com> User-Agent: tin/pre-1.4-980226 (UNIX) (FreeBSD/2.2.6-RELEASE (i386)) Lines: 37 Message-ID: Date: Wed, 25 Nov 1998 17:36:13 GMT NNTP-Posting-Host: 209.127.0.130 X-Trace: news2.giganews.com 912015373 209.127.0.130 (Wed, 25 Nov 1998 11:36:13 CDT) NNTP-Posting-Date: Wed, 25 Nov 1998 11:36:13 CDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!dca1-hub1.news.digex.net!digex!nntp.giganews.com!news2.giganews.com.POSTED!gemini.c-com.net!swaim P98McCabe wrote: : swaim@gemini.c-com.net said: :>I've also seen/heard about people with just enough computer knowledge :>to be dangerous. I'd say that people should know enough to do their :>job, basic maintenance, and to leave the rest alone. : I guess I'm somebody with "just enough" to be dangerous - but where do you : draw the line? I suspect that you actually know enough to past the dangerous level. : I know a whole lot about my Apple II and Vic-20 computers - having torn : them apart, repaired them, modified them, programmed them, crashed them, and : generally played with them since around 1979 (In the case of the Apple II). No smoke? You didn't blow any cards? Then you were relatively tame. : Windows scares the SH** out of me. As the local "Guru", I'm often the : one called to resucitate them when they no longer respond appropriately : to a given stimulus. Usually, I can make them work - even though I don't : "know anything" about them (Heck, I don't even use one!) But you probably know enough to not screw things up even worse. The people I'm thinking of have been known to to do things like try different hardware settings on the hope that it may fix things, or reset a server because a printer's not working. : There's something to be said for a good set of heuristics and "flying by : the seat of your pants." By the time you get to a good set of heuristics you're at least somewhat clueful. : P.S. Another 'Clueless AOL user.' You guys replaced the Delphoids, didn't you? -- Mike Swaim, Avatar of Chaos: Disclaimer:I sometimes lie. Home: swaim@c-com.net Alum: swaim@alumni.rice.edu Quote: "Boingie"^4 Y,W&D ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Wed, 25 Nov 1998 20:20:21 GMT Organization: . Message-ID: <365e656e.7810160@news.uunet.be> References: <766.630T2695T8146524@sky.bus.com> <73f3hl$f8l$2@irk.zetnet.co.uk> NNTP-Posting-Host: pool02b-194-7-146-218.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Lines: 52 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!news-feed1.eu.concert.net!btnet-peer!btnet!dispose.news.demon.net!demon!rill.news.pipex.net!pipex!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach lisard@zetnet.co.uk on 24 Nov 1998 20:05:09 GMT to alt.folklore.computers: > I think MS are trading on lack of knowledge to rip people off. (There. > I've said it explicitly; it saves anyone trying to read it in. ;> ) That's something I won't try to deny. But just like with everything, they didn't invent it. Apple did it first (of the two that is, they in turn may have seen it somewhere else. I don't have PARC in mind here - I'm talking about marketing, not product.) With the MAC, Apple did a much better job at it too. The main reason for its success was that you _really_ didn't have to know anything about a computer, so most buyers never noticed that it wasn't good for anything else but writing fancy documents and some simple database and spreadsheet work (I suppose it's not by accident that that was just the kind of application those users needed most, OTOH - better do good what's most important, than do everything, but do it flaky). If you believe the users, Apple did a better job all over: does that mean it was less of a rip-off, or does it mean that its average user is [was] even less computer-literate? (to Mac users considering the posting of a hot flame: I'm not talking about _you_ , but about the _average_ user ;-) A point of difference to think about: the MAC (at least in its early days, I didn't quite follow the evolution later) was often said to discourage the user from programming. MS instead are building an easy language (VB) into just about every major product they sell (word, access, excel, and through the scripting engine you can download for free, Windows itself). Doesn't this indicate that MS wants to give its illiterate users the opportunity to become a little programming-aware, where Apple tried to keep them dumb? === [I was talking about the first MAC's - hence the past tense - because I never looked at the more recent ones] [And no, I'm not comparing the first MAC's to today's Wintel machines. I'm comparing their target audiences, and how marketing looks/ed at them.] [And yes, this is *my* opinion. I'll neither shoot nor flame you if yours differs] -- "Managing senior programmers is like herding cats." --Dave Platt ###### From: gerglery@usa.net (Fluffy) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 25 Nov 1998 20:45:36 GMT Organization: Ruler of Usenet (all) Message-ID: <73hq9g$gsd$1@meowhost.meow.invalid> References: <766.630T2695T8146524@sky.bus.com> <73f3hl$f8l$2@irk.zetnet.co.uk> <365e656e.7810160@news.uunet.be> X-Trace: mail2news.demon.co.uk 912026755 mail2news:23487 mail2news mail2news.demon.co.uk X-Complaints-To: abuse@demon.net X-Mail2News-Path: news.demon.net!out2.ibm.net!meowhost.meow.invalid X-P-Meow: Meow Mail-Copies-To: never X-URL: http://members.tripod.com/~gerglery/ X-Newsreader: slrn (0.9.5.3-canlock UNIX) Lines: 26 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news-fra1.dfn.de!news-fra.maz.net!news-hh.maz.net!newsfeed.nacamar.de!dispose.news.demon.net!demon!news.demon.co.uk!demon!mail2news.demon.co.uk!not-for-mail Luc Van der Veken wrote: > A point of difference to think about: the MAC (at least in its > early days, I didn't quite follow the evolution later) was often > said to discourage the user from programming. MS instead are > building an easy language (VB) into just about every major > product they sell (word, access, excel, and through the scripting > engine you can download for free, Windows itself). > Doesn't this indicate that MS wants to give its illiterate users > the opportunity to become a little programming-aware, where Apple > tried to keep them dumb? Nah, I don't buy that one. The Windows development options of the time were just as painful. The saving grace was that one could fall back to the DOS and get real work done, rather than fussing around with event loops and all that fun stuff. Just as MS has been working on VB-izing its world, the Apple folks have been putting in hooks for their scripting (applescript, Frontier, perl, etc.). So, each GUI is still on more or less equal footing when it comes to programmability. Where there is a night-and-day difference is in the marketing. MS isn't afraid to brag about its scripting toys, but Apple _still_ seems to be terrified of letting its customers find out that you can do hard stuff if you want =) ###### From: Mike Swaim Subject: Re: Multics Newsgroups: alt.folklore.computers References: <766.630T2695T8146524@sky.bus.com> <73f3hl$f8l$2@irk.zetnet.co.uk> <365e656e.7810160@news.uunet.be> User-Agent: tin/pre-1.4-980226 (UNIX) (FreeBSD/2.2.6-RELEASE (i386)) Lines: 21 Message-ID: <9U072.31715$gI1.5820@news2.giganews.com> Date: Thu, 26 Nov 1998 00:02:13 GMT NNTP-Posting-Host: 209.127.0.130 X-Trace: news2.giganews.com 912038533 209.127.0.130 (Wed, 25 Nov 1998 18:02:13 CDT) NNTP-Posting-Date: Wed, 25 Nov 1998 18:02:13 CDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.cwix.com!207.207.0.26!nntp.giganews.com!news2.giganews.com.POSTED!gemini.c-com.net!swaim Luc Van der Veken wrote: : A point of difference to think about: the MAC (at least in its : early days, I didn't quite follow the evolution later) was often : said to discourage the user from programming. MS instead are : building an easy language (VB) into just about every major : product they sell (word, access, excel, and through the scripting : engine you can download for free, Windows itself). : Doesn't this indicate that MS wants to give its illiterate users : the opportunity to become a little programming-aware, where Apple : tried to keep them dumb? I believe that Apple did have a BASIC in development when the Mac came out. (Some books for it even came out.) The word on the street was that it got killed at MS's request, because it would compete with Microsoft's BASIC. MS BASIC for the Mac did eventually come out, but it sucked and quickly died. -- Mike Swaim, Avatar of Chaos: Disclaimer:I sometimes lie. Home: swaim@c-com.net Alum: swaim@alumni.rice.edu Quote: "Boingie"^4 Y,W&D ###### From: fleetvis@ricochet.net (phil) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Thu, 26 Nov 1998 12:07:55 GMT Organization: Posted via Supernews, http://www.supernews.com - Discussions start here! Lines: 17 Message-ID: <365d422d.22238927@news.ricochet.net> References: <766.630T2695T8146524@sky.bus.com> <73f3hl$f8l$2@irk.zetnet.co.uk> NNTP-Posting-Host: 204.179.134.69 X-Trace: 912081901 F7W8ZAFJM8645CCB3C usenet57.supernews.com X-Complaints-To: newsabuse@supernews.com X-Newsreader: Forte Free Agent 1.11/32.235 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.berkeley.edu!Supernews73!supernews.com!Supernews69!not-for-mail On 24 Nov 1998 20:05:09 GMT, lisard@zetnet.co.uk wrote: >I must chime in here again, having made a comment that originally looked >like "people should know how to program before using a computer". I >didn't actually mean to go that far, but I think that just as people >should have a basic idea of how a car works before they buy said car, >they should have an idea of how a computer works before buying that too. >And the reason for that position, for me, is because a little knowledge >may be a dangerous thing, but a little more will probably stop you from >getting soundly ripped off. But you've just done the same thing. again. > >I think MS are trading on lack of knowledge to rip people off. (There. >I've said it explicitly; it saves anyone trying to read it in. ;> ) And there ae companies don't operate this way? ###### Message-ID: <365D6B2C.4E051A0E@bellsouth.net> From: Max F Lang Reply-To: mflang@bellsouth.net Organization: Home Computing X-Mailer: Mozilla 4.05 [en] (X11; I; Linux 2.0.34 i586) MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: Multics References: <766.630T2695T8146524@sky.bus.com> <73f3hl$f8l$2@irk.zetnet.co.uk> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Lines: 26 Date: Thu, 26 Nov 1998 14:47:09 GMT NNTP-Posting-Host: host-209-214-32-229.mco.bellsouth.net NNTP-Posting-Date: Thu, 26 Nov 1998 09:47:09 EST Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!news.maxwell.syr.edu!newsfeed.atl!news4.mco.POSTED!not-for-mail Mike Swaim wrote: > lisard@zetnet.co.uk wrote: > : should have a basic idea of how a car works before they buy said car, > I know several otherwise intelligent people who've badly damaged cars > because they didn't know about things like changing the oil. One of those people is Bill Gates. He apparently did in a couple of fancy cars in the '80s because he didn't understand the engines needed oil, and the flashing oil light on the panel meant something. Bill Gates, whatever else can be said about him and his actions, is a very intelligent guy. He saw needs, and did what he felt was required to fill them to the benefit of his company and his wallet. That he's the devil incarnate is another issue... :-) > I've also seen/heard about people with just enough computer knowledge to > be dangerous. I'd say that people should know enough to do their job, I think the quote is "The term idiot-proof underestimates the power of the complete idiot." MFLang, a budding Unix and VMS guru... ###### Date: 26 Nov 98 18:05:24 -0800 From: "Charlie Gibbs" Subject: Re: Multics References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> Message-ID: <900.634T1209T10854550@sky.bus.com> Newsgroups: alt.folklore.computers Lines: 25 X-Newsreader: THOR 2.5a (Amiga;TCP/IP) NNTP-Posting-Host: news.skybus.com X-Trace: 26 Nov 1998 18:12:51 -0800, news.skybus.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!howland.erols.net!news-peer1.sprintlink.net!news-backup-west.sprintlink.net!news.sprintlink.net!204.244.4.2!news.westel.com!news.skybus.com!204.244.247.130 In article <73koo5$lln$1@irk.zetnet.co.uk> lisard@zetnet.co.uk (lisard) writes: >On 1998-11-26 fleetvis@ricochet.net(phil) said: > >:On 24 Nov 1998 20:05:09 GMT, lisard@zetnet.co.uk wrote: >: >:>I think MS are trading on lack of knowledge to rip people off. >:>(There. I've said it explicitly; it saves anyone trying to read it >:in. ;> ) >: >:And there ae companies don't operate this way? > >Caldera? Debian? Red Hat? Corel, possibly? Netscape? Many millions more >companies? Sorry, but the vast majority of companies *don't* organise >their entire business strategy on the twin bases that their users are >(a) naive, (b) stupid. Like us, for instance. At least on the development side. Our tech support department, on the other hand, is reminded daily of just how far the average user _hasn't_ come. -- cgibbs@sky.bus.com (Charlie Gibbs) Remove the first period after the "at" sign to reply. ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Thu, 26 Nov 1998 19:21:44 GMT Organization: . Lines: 17 Message-ID: <365fa4a7.3169146@news.uunet.be> References: <733rdt$h9r$1@ligarius.ultra.net> <7357v1$uiv$2@irk.zetnet.co.uk> <365691f6.19845816@news.ricochet.net> <737cc9$e8s@jamesa.cstone.net> <3659f9f9.2389085@news.uunet.be> NNTP-Posting-Host: pool02b-194-7-145-35.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news.tele2.nl!newsfeed1.swip.net!swipnet!masternews.telia.net!newsfeed.ecrc.net!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail +++Off topic+++ (like everything else this week in this thread) A subthread about users not knowing computers, and then some bozo (that was me) starts making those annoying car analogies again. Somewhere else in the same thread, someone else starts about drivers that don't know a car needs other liquids besides gas. A few days later, in rec.humor.funny.reruns: Message-ID: , or on the web at http://www.netfunny.com/rhf/jokes/94q4/gmhelpline.html Brad Templeton, or Maddi Hausmann, or Jim Griffith, or all three of them, must be reading this group. -- "Managing senior programmers is like herding cats." --Dave Platt ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 26 Nov 1998 23:37:41 GMT Message-ID: <73koo5$lln$1@irk.zetnet.co.uk> References: <365d422d.22238927@news.ricochet.net> NNTP-Posting-Host: man-132.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 912123461 22199 194.247.40.168 (26 Nov 1998 23:37:41 GMT) NNTP-Posting-Date: 26 Nov 1998 23:37:41 GMT X-Everything: Net-Tamer V 1.08X Lines: 41 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!dispose.news.demon.net!demon!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-11-26 fleetvis@ricochet.net(phil) said: :Newsgroups: alt.folklore.computers :On 24 Nov 1998 20:05:09 GMT, lisard@zetnet.co.uk wrote: :>I must chime in here again, having made a comment that originally :>looked like "people should know how to program before using a :>computer". I didn't actually mean to go that far, but I think that :>just as people should have a basic idea of how a car works before :>they buy said car, they should have an idea of how a computer :>works before buying that too. And the reason for that position, :>for me, is because a little knowledge may be a dangerous thing, :>but a little more will probably stop you from getting soundly :ripped off. :But you've just done the same thing. again. No I haven't. What makes you think I did? Knowing how a computer works is not the same as knowing how to program it. Not by a long way. And I'm not talking about bits, bytes, and assembler coding, for heaven's sake! Just basic things like what a "file format" is, what "proprietary" really means, the basic currency of information on the web, how to find out how to do things, the principles of multitasking, maybe a few commands from the command line (just in case everything goes pear-shaped), the concept of a hierarchical filing system, protection. Basic concepts that will get used every day. Most importantly, that there are alternatives, that what you get free is not always or necessarily the best way to do something, how valuable software is, etc. :>I think MS are trading on lack of knowledge to rip people off. :>(There. I've said it explicitly; it saves anyone trying to read it :in. ;> ) :And there ae companies don't operate this way? Caldera? Debian? Red Hat? Corel, possibly? Netscape? Many millions more companies? Sorry, but the vast majority of companies *don't* organise their entire business strategy on the twin bases that their users are (a) naive, (b) stupid. -- Communa (lisard@zetnet.co.uk) -- you know soft spoken changes nothing ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 26 Nov 1998 23:37:43 GMT Message-ID: <73koo7$lln$2@irk.zetnet.co.uk> References: <365e656e.7810160@news.uunet.be> NNTP-Posting-Host: man-132.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 912123463 22199 194.247.40.168 (26 Nov 1998 23:37:43 GMT) NNTP-Posting-Date: 26 Nov 1998 23:37:43 GMT X-Everything: Net-Tamer V 1.08X Lines: 28 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!dispose.news.demon.net!demon!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-11-25 lucvdv@null.net(LucVanderVeken) said: :A point of difference to think about: the MAC (at least in its :early days, I didn't quite follow the evolution later) was often :said to discourage the user from programming. MS instead are Well, except for MacForth, which was released very early into the Mac's lifespan and for quite a while was the only development environment you could fit inside a 128k Mac. (Naturally. ;> ) :building an easy language (VB) into just about every major :product they sell (word, access, excel, and through the scripting :engine you can download for free, Windows itself). Easy and crap. Frankly, I'd rather not have to worry. Besides which, Apple had HyperCard, which (I believe) fulfilled about the same function. And I don't believe Word 2 had even WordBasic, let alone VB. Besides which, you aren't really comparing apples with apples (ahem). The Mac was designed as a complete appliance, not (so much) as a general purpose computer. It was explicitly designed for people who weren't programmers, and weren't going to program it. That's fair enough; they weren't presenting this thing as a general solution and seeking to annex every single box in the world with it. MS are, and their product isn't up to it, either. That's worrying, and crooked. -- Communa (lisard@zetnet.co.uk) -- you know soft spoken changes nothing ###### From: Duane Hentrich Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Fri, 27 Nov 1998 13:33:58 -0800 Organization: DHL Systems, Inc. Lines: 8 Message-ID: <365F1AC6.55FF56AF@systems.dhl.com> References: <73mlqs$2ra$1@news.seed.net.tw> NNTP-Posting-Host: 199.41.203.136 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 4.06 [en] (X11; U; HP-UX B.10.20 9000/778) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news.shore.net!uunet!in1.uu.net!not-for-mail Dan Strychalski wrote: > Yes, YES, *Y*E*S*. Happiness is a waiting command prompt. And nirvana is a long row of toggle switches under Der Blinken Lights. -- Hypertext isn't what it's marked up to be. | dhentric@systems.dhl.com ###### From: Dan Strychalski Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 27 Nov 1998 17:00:12 GMT Organization: Cameo Communications, Inc. Lines: 20 Message-ID: <73mlqs$2ra$1@news.seed.net.tw> NNTP-Posting-Host: 192.72.104.4 Originator: dski@ Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newscore.univie.ac.at!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news-peer.gip.net!news.gsl.net!gip.net!spring.edu.tw!feeder.seed.net.tw!news.seed.net.tw!!dski Communa expostulated -- > [...] Knowing how a computer works is not the same as knowing how to > program it. Not by a long way. And I'm not talking about bits, bytes, and > assembler coding, for heaven's sake! [...] I don't seem to have the particular talent set needed to make a programmer, but bits and bytes were the first things that "clicked" for me when I was trying to understand how a computer works. I would include them among the basic things learners need to be exposed to close to the beginning. The other stuff you mention -- right on the money. These are necessary concepts, and not at all hard to grasp when presented right. > Just basic things like what a "file format" is, what "proprietary" > really means, ...maybe a few commands from the command line.... Yes, YES, *Y*E*S*. Happiness is a waiting command prompt. Dan Strychalski dski at cameonet, cameo, com, tw (no _x_) Apologies for the anti-spam devices and non-threading newsreader. ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Fri, 27 Nov 1998 20:50:45 GMT Organization: . Lines: 56 Message-ID: <36600eec.6909655@news.uunet.be> References: <365e656e.7810160@news.uunet.be> <73koo7$lln$2@irk.zetnet.co.uk> NNTP-Posting-Host: pool02b-194-7-150-161.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!isdnet!newsfeed.icl.net!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach lisard@zetnet.co.uk on 26 Nov 1998 23:37:43 GMT to alt.folklore.computers: > Well, except for MacForth, which was released very early into the Mac's > lifespan By Apple? (I really don't know) [VB] > Easy and crap. As for VBScript, true. VBS in itself isn't even that bad - if you keep in mind that for simple things like copying a few files you better write a .cmd or a .bat instead of a .vbs - it's the documentation that really stinks (with 2 different online help systems available, publishing a reference work as a word document isn't my idea of a good job.) But look at it like the people it's aimed at, not like a seasoned programmer. VB is easy to learn, which isn't exactly true for Forth (at least they told me it was hard before I started with it - actually I found learning it not that hard, doing something useful in it just took too much work - but I'm talking about [obafc] TRS-80 Forth now, I suppose today's versions don't require you to reinvent the wheel every day anymore.) Maybe you'd better have thought of HyperCard here, instead of further on (btw, I forgot all about it when I wrote that post - so I'd better admit here that I wasn't very right about programming possibilities). As for the full VB, it's not that crappy. I used to think "I'll prototype that user interface in VB and then rewrite it in VC when I'm satisfied about looks and functionality" - but after I found out that for those things (user interfaces) it offers ten times the ease of programming and debugging versus no worthwhile difference in speed and size (the VB runtime is slightly larger than the MFC one - both :( - and its executables are actually smaller), I just kept it in VB, and do only the real work in a DLL written in C or C++. > Frankly, I'd rather not have to worry. Besides which, > Apple had HyperCard, which (I believe) fulfilled about the same > function. And I don't believe Word 2 had even WordBasic, let alone VB. That's true (as far as I know HyperCard - which is all hear-say and magazine-read). > Besides which, you aren't really comparing apples with apples (ahem). Compare 2 items, and I'll tell you tomorrow how many posts on dejanews say they aren't comparable :-) -- "Managing senior programmers is like herding cats." --Dave Platt ###### From: fleetvis@ricochet.net (phil) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sat, 28 Nov 1998 11:41:01 GMT Organization: Posted via Supernews, http://www.supernews.com - Discussions start here! Lines: 57 Message-ID: <365fd57c.3293716@news.ricochet.net> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> NNTP-Posting-Host: 204.179.136.64 X-Trace: 912252718 F7W8ZAFJM8840CCB3C usenet54.supernews.com X-Complaints-To: newsabuse@supernews.com X-Newsreader: Forte Free Agent 1.11/32.235 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.berkeley.edu!Supernews73!supernews.com!Supernews69!not-for-mail On 26 Nov 1998 23:37:41 GMT, lisard@zetnet.co.uk wrote: >On 1998-11-26 fleetvis@ricochet.net(phil) said: > :But you've just done the same thing. again. > >No I haven't. What makes you think I did? Knowing how a computer works >is not the same as knowing how to program it. Not by a long way. And I'm >not talking about bits, bytes, and assembler coding, for heaven's sake! >Just basic things like what a "file format" is, what "proprietary" >really means, the basic currency of information on the web, how to find >out how to do things, the principles of multitasking, maybe a few >commands from the command line (just in case everything goes >pear-shaped), the concept of a hierarchical filing system, protection. >Basic concepts that will get used every day. Most importantly, that >there are alternatives, that what you get free is not always or >necessarily the best way to do something, how valuable software is, etc. What i meant to say was that you have imposed an even more onerous burden on the user. No longer do they have to know how to use a computer, but they have to know how a computer works. There are a surprising number of programmers that have only a vague idea of how a computer works and that, for example, couldn't write an interrupt service to save their lives, and why should they? Everyone uses stuff that they don't understand. > :>I think MS are trading on lack of knowledge to rip people off. > :>(There. I've said it explicitly; it saves anyone trying to read it > :in. ;> ) > > :And there ae companies don't operate this way? > >Caldera? Debian? Red Hat? Corel, possibly? Netscape? Many millions more >companies? Sorry, but the vast majority of companies *don't* organise >their entire business strategy on the twin bases that their users are >(a) naive, (b) stupid. If a company has a large market share or is in a large market, then the vast majority of companies *do* organise their business strategy on the basis that their potential consumers are both naive and stupid. This is what advertising is for. For high-tech companies selling to high-tech consumers, there is a good argument for giving the impression that the customers are not stupid, until those companies move into the mass consumer market or position themselves for so doing. The more successful a company is, the more stupid it thinks its customers are.. Compare the Microsoft "Where do you want to go today?" with the Cisco (is it cisco?) "Are you ready?". Same shit. Different company. What about the Apple ads? Scary. I think iIt's disingenious to overestimate the knowledge of the general public (and that includes me). For example,the Un*x sound bite could be: "rm -r * - you know it makes sense" and i'd buy it. phil ###### From: genew@vip.net (Gene Wirchenko) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sun, 29 Nov 1998 00:14:36 GMT Organization: Posted via Supernews, http://www.supernews.com - Discussions start here! Lines: 46 Message-ID: <366056e7.1022386@news.vip.net> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> Reply-To: genew@vip.net NNTP-Posting-Host: 204.209.212.52 X-Trace: 912298216 A01OARAUVD434CCD1C usenet76.supernews.com X-Complaints-To: newsabuse@supernews.com X-Newsreader: Forte Free Agent 1.1/32.230 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.berkeley.edu!Supernews73!supernews.com!Supernews69!not-for-mail fleetvis@ricochet.net (phil) wrote: >On 26 Nov 1998 23:37:41 GMT, lisard@zetnet.co.uk wrote: > >>On 1998-11-26 fleetvis@ricochet.net(phil) said: > >> :But you've just done the same thing. again. >> >>No I haven't. What makes you think I did? Knowing how a computer works >>is not the same as knowing how to program it. Not by a long way. And I'm >>not talking about bits, bytes, and assembler coding, for heaven's sake! >>Just basic things like what a "file format" is, what "proprietary" >>really means, the basic currency of information on the web, how to find >>out how to do things, the principles of multitasking, maybe a few >>commands from the command line (just in case everything goes >>pear-shaped), the concept of a hierarchical filing system, protection. >>Basic concepts that will get used every day. Most importantly, that >>there are alternatives, that what you get free is not always or >>necessarily the best way to do something, how valuable software is, etc. > >What i meant to say was that you have imposed an even more onerous >burden on the user. No longer do they have to know how to use a >computer, but they have to know how a computer works. There are a >surprising number of programmers that have only a vague idea of how a >computer works and that, for example, couldn't write an interrupt >service to save their lives, and why should they? Everyone uses stuff >that they don't understand. It isn't necessarily a matter of not understanding. If I had to write an ISR, I could probably figure it out. Why don't I? Because I'm generally an applications programmer/analyst. Let someone else take the time to learn how to write ISRs as he'll be writing far more than I will. Translation: Assuming I can get an ISR from some[where|one] else, I'm better off spending my time in my area of expertise. [snip] Sincerely, Gene Wirchenko Computerese Irregular Verb Conjugation: I have preferences. You have biases. He/She has prejudices. ###### From: "Andrew Cadley" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sun, 29 Nov 1998 21:09:22 -0000 Organization: University of East Anglia, Norwich, Norfolk, NR47TJ, UK Lines: 22 Message-ID: <73sd9a$p8s@cpca3.uea.ac.uk> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <366056e7.1022386@news.vip.net> NNTP-Posting-Host: oak6d.vlg1.uea.ac.uk X-Newsreader: Microsoft Outlook Express 4.72.2120.0 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2120.0 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!rill.news.pipex.net!pipex!server1.netnews.ja.net!cpca3.uea.ac.uk!news Gene Wirchenko wrote in message <366056e7.1022386@news.vip.net>... :fleetvis@ricochet.net (phil) wrote: : It isn't necessarily a matter of not understanding. If I had to :write an ISR, I could probably figure it out. Why don't I? Because :I'm generally an applications programmer/analyst. Let someone else :take the time to learn how to write ISRs as he'll be writing far more :than I will. Translation: Assuming I can get an ISR from :some[where|one] else, I'm better off spending my time in my area of :expertise. Which is precisely the point, I think. You don't need to go away and work out how to write an ISR, so why should any user need to know anything more than how to use the software they need? They are better off spending their time in *their* area of expertise, be it graphics, accountancy, designing widgets or whatever. AndyC ###### From: Richard Lamb Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sun, 29 Nov 1998 23:16:44 -0600 Organization: Tapestry Software Lines: 23 Message-ID: <36622A3C.72A3F80E@flash.net> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <366056e7.1022386@news.vip.net> <73sd9a$p8s@cpca3.uea.ac.uk> Reply-To: lamb01@flash.net NNTP-Posting-Host: p126.amax3.dialup.sat1.flash.net Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Sender: "Richard Lamb" (Unverified) X-Mailer: Mozilla 4.04 [en]C-FLASHNET (Win95; I) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!isdnet!howland.erols.net!nntp.flash.net!excalibur.flash.net!not-for-mail About what I'd expect from a C+- coder, hmmm? Andrew Cadley wrote: > > Gene Wirchenko wrote in message <366056e7.1022386@news.vip.net>... > :fleetvis@ricochet.net (phil) wrote: > > : It isn't necessarily a matter of not understanding. If I had to > :write an ISR, I could probably figure it out. Why don't I? Because > :I'm generally an applications programmer/analyst. Let someone else > :take the time to learn how to write ISRs as he'll be writing far more > :than I will. Translation: Assuming I can get an ISR from > :some[where|one] else, I'm better off spending my time in my area of > :expertise. > > Which is precisely the point, I think. You don't need to go away and work > out how to write an ISR, so why should any user need to know anything more > than how to use the software they need? They are better off spending their > time in *their* area of expertise, be it graphics, accountancy, designing > widgets or whatever. > > AndyC ###### From: Richard Lamb Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sun, 29 Nov 1998 23:20:25 -0600 Organization: Tapestry Software Lines: 3 Message-ID: <36622B19.A621FDD0@flash.net> References: <73mlqs$2ra$1@news.seed.net.tw> <365F1AC6.55FF56AF@systems.dhl.com> Reply-To: lamb01@flash.net NNTP-Posting-Host: p126.amax3.dialup.sat1.flash.net Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Sender: "Richard Lamb" (Unverified) X-Mailer: Mozilla 4.04 [en]C-FLASHNET (Win95; I) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newscore.univie.ac.at!howland.erols.net!nntp.flash.net!excalibur.flash.net!not-for-mail But I really LIKE assembly! Anybody want a peek at my Visual Assembler? ###### From: Mike Swaim Subject: Re: Multics Newsgroups: alt.folklore.computers References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> User-Agent: tin/pre-1.4-980226 (UNIX) (FreeBSD/2.2.6-RELEASE (i386)) Lines: 27 Message-ID: Date: Mon, 30 Nov 1998 05:08:27 GMT NNTP-Posting-Host: 209.127.0.130 X-Trace: news1.giganews.com 912402507 209.127.0.130 (Sun, 29 Nov 1998 23:08:27 CDT) NNTP-Posting-Date: Sun, 29 Nov 1998 23:08:27 CDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!isdnet!newsfeed.cwix.com!207.207.0.26!nntp.giganews.com!news1.giganews.com.POSTED!gemini.c-com.net!swaim phil wrote: : What i meant to say was that you have imposed an even more onerous : burden on the user. No longer do they have to know how to use a : computer, but they have to know how a computer works. To a certain extent, yes, just as there's a lot more to operating a car than just knowing how to drive. (And that's with automobile design being a lot more mature than computer design, and most people having their cars checked at least twice a year. How many people have a professional look over their computers twice a year?) : There are a : surprising number of programmers that have only a vague idea of how a : computer works and that, for example, couldn't write an interrupt : service to save their lives, and why should they? And they tend to write bad programs. I remember someone in this very newsgroup posting many years ago that all professional programmers should be required to write and debug at least one program in assembler, so they'd have some understanding of how the computer really worked. Then when I got around to taking the intro computing class at school, a portion of the class was taught in FORTRAN for similar reasons. I happen to agree with both of them that if you're going to be making a living programming computers, you should have an idea of how they work. -- Mike Swaim, Avatar of Chaos: Disclaimer:I sometimes lie. Home: swaim@c-com.net Alum: swaim@alumni.rice.edu Quote: "Boingie"^4 Y,W&D ###### From: jcmorris@mwunix.mitre.org (Joe Morris) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 30 Nov 1998 15:16:04 GMT Organization: The MITRE Corporation Lines: 76 Message-ID: <73ucrk$gg4@top.mitre.org> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> NNTP-Posting-Host: mwunix.mitre.org Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.cwix.com!204.97.128.3!usenet.logical.net!news.tufts.edu!blanket.mitre.org!news.mitre.org!mwunix!jcmorris Mike Swaim writes: >phil wrote: >: There are a >: surprising number of programmers that have only a vague idea of how a >: computer works and that, for example, couldn't write an interrupt >: service to save their lives, and why should they? > And they tend to write bad programs. I remember someone in this very >newsgroup posting many years ago that all professional programmers should >be required to write and debug at least one program in assembler, so >they'd have some understanding of how the computer really worked. Then >when I got around to taking the intro computing class at school, a portion >of the class was taught in FORTRAN for similar reasons. I happen to agree >with both of them that if you're going to be making a living programming >computers, you should have an idea of how they work. While there is a legitimate range of opinions about how much an application programmer needs to know about the underlying details of a computer, I'll strongly agree that anyone with pretentions of working in *any* field needs to be at least somewhat familiar with what's going on below the level where he or she is working. In the computer field this doesn't mean that a business application programmer writing an A/P package in COBOL needs to be able to design the mask for a new Pentium chip, but that same programmer should be more than passingly familiar with the organization of a computer and how the application program interacts with it. In the days between the end of the era when programmers were actually at the console of the (mainframe) computer when their program ran, and the arrival of online systems with interactive user access, there was a major problem caused by application programmers who were apparently incapable of any understanding of just how thoroughly a badly-written computer program could trash computer performance. When running on a single-user system (such as IBSYS or FMS) there was a self-correcting mechanism: the user was charged by wall-clock time without regard to the number of CPU cycles used, so an inefficient program would generate huge bills -- something that was difficult for management to ignore, unless it was on a no-charge account. On a multi-user system, however, where the billing was based on CPU time and maybe the number of I/O operations, it was absurdly easy for an I/O-bound task to wreck performance of the entire system without providing any feedback to the end user whose program was responsible for the trouble. Two war stories on this subject, both of which occurred in my shop, which at the time was a large state university: * An application programmer was tasked to extract random entries from the student master file. This was on IBSYS, on a machine with no disk so the only bulk storage was 1/2" magnetic tape on 125 ips drives. Knowing the number of entries on the tape, his program would: 1) Generate a random number in [0,1] 2) Multiply the random number by the number of entries on the tape 3) Space down the tape that number of entries 4) Read the next entry 5) Rewind the tape 6) Go back to step 1 * On a 360/40 running MFT: a USDA-distributed program named SEMIS (State Extension Management Information System) was used to collect data from the agricultural extension service offices. It was written in basic COBOL with the admirable intent of being usable on just about any small system, but that meant that it had no ability for other than sequential access to files. Like the example above it spent most of its time chugging through files (at least they were on disk!), and on the 360/40 (2 usec memory cycle for 2 bytes per cycle) this program ran a wallclock-to-CPU time ratio of 3600:1. Only when the personal computer restored the "programmer as operator" mode did many of the application programmers begin to appreciate the interaction of program design and computer architecture -- and of course some of the programmers still don't show any understanding of the issues. Joe Morris ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 1 Dec 1998 19:43:37 GMT Message-ID: <741gt9$88i$4@irk.zetnet.co.uk> References: <365fd57c.3293716@news.ricochet.net> NNTP-Posting-Host: man-141.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 912541417 8466 194.247.40.180 (1 Dec 1998 19:43:37 GMT) NNTP-Posting-Date: 1 Dec 1998 19:43:37 GMT X-Everything: Net-Tamer V 1.08X Lines: 78 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!dispose.news.demon.net!demon!demeter.clara.net!news.clara.net!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-11-28 fleetvis@ricochet.net(phil) said: :What i meant to say was that you have imposed an even more onerous :burden on the user. No longer do they have to know how to use a :computer, but they have to know how a computer works. There are a :surprising number of programmers that have only a vague idea of how :a computer works and that, for example, couldn't write an interrupt :service to save their lives, and why should they? Everyone uses :stuff that they don't understand. Yes, but don't they know what an interrupt *is*? Besides, most programmers I know do tend to have some idea of how a computer works, and the ones that don't don't tend to be terrific programmers. (Try writing a VB program without using events, for example.) In any case, what I mean is that people shouldn't regard buying a computer as a decision in the same vein as buying (say) a washing machine or a microwave oven. Or even a cellphone. There are all kinds of variables that it is difficult to grasp, and whilst people would shrink from buying a microwave without understanding what they were getting, they seem to show no such reticence from buying a computer. Frankly, that's silly. Nobody who doesn't want to should know assembler, certainly. But everyone who buys a computer should know about file systems, because they'll be using one every day - certainly when they get past the "This is my first letter in

REALLY BIG FONTS

" stage. And everyone should be aware that buying Microsoft software will pretty much lock them into Microsoft software for life, which can be somewhat awkward once they outgrow it. :If a company has a large market share or is in a large market, then :the vast majority of companies *do* organise their business strategy :on the basis that their potential consumers are both naive and :stupid. This is what advertising is for. For high-tech companies :selling to high-tech consumers, there is a good argument for giving :the impression that the customers are not stupid, until those :companies move into the mass consumer market or position themselves :for so doing. The more successful a company is, the more stupid it :thinks its customers are.. Erm, that could be read as "successful companies think their products are shite". (In addition to several other readings, of course; that's just the one that jumped out at me first.) I don't think that is the case. I think Microsoft have picked up on Apple's lead and are trying to produce software where the user doesn't have to know anything "techie". (Remember "The computer for the rest of us"?) I think they're extending the concept to include users who don't know anything techie about *anything*, including the core competence areas that their products are designed to serve. (Eg. system admin, database admin, word processing, desktop publishing...) Microsoft are a long way from ever producing a professioal-quality product, ie. something that someone good at their job but technologically naive can use quickly and productively. (To a large extent, Apple cornered that market fairly quickly.) They seem to be content on concentrating on the Sun-reader market, to use an Anglicism. (dunno what the US equivalent would be... New York Post rather than New York Times, et al?) :Compare the Microsoft "Where do you want to go today?" with the :Cisco (is it cisco?) "Are you ready?". Same shit. Different company. (Thought it was Gillette Ladyshave, personally, but *ahem*.) It's hard to feel bad about Cisco, considering their origins. But comparing ad slogans is hardly saying anything about the products; I'd be very surprised if Cisco were selling routers to people who didn't know why they needed them - *exactly* the group MS seem to be targeting. :What about the Apple ads? Scary. I think it's disingenious to :overestimate the knowledge of the general public (and that includes :me). For example,the Un*x sound bite could be: "rm -r * - you know :it makes sense" and i'd buy it. Overestimating initial knowledge is one thing. What distresses me about Microsoft is the way they underestimate people's capacity to acquire new knowledge. -- Communa (lisard@zetnet.co.uk) -- you know soft spoken changes nothing ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 1 Dec 1998 19:43:40 GMT Lines: 54 Message-ID: <741gtc$88i$5@irk.zetnet.co.uk> References: <36600eec.6909655@news.uunet.be> NNTP-Posting-Host: man-141.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 912541420 8466 194.247.40.180 (1 Dec 1998 19:43:40 GMT) NNTP-Posting-Date: 1 Dec 1998 19:43:40 GMT X-Everything: Net-Tamer V 1.08X Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!baron.netcom.net.uk!netcom.net.uk!demeter.clara.net!news.clara.net!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-11-27 lucvdv@null.net(LucVanderVeken) said: :Also sprach lisard@zetnet.co.uk: :> Well, except for MacForth, which was released very early into the :>Mac's lifespan :By Apple? (I really don't know) No, by a distinct company which has since been gobbled up by that corporate giant, Forth Inc. ;> :> Easy and crap. :As for VBScript, true. VBS in itself isn't even that bad - if you :keep in mind that for simple things like copying a few files you :better write a .cmd or a .bat instead of a .vbs - it's the :documentation that really stinks (with 2 different online help :systems available, publishing a reference work as a word document :isn't my idea of a good job.) Well, let's see. VBScript is basically a different front end to the same language as JavaScript, which I suppose means that you can assume the actual language to not be so limited. It's not a bad language, and it's great for tying COM components together... ish. Unfortunately, the quality of the COM components themselves seems to leave a lot to be desired. I'm still trying to get over the limitations of their Collection class. Eurgh! But what I was thinking of were the earlier versions of VB, which (compared to the current version, which *still* imposes an arbitrary restriction on which bits of the API you can call) were frankly not up to it. Great development environment, love the form editor (well, actually no, but that's a separate issue) but the basic language just had too many limitations on it for anything beyond a self-contained program. :VB is easy to learn, which isn't exactly true for Forth (at least :they told me it was hard before I started with it - actually I :found learning it not that hard, doing something useful in it :just took too much work - but I'm talking about [obafc] TRS-80 :Forth now, I suppose today's versions don't require you to :reinvent the wheel every day anymore.) Forth has moved forward too, yes. There are a fair few good commercial versions, apparently, and in the free world you have Win32Forth and Gforth, which seem to be shaping up as fairly heavyweight contenders. :That's true (as far as I know HyperCard - which is all hear-say :and magazine-read). Ditto. (to make that clear :> ) -- Communa (lisard@zetnet.co.uk) -- you know soft spoken changes nothing ###### Date: 01 Dec 98 16:19:15 -0800 From: "Charlie Gibbs" Subject: Re: Multics References: <365fd57c.3293716@news.ricochet.net> <741gt9$88i$4@irk.zetnet.co.uk> Message-ID: <3639.639T780T9793813@sky.bus.com> Newsgroups: alt.folklore.computers Lines: 86 X-Newsreader: THOR 2.5a (Amiga;TCP/IP) NNTP-Posting-Host: news.skybus.com X-Trace: 1 Dec 1998 17:59:12 -0800, news.skybus.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!isdnet!howland.erols.net!news-peer1.sprintlink.net!news-backup-west.sprintlink.net!news.sprintlink.net!news.westel.com!news.skybus.com!204.244.247.120 In article <741gt9$88i$4@irk.zetnet.co.uk> lisard@zetnet.co.uk (lisard) writes: >In any case, what I mean is that people shouldn't regard buying a >computer as a decision in the same vein as buying (say) a washing >machine or a microwave oven. Or even a cellphone. Unfortunately, not only do people do so, but it's generally regarded as a Good Thing. >Nobody who doesn't want to should know assembler, certainly. But >everyone who buys a computer should know about file systems, because >they'll be using one every day - certainly when they get past the >"This is my first letter in

REALLY BIG FONTS

" stage. And even then, if they decide to save that little gem. >And everyone should be aware that buying Microsoft software will >pretty much lock them into Microsoft software for life, Do I hear a snicker from the direction of Redmond? >which can be somewhat awkward once they outgrow it. Many people probably never will outgrow it - especially if they can be enticed by flashy screens into keeping their expectations suitably low. >:If a company has a large market share or is in a large market, then >:the vast majority of companies *do* organise their business strategy >:on the basis that their potential consumers are both naive and >:stupid. Based on my tech support experience, this is a safe bet. >:For high-tech companies selling to high-tech consumers, there is >:a good argument for giving the impression that the customers are >:not stupid, Agreed. >:until those companies move into the mass consumer market or position >:themselves for so doing. The more successful a company is, the more >:stupid it thinks its customers are.. That's just a fact of life. There's only room for so many people in that little corner at the upper end of the bell curve. If you're going after a bigger part of the market, you have no choice but to slide down the curve. Remember, half of all people are below average intelligence. :-) >Microsoft are a long way from ever producing a professioal-quality >product, ie. something that someone good at their job but >technologically naive can use quickly and productively. (To a >large extent, Apple cornered that market fairly quickly.) But Microsoft isn't after a professional-quality product. They want to monopolize the market, and it's a waste of time to concentrate too much on quality when most of their customers (remember that bell curve?) either don't know the difference or can be persuaded by fancy advertising for a lot less than true quality would cost. >They seem to be content on concentrating on the Sun-reader market, >to use an Anglicism. (dunno what the US equivalent would be... New >York Post rather than New York Times, et al?) USA Today, maybe? There probably is no exact equivalent, since I don't think any American tabloid duplicates the Sun's page 3. :-) >Overestimating initial knowledge is one thing. What distresses me >about Microsoft is the way they underestimate people's capacity to >acquire new knowledge. When I'm feeling especially paranoid, I'm inclined to replace "underestimate" above with "undermine". After all, one of the reasons Microsoft hires people right out of school is that such people don't have enough experience to realize that there are viable alternatives to the Microsoft Way. And the "paradigm of the week" philosophy that's sweeping the industry is designed in part to deal with the rest of us. If someone has enough knowledge to be potential trouble, you can remove the threat by invalidating his knowledge. -- cgibbs@sky.bus.com (Charlie Gibbs) Remove the first period after the "at" sign to reply. ###### Newsgroups: alt.folklore.computers From: ehrice@his.com (Edward Rice) Subject: Re: Multics Message-ID: Organization: NDS Date: Tue, 01 Dec 1998 23:27:15 -0500 References: <766.630T2695T8146524@sky.bus.com> <73f3hl$f8l$2@irk.zetnet.co.uk> <365e656e.7810160@news.uunet.be> <9U072.31715$gI1.5820@news2.giganews.com> NNTP-Posting-Host: ehrice.his.com Lines: 18 X-Authenticated-User: ehrice Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!newspeer1.nac.net!newspeer.monmouth.com!news.lightlink.com!news4.his.com!user In article <9U072.31715$gI1.5820@news2.giganews.com>, Mike Swaim wrote: > I believe that Apple did have a BASIC in development when the Mac came > out. (Some books for it even came out.) The word on the street was that it > got killed at MS's request, because it would compete with Microsoft's > BASIC. MS BASIC for the Mac did eventually come out, but it sucked and > quickly died. MacBASIC, which came out of THINK Technologies (which was later absorbed into Symantec). MacBASIC came out in a from-Apple flavor originally, only as an interpreter. Later, THINK released the same thing, plus a Pascal compiler environment. LOVELY programmer interface and debugging tools -- far better than anything else I've seen. ###### Newsgroups: alt.folklore.computers Subject: Re: Multics References: <365fd57c.3293716@news.ricochet.net> <741gt9$88i$4@irk.zetnet.co.uk> <3639.639T780T9793813@sky.bus.com> From: csess@blaze.trentu.ca (Eric S. Smith: Left-Field Marshal) Organization: Canadian Federation of Anarchists Board of Governors Originator: ess@crunch-not-really-uucp Lines: 48 Message-ID: Date: Wed, 02 Dec 1998 02:51:28 GMT NNTP-Posting-Host: 207.61.241.166 NNTP-Posting-Date: Tue, 01 Dec 1998 21:51:28 EDT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsfeed.cwix.com!130.185.14.36!torn!news1.bellglobal.com!198.235.216.4.POSTED!not-for-mail In article <3639.639T780T9793813@sky.bus.com>, Charlie Gibbs wrote: >In article <741gt9$88i$4@irk.zetnet.co.uk> lisard@zetnet.co.uk (lisard) >writes: >>In any case, what I mean is that people shouldn't regard buying a >>computer as a decision in the same vein as buying (say) a washing >>machine or a microwave oven. Or even a cellphone. >Unfortunately, not only do people do so, but it's generally >regarded as a Good Thing. Given the amount of research and inspection I'd do before forking out for a washing machine, microwave, or cellular phone (phone esp., given the ongoing albatross of monthly bills), I'd instinctively say that of *course* it should be in the same vein. Naturally, the world is full of people who're more foolish than I am... And here's a thought: by the time you're in a position to buy a washing machine, you know what you can do with it, and, vaguely, how. It's entirely possible to buy a computer without even knowing that much -- perhaps the problem here isn't so much that appliance purchases require less of the purchaser than computer purchases, but that the purchaser is rarely a complete appliance newbie. More people know what a washing machine cycle is than know what a CPU cycle is... The idea of "computer as appliance" isn't as repulsive to me as it is to some people (many of whom are reading this right now, I imagine...). Really, truly, ideally, buying a home micro ought to be a pretty simple process, and using one ought to be simple, too (but see above). The problem with PCs is that you've got a hopped up piece of office equipment trying to be cuddly, and it's like using a filing cabinet for a toybox -- you can do it, and, in some ways, it's better than a plain ol' toybox, but... And a filing cabinet with only one, slighly narrower drawer (the PC Jr by analogy) isn't the way to go, either. The iMac as pitched looks like a better attempt, but it's still the MacOS running on the thing, and even that's a bit hairy. I'm not saying that the thing could ever be made drool-proof, but if someone just wants to poke at Web pages and write letters, I don't see why they should be plunking down $2500 for some cruft-laden overpowered adding machine. I can buy a stereo system, a ghetto blaster with tape and CD player, or a clock radio. Where are the similar choices in home computing? I submit that this may be *part* of the problem... --Eric Smith ###### Message-ID: <366C8171.62693A07@loudoun.com> Date: Mon, 07 Dec 1998 20:31:30 -0500 From: Robert & Debbie Fetter X-Mailer: Mozilla 4.5 [en] (Win98; I) X-Accept-Language: en MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: Multics References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Host: 204.188.47.192 X-Trace: 7 Dec 1998 20:32:34 -0500, 204.188.47.192 Lines: 44 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!news-peer.gip.net!news.gsl.net!gip.net!cpk-news-hub1.bbnplanet.com!news.gtei.net!newsfeed.cwix.com!204.71.1.51!spamkiller2.cwix.com!pull-feed.cwix.com!vangogh.loudoun.com!204.188.47.192 Joe Morris wrote: > > Only when the personal computer restored the "programmer as operator" > mode did many of the application programmers begin to appreciate the > interaction of program design and computer architecture -- and of course > some of the programmers still don't show any understanding of the issues. > > Joe Morris I have to agree. I started out in this business as an operator of 2nd/3rd generation systems (Honeywell H-200/H-800, IBM 1620, IBM 360) and then grew into software. It was a firm belief of mine then that all software developers should have had a stint in the computer room, just so they could *watch* their software run. Only then would the silliness of having tape sorts vs. disk/drum sorts and single pass with embedded sorts (with sort output validated and applied) vs. pass for edit, pass for sort, pass for validity, pass for apply be obvious. Maybe it was the endless hours of running to and from tape drives that pounded this into my head -- anyways, I still look for those extra instructions in loops and those extra and needless portions of objects to eliminate. The worst I've seen recently is an application with 50K+ bytes of text data, and the application treats it as an object: "go get this item" (parse the whole damn thing, get the one damn thing, clean up), then "go get this item" (do it again), then ... And, this from the leading edge of C++ minds (or so I've been told). On 400MH Pentium II boxes, still takes countable clock seconds, where a blink of an eye would do. When presented with this, though, the attitude was 'well, that *is* a lot to do' (????) It's only 'a lot to do' if you *do* it. Some of the best solutions in this field is obtained by knowing what NOT to do (IMHO). ###### From: Anne & Lynn Wheeler Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 07 Dec 1998 23:15:06 -0800 Organization: Wheeler&Wheeler Lines: 17 Message-ID: References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> Reply-To: Anne & Lynn Wheeler NNTP-Posting-Host: lynn-19.garlic.com X-Newsreader: Gnus v5.6.44/Emacs 20.3 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!news.maxwell.syr.edu!news.shore.net!uunet!in3.uu.net!bulb.garlic.com!not-for-mail summer after first programming intro class when I was a sophomore ... they would left me have the machine room to myself from 8am sat. until 8am monday .... 48hr straight thru shift (initially only a 709, 1401, and a 360/30 ... plus misc. other things ... made it a little hard to make a monday morning summer school class). I quickly acquired the habit of first thing cleaning all the tape drives, card reader, and card punch ... before even bothering to proceed. For those 48 hrs ... everything in the room was my personal computer ... had to develop habit of going w/o sleep for 50-60hr stretches. nothing substitutes for hands on experience (and it also beats washing dishes in cafeteria) -- -- Anne & Lynn Wheeler | lynn@garlic.com, finger for pgp key http://www.garlic.com/~lynn/ ###### From: nailed_barnacle@junkfree.hotmail.com (barnacle) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Tue, 08 Dec 1998 07:49:46 GMT Organization: [posted via Easynet Ltd] Lines: 24 Message-ID: <74ilqp$fs1$1@quince.news.easynet.net> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> NNTP-Posting-Host: nbarnes.easynet.co.uk X-Trace: quince.news.easynet.net 913103513 16257 194.154.98.206 (8 Dec 1998 07:51:53 GMT) X-Complaints-To: abuse@easynet.net NNTP-Posting-Date: 8 Dec 1998 07:51:53 GMT X-Newsreader: News Xpress 2.01 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!easynet-tele!quince.news.easynet.net!easynet-quince!egbert In article <366C8171.62693A07@loudoun.com>, Robert & Debbie Fetter wrote: ---snip--- >the leading edge of C++ minds (or so I've been told). On 400MH Pentium II > boxes, >still >takes countable clock seconds, where a blink of an eye would do. > When presented with this, though, the attitude was 'well, that *is* a lot to >do' (????) >It's only 'a lot to do' if you *do* it. Some of the best solutions in this > field >is obtained by knowing what NOT to do (IMHO). > > Well said! Everytime I hear how fast the next generation of processors/memory/discs is going to be, I wonder how much bigger/slower/more useless bells and whistles the next generation of software bloat will be...I ask you - is it reasonable that a 333MHz processor with 128M ram and an 8 gig disk should be considered an 'entry level' system when the average user wants to do nothing more than play games, browse the net, and play games? barnacle nailed-barnacle.home.ml.org ###### From: fleetvis@ricochet.net (phil) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Tue, 08 Dec 1998 07:51:11 GMT Organization: Posted via RemarQ, http://www.remarQ.com - Discussions start here! Lines: 37 Message-ID: <366cd500.8592695@news.ricochet.net> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> NNTP-Posting-Host: 204.179.130.44 X-Trace: 913103206 F7W8ZAFJM822CCCB3C usenet53.supernews.com X-Complaints-To: newsabuse@remarQ.com X-Newsreader: Forte Free Agent 1.11/32.235 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!feeder.qis.net!newsfeed-east.supernews.com!SupernewsUK!supernews.com!SnNA!Supernews69!not-for-mail On 30 Nov 1998 15:16:04 GMT, jcmorris@mwunix.mitre.org (Joe Morris) wrote: >Mike Swaim writes: > >>phil wrote: > >>: There are a >>: surprising number of programmers that have only a vague idea of how a >>: computer works and that, for example, couldn't write an interrupt >>: service to save their lives, and why should they? > > >While there is a legitimate range of opinions about how much an application >programmer needs to know about the underlying details of a computer, I'll >strongly agree that anyone with pretentions of working in *any* field >needs to be at least somewhat familiar with what's going on below the >level where he or she is working. In the computer field this doesn't >mean that a business application programmer writing an A/P package in COBOL >needs to be able to design the mask for a new Pentium chip, but that same >programmer should be more than passingly familiar with the organization of >a computer and how the application program interacts with it. > Exactly. It's how deep that knowledge has to go that matters. I chose an ISR because it was an obvious discontinuity. A device driver or a kernel service might be more appropriate on more recent machines. Today i think that the paradigm has shifted significantly. The successful programmer these days knows how to integrate components together and the significance of the functionality of a computer is a very bad second to the functionality of the tools available (general purpose stuff, not embedded systems), whatever they reveal of the underlying architecture. I know this is AFC and so these is fightin' words, but it's the way i see it. phil. ###### From: "Samael" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Tue, 8 Dec 1998 13:08:19 -0000 Lines: 22 Message-ID: <366d24c4.0@192.168.0.20> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> NNTP-Posting-Host: 195.147.135.72 X-Newsreader: Microsoft Outlook Express 4.72.3155.0 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3155.0 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!howland.erols.net!news-peer.gip.net!news-lond.gip.net!news.gsl.net!gip.net!nntp.news.xara.net!xara.net!news.itg.net.uk!192.168.0.20!192.168.0.1 barnacle wrote in message <74ilqp$fs1$1@quince.news.easynet.net>... >Well said! Everytime I hear how fast the next generation of >processors/memory/discs is going to be, I wonder how much bigger/slower/more >useless bells and whistles the next generation of software bloat will be...I >ask you - is it reasonable that a 333MHz processor with 128M ram and an 8 gig >disk should be considered an 'entry level' system when the average user wants >to do nothing more than play games, browse the net, and play games? When my net browsing contains large amounts of streaming video and my games a throwing large amounts of textured mip-mapped polygons at the screen, whiolst attempting to simulate basic pack cooperation for the bad guys shooting at me, then yes. I want a faster machine, dammit. Samael ###### From: deke.spamblock@generous.net Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Tue, 08 Dec 1998 13:38:32 GMT Organization: GenerousCity is a virtue - find romance at http://generous.net Lines: 29 Message-ID: <366d2a29.12511127@news.bright.net> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> NNTP-Posting-Host: paul-cas2-cs-29.dial.bright.net X-Newsreader: Forte Free Agent 1.11/32.235 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!194.165.93.117.MISMATCH!newshub.bart.net!newsfeed.wirehub.nl!howland.erols.net!vixen.cso.uiuc.edu!NewsNG.Chicago.Qual.Net!news.bright.net!not-for-mail On Tue, 08 Dec 1998 07:49:46 GMT, nailed_barnacle@junkfree.hotmail.com (barnacle) wrote: >Well said! Everytime I hear how fast the next generation of >processors/memory/discs is going to be, I wonder how much bigger/slower/more >useless bells and whistles the next generation of software bloat will be...I >ask you - is it reasonable that a 333MHz processor with 128M ram and an 8 gig >disk should be considered an 'entry level' system when the average user wants >to do nothing more than play games, browse the net, and play games? "Nothing more"? What do you think would possibly requires more horsepower than playing games? Those guys are trying to compete with 64-bit systems from Sega, Nintendo, etc. For browsing the net, there's no point in most people having anything more than a 486SX25 unless they download through a satellite dish. Living more than 3 miles from the phone station, so I can't even take advantage of 56K modems, so I get stuck with 26.4K or 28.8K on my 33.6K modem.... Oh, well, it beats hell out of that 1200 baud modem I have in the closet.... deke ------------------------ Let love find you! http://generous.net A list for flirting GenerousSingles-subscribe@onelist.com Over The Hill Gang GenerousSinglesOTHG-subscribe@onelist.com Personal Ads list GenerousProfiles-subscribe@onelist.com ###### From: nailed_barnacle@junkfree.hotmail.com (barnacle) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Tue, 08 Dec 1998 17:09:19 GMT Organization: [posted via Easynet Ltd] Lines: 29 Message-ID: <74jmk1$1kvu$1@quince.news.easynet.net> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d24c4.0@192.168.0.20> NNTP-Posting-Host: nbarnes.easynet.co.uk X-Trace: quince.news.easynet.net 913137089 54270 194.154.98.206 (8 Dec 1998 17:11:29 GMT) X-Complaints-To: abuse@easynet.net NNTP-Posting-Date: 8 Dec 1998 17:11:29 GMT X-Newsreader: News Xpress 2.01 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news.tele2.nl!newsfeed1.swip.net!swipnet!masternews.telia.net!newsfeed.ecrc.net!newsfeed.nacamar.de!easynet-tele!quince.news.easynet.net!easynet-quince!egbert In article <366d24c4.0@192.168.0.20>, "Samael" wrote: > >barnacle wrote in message <74ilqp$fs1$1@quince.news.easynet.net>... >>Well said! Everytime I hear how fast the next generation of >>processors/memory/discs is going to be, I wonder how much >bigger/slower/more >>useless bells and whistles the next generation of software bloat will >be...I >>ask you - is it reasonable that a 333MHz processor with 128M ram and an 8 >gig >>disk should be considered an 'entry level' system when the average user >wants >>to do nothing more than play games, browse the net, and play games? > > >When my net browsing contains large amounts of streaming video and my games >a throwing large amounts of textured mip-mapped polygons at the screen, >whiolst attempting to simulate basic pack cooperation for the bad guys >shooting at me, then yes. I want a faster machine, dammit. > >Samael Fair comment Samael...but I see all this extra power and it's not used by the user...it's the bloody operating system (no names but it's small and flexible) adding improved uselessness...do I really need dancing paperclips? > barnacle nailed-barnacle.home.ml.org ###### From: Kin Hoong CHUNG Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 9 Dec 1998 01:30:25 GMT Organization: University of New South Wales Lines: 17 Message-ID: <74kjrh$jbn$2@mirv.unsw.edu.au> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> NNTP-Posting-Host: alpha.maths.unsw.edu.au User-Agent: tin/pre-1.4-980618 (UNIX) (OSF1/V4.0 (alpha)) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.cwix.com!192.26.210.166!sunqbc.risq.qc.ca!news.uow.edu.au!news.usyd.edu.au!unsw.edu.au!not-for-mail deke.spamblock@generous.net wrote: : "Nothing more"? What do you think would possibly requires more horsepower than : playing games? Those guys are trying to compete with 64-bit systems from Sega, : Nintendo, etc. There are many things which require computing power than any game one could imagine. Some of them are even important enough that people will buy bleeding-edge supercomputers to deal with them. The standard example is weather forecasting. Another is mathematical modelling of anything from bridges to airflow around aircraft wings. There are also some people who think that factorising very large primes is an important pursuit too. Cheers, Kin Hoong ###### From: "Samael" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Wed, 9 Dec 1998 10:22:53 -0000 Lines: 40 Message-ID: <366e4f83.0@192.168.0.20> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d24c4.0@192.168.0.20> <74jmk1$1kvu$1@quince.news.easynet.net> NNTP-Posting-Host: 195.147.139.23 X-Newsreader: Microsoft Outlook Express 4.72.3155.0 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3155.0 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news-feed.inet.tele.dk!bofh.vszbr.cz!news-lond.gip.net!news.gsl.net!gip.net!nntp.news.xara.net!xara.net!news.itg.net.uk!192.168.0.20!192.168.0.1 barnacle wrote in message <74jmk1$1kvu$1@quince.news.easynet.net>... >In article <366d24c4.0@192.168.0.20>, "Samael" wrote: >> >>barnacle wrote in message <74ilqp$fs1$1@quince.news.easynet.net>... >>>Well said! Everytime I hear how fast the next generation of >>>processors/memory/discs is going to be, I wonder how much >>bigger/slower/more >>>useless bells and whistles the next generation of software bloat will >>be...I >>>ask you - is it reasonable that a 333MHz processor with 128M ram and an 8 >>gig >>>disk should be considered an 'entry level' system when the average user >>wants >>>to do nothing more than play games, browse the net, and play games? >> >> >>When my net browsing contains large amounts of streaming video and my games >>a throwing large amounts of textured mip-mapped polygons at the screen, >>whiolst attempting to simulate basic pack cooperation for the bad guys >>shooting at me, then yes. I want a faster machine, dammit. >> >>Samael >Fair comment Samael...but I see all this extra power and it's not used by the >user...it's the bloody operating system (no names but it's small and flexible) >adding improved uselessness...do I really need dancing paperclips? Yes, our office secretary, who finds it invaluable. Also, my mum. Other non-computer people who find it infinitely easier to interface with a program that has an anthropomorphic front end. Samael ###### From: DANN@PLAN9.GREYCAT.COM (Dann Lunsford) Newsgroups: alt.folklore.computers Subject: Re: Multics References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> Message-ID: User-Agent: slrn/0.9.5.4 (VMS) NNTP-Posting-Host: plan9.greycat.com X-NNTP-Posting-Host: plan9.greycat.com Date: 9 Dec 1998 10:33:08 +0800 X-Trace: 9 Dec 1998 10:33:08 +0800, plan9.greycat.com Organization: "CalWeb Internet Services Inc. Your affordable ISP." Lines: 8 X-NNTP-Posting-Host: news.calweb.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newsfeed.uk.ibm.net!ibm.net!news-feed.inet.tele.dk!bofh.vszbr.cz!chippy.visi.com!news-out.visi.com!calwebnntp!calwebnnrp!DANN In article <74kjrh$jbn$2@mirv.unsw.edu.au>, Kin Hoong CHUNG wrote: >There are also >some people who think that factorising very large primes is an important >pursuit too. ^^^^^^^^^^^ ^^^^^^ Hmmm... Bill Gates, posting under a pseudonym, via a .au relay? Whoda thought? :-) ###### Date: 09 Dec 98 11:28:42 -0800 From: "Charlie Gibbs" Subject: Re: Multics References: <366e4f83.0@192.168.0.20> Message-ID: <1595.647T2126T6885175@sky.bus.com> Newsgroups: alt.folklore.computers Lines: 36 X-Newsreader: THOR 2.5a (Amiga;TCP/IP) NNTP-Posting-Host: news.skybus.com X-Trace: 9 Dec 1998 12:26:06 -0800, news.skybus.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!howland.erols.net!news-peer1.sprintlink.net!news-in-east1.sprintlink.net!news.sprintlink.net!news.westel.com!news.skybus.com!204.244.247.121 In article <366e4f83.0@192.168.0.20> Samael@dial.pipex.com (Samael) writes: >barnacle wrote in message <74jmk1$1kvu$1@quince.news.easynet.net>... >>Fair comment Samael...but I see all this extra power and it's not >>used by the user...it's the bloody operating system (no names but >>it's small and flexible) adding improved uselessness...do I really >>need dancing paperclips? > >Yes, our office secretary, who finds it invaluable. >Also, my mum. >Other non-computer people who find it infinitely easier to interface >with a program that has an anthropomorphic front end. But do they actually manage to do more with those dancing paper clips than with a text-based program? Now that the average user has more computer power on his or her desk than it took to put a man on the moon, do they do any better a job of writing a letter? We produce a software package that runs under MS-DOS and Windows (as well as Unix). Our tech support experience is that users don't find the Windows version that much easier to use at all. (The screens are much prettier, though.) From a support standpoint, Windows is a nightmare - it's much easier to have the user type a few keystrokes and read what comes up on the screen than to have him manoeuvre a mouse around a screen and click on something that might not even be visible. Remember, we're working through a telephone and can't see the user's screen. Although GUIs do have their advantages, much of their popularity is a matter of style over substance. -- cgibbs@sky.bus.com (Charlie Gibbs) Remove the first period after the "at" sign to reply. ###### Newsgroups: alt.folklore.computers From: m@mbsks.franken.de (Matthias Bruestle) Subject: Re: Multics References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> X-No-Archive: yes Organization: PDP-11/34A Date: Wed, 9 Dec 1998 19:45:07 GMT X-Newsreader: TIN [version 1.2 PL2] Message-ID: <1998Dec9.194507.18383@mbsks.franken.de> Lines: 23 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!howland.erols.net!newsfeed.nacamar.de!uni-erlangen.de!chico.franken.de!mbsks.franken.de!m Mahlzeit Kin Hoong CHUNG (khchung@maths.unsw.edu.au) wrote: > There are many things which require computing power than any game one > could imagine. Some of them are even important enough that people will > buy bleeding-edge supercomputers to deal with them. The standard The difference is not the required computing power. There are many things, which require infinite computing power, and one of them is gaming. The difference is budget. If everybody could buy a 1024 processor computer for US$1000 you can be sure there were games which need these computers. Mahlzeit endergone Zwiebeltuete -- PGP: SIG:C379A331 ENC:F47FA83D I LOVE MY PDP-11/34A, M70 and MicroVAXII! -- A lot of occult writers have made some amazing guesses as to what that means. Actually, it doesn't mean a damned thing. ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 9 Dec 1998 20:31:43 GMT Message-ID: <74mmnf$4l7$7@irk.zetnet.co.uk> References: <366d2a29.12511127@news.bright.net> NNTP-Posting-Host: man-178.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 913235503 4775 194.247.40.226 (9 Dec 1998 20:31:43 GMT) NNTP-Posting-Date: 9 Dec 1998 20:31:43 GMT X-Everything: Net-Tamer V 1.08X Lines: 12 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newsfeed.uk.ibm.net!ibm.net!news-lond.gip.net!news.gsl.net!gip.net!dispose.news.demon.net!demon!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-12-08 deke.spamblock@generous.net said: :"Nothing more"? What do you think would possibly requires more :horsepower than playing games? Those guys are trying to compete :with 64-bit systems from Sega, Nintendo, etc. ..that sell for 100 quid all in and plug in to the family TV... It'll be a while before PC prices fall to this level, if they ever do. And frankly, I doubt they'll ever be as good. -- Communa (lisard@zetnet.co.uk) -- you know soft spoken changes nothing ###### Message-ID: <366F0E61.275D@gazonk.del> Date: Wed, 09 Dec 1998 18:57:21 -0500 From: "Foobar T. Clown" Organization: Blurp X-Mailer: Mozilla 3.01 (WinNT; I) MIME-Version: 1.0 Newsgroups: alt.folklore.computers Subject: Re: Multics References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit NNTP-Posting-Host: 198.133.110.49 X-Trace: 9 Dec 1998 18:58:46 -0500, 198.133.110.49 Lines: 17 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newshub.northeast.verio.net!ix.netcom.com!zeus.nomos.com!198.133.110.49 deke.spamblock@generous.net wrote: > > "Nothing more"? What do you think would possibly requires more > horsepower than playing games? Those guys are trying to compete with > 64-bit systems from Sega, Nintendo, etc. Trying? I don't think those game systems can touch an up-to-date, multimedia PC. Don't forget that your PC has a minimum of two processors: One of them is the Pentium whatever, and the other one is the "accelerator" on your video adaptor. I don't know what's in a 64-bit game console, but my guess is that any one of today's hot video accelerators could beat it up without needing help from your system's CPU. Consider, especially, that those game consoles only have to do T.V. rez graphics at a leisurely, 30Hz or 25Hz frame rate. ###### From: "Andrew Cadley" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Thu, 10 Dec 1998 12:09:43 -0000 Organization: University of East Anglia, Norwich, Norfolk, NR47TJ, UK Lines: 47 Message-ID: <74odpk$9j4@cpca3.uea.ac.uk> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <366F0E61.275D@gazonk.del> NNTP-Posting-Host: oak6d.vlg1.uea.ac.uk Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit 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!news-ge.switch.ch!news.maxwell.syr.edu!rill.news.pipex.net!pipex!server1.netnews.ja.net!server2.netnews.ja.net!cpca3.uea.ac.uk!news Foobar T. Clown wrote in message <366F0E61.275D@gazonk.del>... :deke.spamblock@generous.net wrote: :> :> "Nothing more"? What do you think would possibly requires more :> horsepower than playing games? Those guys are trying to compete with :> 64-bit systems from Sega, Nintendo, etc. : :Trying? : :I don't think those game systems can touch an up-to-date, multimedia :PC. Don't forget that your PC has a minimum of two processors: One of :them is the Pentium whatever, and the other one is the "accelerator" on :your video adaptor. I don't know what's in a 64-bit game console, but :my guess is that any one of today's hot video accelerators could beat it :up without needing help from your system's CPU. Well in an N64: "64-bit MIPS R4300 RISC CPU running at 93.75 Mhz (125 MIPs) with true 64-bit data path and registers 64-bit RISC "Reality Immersion" graphics co-processor running at 62.5 Mhz (100+ MFLOPS) with onboard rendering processor to handle textures, anti-aliasing, z-buffering Geometry engine provides support for hardware based shading, z-buffering, perspective correct texture mapping, tri-linear mip-mapping, anti-aliasing, 256-level alpha channel, LOD Management. Sprite effects include rotation, scaling, anti-aliasing, 256-level alpha channel effects. Overall, the Nintendo 64 is capable of rendering about 160,000 polygons with all hardware features enabled." Which s certainly a lot more grunt than the average home PC, especially when you consider the difference in price. :Consider, especially, that those game consoles only have to do T.V. rez :graphics at a leisurely, 30Hz or 25Hz frame rate. TV is usually 50Hz (PAL) or 60Hz (NTSC). The N64 can do 640*480 21 bit colour output, which is pretty good even on a TV. AndyC ###### Newsgroups: alt.folklore.computers From: "Ralph Wade Phillips" Subject: Re: Multics X-Nntp-Posting-Host: zs150050.shrv.bna.boeing.com Message-ID: X-Mimeole: Produced By Microsoft MimeOLE V4.72.3110.3 Lines: 37 Sender: nntp@news.boeing.com (Boeing NNTP News Access) Organization: The Boeing Company X-Newsreader: Microsoft Outlook Express 4.72.3110.5 References: <366e4f83.0@192.168.0.20> <1595.647T2126T6885175@sky.bus.com> Date: Thu, 10 Dec 1998 15:23:46 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!surfnet.nl!newsfeed.wirehub.nl!newspeer1.nac.net!logbridge.uoregon.edu!netnews1.nw.verio.net!netnews.nwnet.net!xyzzy!not-for-mail Hi, Charlie! Charlie Gibbs wrote in message <1595.647T2126T6885175@sky.bus.com>... (snippity) > >We produce a software package that runs under MS-DOS and Windows (as >well as Unix). Our tech support experience is that users don't find >the Windows version that much easier to use at all. (The screens are >much prettier, though.) From a support standpoint, Windows is a >nightmare - it's much easier to have the user type a few keystrokes >and read what comes up on the screen than to have him manoeuvre a >mouse around a screen and click on something that might not even be >visible. Remember, we're working through a telephone and can't see >the user's screen. > >Although GUIs do have their advantages, much of their popularity >is a matter of style over substance. > I submit then that your programmers have mush-for-brains when it comes to a GUI's human interface component. You SHOULD be able to do ANYTHING from the keyboard, even if it's Alt-F or Ctrl-C to get there. No, Microsoft doesn't have it right, and the Mac is even further retarded in that respect. BUT - you should NEVER require a "mouse" of any kind to get the program to operate. When talking people through over the phone, I typically will tell them "Please put the mouse down. We will do this from the keyboard" for such things as changing passwords, etc. RwP ###### From: "Derry Hamilton" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Thu, 10 Dec 1998 16:43:26 -0000 Organization: edNET - Internet Access for Edinburgh Lines: 17 Message-ID: <74otna$oej$1@monteverdi.ednet.co.uk> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d24c4.0@192.168.0.20> <74jmk1$1kvu$1@quince.news.easynet.net> <366e4f83.0@192.168.0.20> NNTP-Posting-Host: dialup14.ednet.co.uk X-Trace: monteverdi.ednet.co.uk 913308202 25043 195.89.132.142 (10 Dec 1998 16:43:22 GMT) X-Complaints-To: abuse@ednet.co.uk NNTP-Posting-Date: 10 Dec 1998 16:43:22 GMT X-Newsreader: Microsoft Outlook Express 4.72.2106.4 X-Mimeole: Produced By Microsoft MimeOLE V4.72.2106.4 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!btnet-peer!btnet-feed1!btnet!newsfeed.ednet.co.uk!news.ednet.co.uk!not-for-mail Samael wrote in message <366e4f83.0@192.168.0.20>... >>Fair comment Samael...but I see all this extra power and it's not used by >the >>user...it's the bloody operating system (no names but it's small and >flexible) >>adding improved uselessness...do I really need dancing paperclips? > > >Yes, our office secretary, who finds it invaluable. >Also, my mum. >Other non-computer people who find it infinitely easier to interface with a >program that has an anthropomorphic front end. Alternativly, the number of clients (of all abilities) that have demanded we remove the blasted thing has made it SOP here to install Office without it. ###### Date: 10 Dec 98 11:00:46 -0800 From: "Charlie Gibbs" Subject: Re: Multics References: <366e4f83.0@192.168.0.20> <1595.647T2126T6885175@sky.bus.com> Message-ID: <2631.648T2320T6605417@sky.bus.com> Newsgroups: alt.folklore.computers Lines: 62 X-Newsreader: THOR 2.5a (Amiga;TCP/IP) NNTP-Posting-Host: news.skybus.com X-Trace: 10 Dec 1998 13:22:02 -0800, news.skybus.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!howland.erols.net!news-peer1.sprintlink.net!news-backup-west.sprintlink.net!news.sprintlink.net!news.westel.com!news.skybus.com!204.244.247.104 In article ralphp@techie.com (Ralph Wade Phillips) writes: >Hi, Charlie! > >Charlie Gibbs wrote in message <1595.647T2126T6885175@sky.bus.com>... > >(snippity) >> >>We produce a software package that runs under MS-DOS and Windows (as >>well as Unix). Our tech support experience is that users don't find >>the Windows version that much easier to use at all. (The screens are >>much prettier, though.) From a support standpoint, Windows is a >>nightmare - it's much easier to have the user type a few keystrokes >>and read what comes up on the screen than to have him manoeuvre a >>mouse around a screen and click on something that might not even be >>visible. Remember, we're working through a telephone and can't see >>the user's screen. >> >>Although GUIs do have their advantages, much of their popularity >>is a matter of style over substance. > > I submit then that your programmers have mush-for-brains when >it comes to a GUI's human interface component. I suppose I could take that as an insult, since I _am_ the programmer. But I'll keep the flamethrower holstered because you've just misunderstood something that I probably didn't write too clearly. Read on... >You SHOULD be able to do ANYTHING from the keyboard, even if it's >Alt-F or Ctrl-C to get there. Agreed, and that's exactly what I've done. In fact, you can stay with the keyboard on our Windows version and it feels exactly like the MS-DOS version (except slower, of course :-). But it's amazing how many users seem to believe that if you _can_ click on a button, then you _must_ click on it (instead of pressing , , or whatever). >No, Microsoft doesn't have it right, and the Mac is even further >retarded in that respect. BUT - you should NEVER require a "mouse" >of any kind to get the program to operate. Amen. The problems I referred to above occur when we're trying to get the user to start up our program, or look around in the system in other ways. One of our favourites is the luser who called up and said, "My TV went away!" After the usual game of "20 Questions", it turned out that he had opened another window which obscured ours. > When talking people through over the phone, I typically >will tell them "Please put the mouse down. We will do this from >the keyboard" for such things as changing passwords, etc. Yup. One of the first things we do if we want to check things out is to have them open an MS-DOS command prompt. Very foreign, very clunky - and very effective. -- cgibbs@sky.bus.com (Charlie Gibbs) Remove the first period after the "at" sign to reply. ###### From: genew@vip.net (Gene Wirchenko) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Thu, 10 Dec 1998 20:10:49 GMT Organization: Posted via RemarQ, http://www.remarQ.com - Discussions start here! Lines: 42 Message-ID: <36700b2d.3632427@news.vip.net> References: <366e4f83.0@192.168.0.20> <1595.647T2126T6885175@sky.bus.com> Reply-To: genew@vip.net NNTP-Posting-Host: 204.209.212.27 X-Trace: 913320502 A01OARAUVD41BCCD1C usenet78.supernews.com X-Complaints-To: newsabuse@remarQ.com X-Newsreader: Forte Free Agent 1.1/32.230 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newscore.univie.ac.at!howland.erols.net!news-peer.gip.net!news-lond.gip.net!news.gsl.net!gip.net!nntp.news.xara.net!xara.net!SupernewsUK!supernews.com!SnNA!Supernews69!not-for-mail "Ralph Wade Phillips" wrote: >Hi, Charlie! > >Charlie Gibbs wrote in message <1595.647T2126T6885175@sky.bus.com>... [snip] >>Although GUIs do have their advantages, much of their popularity >>is a matter of style over substance. >> > > > I submit then that your programmers have mush-for-brains when it >comes to a GUI's human interface component. You SHOULD be able to do >ANYTHING from the keyboard, even if it's Alt-F or Ctrl-C to get there. No, >Microsoft doesn't have it right, and the Mac is even further retarded in >that respect. BUT - you should NEVER require a "mouse" of any kind to get >the program to operate. I don't think so. The control keys are only part of it. I have tasks that I want done frequently. The user-hostile, inefficient CLI approach is to use batch files. The user-friendly, efficient GUI way is to... ah, blast. Let me rearrange the adjectives. The user-friendly, efficient CLI approach is to use batch files. The user-hostile, inefficient GUI way is to do everything by hand. > > When talking people through over the phone, I typically will tell >them "Please put the mouse down. We will do this from the keyboard" for >such things as changing passwords, etc. "Please put the mouse down and put your hands on the keybaord where we can see them." Sincerely, Gene Wirchenko Computerese Irregular Verb Conjugation: I have preferences. You have biases. He/She has prejudices. ###### From: bayko@pollux.cs.uregina.ca (John Bayko) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 10 Dec 1998 22:10:22 GMT Organization: University of Regina, Dept. of Computer Science Lines: 14 Message-ID: <74pgse$nm0$1@sue.cc.uregina.ca> References: <365d422d.22238927@news.ricochet.net> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <366F0E61.275D@gazonk.del> NNTP-Posting-Host: pollux.cs.uregina.ca Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news.tele2.nl!newsfeed1.swip.net!swipnet!newsfeed.direct.ca!cyclone.bc.net!mongol.sasknet.sk.ca!news.uregina.ca!not-for-mail In article <366F0E61.275D@gazonk.del>, Foobar T. Clown wrote: [...] >Don't forget that your PC has a minimum of two processors: One of >them is the Pentium whatever, and the other one is the "accelerator" on >your video adaptor. [...] Probably more - your keyboard could have more computing power than a VIC-20 did way back when (and probably a varient of the same CPU). -- John Bayko (Tau). bayko@cs.uregina.ca http://www.cs.uregina.ca/~bayko ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Thu, 10 Dec 1998 23:49:13 GMT Organization: . Lines: 78 Message-ID: <367133b9.52105223@news.uunet.be> References: <366e4f83.0@192.168.0.20> <1595.647T2126T6885175@sky.bus.com> NNTP-Posting-Host: pool02b-194-7-150-33.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!newsfeed.tli.de!news-feed.inet.tele.dk!bofh.vszbr.cz!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach "Ralph Wade Phillips" on Thu, 10 Dec 1998 15:23:46 GMT to alt.folklore.computers: > ... You SHOULD be able to do > ANYTHING from the keyboard, even if it's Alt-F or Ctrl-C to get there. No, Compared to emacs, those are easy combinations :-) That (your remark about keyboard, not mine about emacs) is what I always kept in mind while working on a project that's just about to get shipped now. I kinda like keyboard shortcuts myself, you see ;-) I even did my best to make navigation using the keyboard not just easy - but plain not necessary wherever I could. For example in a form where information is entered, you start out at the first field automatically when the form is loaded, the enter key moves you from field to field (except in one field, which is a multi-line one: there it inserts a line break and it only leaves the field if you use it in an empty line), and when the last field gets the focus, the "OK" button is made default (when they skip back to another field, the default status is reset again). There are a few spots where the mouse _is_ easier, and that's where I put all possible functionality in context menus under the right mouse button, customizing the menu for the selected item before I show it (you could use the menu key I guess, but you'd have to navigate to the right choice with the arrow keys - too many choices, and too many of them change with the selected item, to assign keys: it would only get confusing). Can you imagine how I felt a few months ago when I was demonstrating the first beta copy that was going to be used outside of our company HQ, at the client's site, and got into a conversation like this: cl> Heh - it's Windows 95! (no it wasn't, it was NT.) cl> I don't like that. I'd much have preferred a DOS program, I'm used to do everything with the keyboard. me> That's no problem, miss. You can do anything you can do with the mouse with the keyboard as well. And you can even do most things in more than one way - still using only the keyboard: for example, to zgrublyfsck, you press ALT-M Z (menu drops down, selection Z is executed), or [pressing ALT-M again] see that 'F12' there next to that 'zgrublyfsck' choice? Just press F12 right from the start, and it will be done in a single keypress. cl> I'd still rather have had a DOS program. Do you have a cheat sheet for the keys? me> No need - you can see them on screen all the time. If you see a button with an underlined "T" in its name, just hold ALT and press T to execute that function. [some other examples with other controls, menus, shortcut keys for menu choices, ESC key to cancel, etc.] cl> [etc etc DOS etc DOS DOS etc etc] me> [etc etc etc] Several weeks later, she has constantly been using the mouse, and she has constantly been complaining my ears off that it's not a DOS program, and that she'd rather use the keyboard. And I'm _still_ wondering why she doesn't. It's not that it doesn't work, because I use it myself more than half the time. Maybe some people just don't like changes, even _after_ they got used to them? -- "Managing senior programmers is like herding cats." --Dave Platt ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Thu, 10 Dec 1998 23:49:13 GMT Organization: . Lines: 30 Message-ID: <36734780.57168153@news.uunet.be> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d24c4.0@192.168.0.20> <74jmk1$1kvu$1@quince.news.easynet.net> <366e4f83.0@192.168.0.20> <74otna$oej$1@monteverdi.ednet.co.uk> NNTP-Posting-Host: pool02b-194-7-150-33.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!diablo.theplanet.net!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach "Derry Hamilton" on Thu, 10 Dec 1998 16:43:26 -0000 to alt.folklore.computers: > Alternativly, the number of clients (of all abilities) that have demanded we ^^^^^^^^^^^^^ > remove the blasted thing has made it SOP here to install Office without it. If they demanded you to remove it instead of clicking on that little X themselves, I wonder how broad that "all" is. Not that I think you're wrong - Bob wasn't that much of a success either. Personally, I don't have it on screen all the time, but I do click it ON when I can't remember where some seldom-used function is (I don't use office that much - in fact I only found out today how easy Word's [simple] 3d drawing capabilities are). It may be just an impression, but it feels like Einstein always gets me to the right answer a lot faster than the classical-style online help did (at first I refused to use this newfangled help too - I was much to computer literate for such a beginner's gadget). Now I'm wondering, would the paperclip give different answers than Einstein or the robot? -- "Managing senior programmers is like herding cats." --Dave Platt ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Thu, 10 Dec 1998 23:49:15 GMT Organization: . Lines: 20 Message-ID: <36744b53.58147201@news.uunet.be> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <1998Dec9.194507.18383@mbsks.franken.de> NNTP-Posting-Host: pool02b-194-7-150-33.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news-raspail.gip.net!news-lond.gip.net!news.gsl.net!gip.net!diablo.theplanet.net!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach m@mbsks.franken.de (Matthias Bruestle) on Wed, 9 Dec 1998 19:45:07 GMT to alt.folklore.computers: > If everybody could buy > a 1024 processor computer for US$1000 you can be sure there were > games which need these computers. Just wait a couple (?) of years, and that kind of power will be in every store. Games will come in 128GB write-inhibited flash rom discs (drives connected to your mainboard though an optical fiber bundle), run in multithreading virtual neuron basic interpreters, and draw every pixel of their 3-d image individually with PSET instructions. Or something like that. [No, I'm not kidding. At least not completely :)] -- "Managing senior programmers is like herding cats." --Dave Platt ###### From: deke.spamblock@generous.net Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Fri, 11 Dec 1998 03:18:36 GMT Organization: GenerousCity is a virtue - find romance at http://generous.net Lines: 29 Message-ID: <36708e38.95649957@news.bright.net> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <1998Dec9.194507.18383@mbsks.franken.de> <36744b53.58147201@news.uunet.be> NNTP-Posting-Host: paul-cas1-cs-38.dial.bright.net X-Newsreader: Forte Free Agent 1.11/32.235 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!news.maxwell.syr.edu!feeder1.wwnet.net!streamer1.cleveland.iagnet.net!NewsNG.Chicago.Qual.Net!news.bright.net!not-for-mail On Thu, 10 Dec 1998 23:49:15 GMT, lucvdv@null.net (Luc Van der Veken) wrote: >Also sprach m@mbsks.franken.de (Matthias Bruestle) on Wed, 9 Dec >1998 19:45:07 GMT to alt.folklore.computers: >> If everybody could buy >> a 1024 processor computer for US$1000 you can be sure there were >> games which need these computers. >Just wait a couple (?) of years, and that kind of power will be >in every store. #define couple 7 Yes. Including the corner donut shop. That chip with 1024 processors in it will be running the thermostat. Woz's law: sooner or later, every chip costs $1.75 deke ------------------------ Let love find you! http://generous.net A list for flirting GenerousSingles-subscribe@onelist.com Over The Hill Gang GenerousSinglesOTHG-subscribe@onelist.com Personal Ads list GenerousProfiles-subscribe@onelist.com ###### From: pg@sff.net (Paul Guertin) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Fri, 11 Dec 1998 10:24:05 GMT Organization: Amalgamated Karma Lines: 12 Message-ID: <3670c092.2894255@news.newsguy.com> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> NNTP-Posting-Host: p-562.newsdawg.com X-Newsreader: Forte Free Agent 1.11/32.235 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!news.idt.net!newsfeed.berkeley.edu!pln-w!spln!extra.newsguy.com!newsp.newsguy.com!news1 On 9 Dec 1998 01:30:25 GMT, Kin Hoong CHUNG wrote: >There are many things which require computing power than any game one >could imagine. [...] There are also >some people who think that factorising very large primes is an important >pursuit too. Ha! I can factor arbitrarily large primes in my head! Paul Guertin pg@sff.net ###### Date: 11 Dec 98 10:43:50 -0800 From: "Charlie Gibbs" Subject: Re: Multics References: <366e4f83.0@192.168.0.20> <1595.647T2126T6885175@sky.bus.com> <2631.648T2320T6605417@sky.bus.com> <36712536.0@192.168.0.20> Message-ID: <1111.649T2525T6435686@sky.bus.com> Newsgroups: alt.folklore.computers Lines: 28 X-Newsreader: THOR 2.5a (Amiga;TCP/IP) NNTP-Posting-Host: news.skybus.com X-Trace: 11 Dec 1998 13:33:57 -0800, news.skybus.com Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newshub.northeast.verio.net!news-pen-3.sprintlink.net!news-in-east1.sprintlink.net!news.sprintlink.net!news.westel.com!news.skybus.com!204.244.247.120 In article <36712536.0@192.168.0.20> Samael@dial.pipex.com (Samael) writes: >Charlie Gibbs wrote in message <2631.648T2320T6605417@sky.bus.com>... > >>Amen. The problems I referred to above occur when we're trying >>to get the user to start up our program, or look around in the >>system in other ways. One of our favourites is the luser who >>called up and said, "My TV went away!" After the usual game of >>"20 Questions", it turned out that he had opened another window >>which obscured ours. > >So you had a user who was working with an operating system he'd had >no training with. > >And you were surprised you had problems? Entertained, perhaps. Not suprised. After all, this is the Real World [tm], where software packages wind up in random mom-and-pop motels and get used by people who barely know what a computer is. But the marketroids say that Windoze is so easy to use that anyone can use it without training... and there are millions of people out there who are more than willing to believe them. -- cgibbs@sky.bus.com (Charlie Gibbs) Remove the first period after the "at" sign to reply. ###### From: "Samael" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Fri, 11 Dec 1998 13:58:56 -0000 Lines: 18 Message-ID: <36712536.0@192.168.0.20> References: <366e4f83.0@192.168.0.20> <1595.647T2126T6885175@sky.bus.com> <2631.648T2320T6605417@sky.bus.com> NNTP-Posting-Host: 195.147.140.180 X-Newsreader: Microsoft Outlook Express 4.72.3155.0 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3155.0 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsgate.cistron.nl!het.net!news-feed.inet.tele.dk!bofh.vszbr.cz!news-lond.gip.net!news.gsl.net!gip.net!nntp.news.xara.net!xara.net!news.itg.net.uk!192.168.0.20!192.168.0.1 Charlie Gibbs wrote in message <2631.648T2320T6605417@sky.bus.com>... >Amen. The problems I referred to above occur when we're trying >to get the user to start up our program, or look around in the >system in other ways. One of our favourites is the luser who >called up and said, "My TV went away!" After the usual game of >"20 Questions", it turned out that he had opened another window >which obscured ours. So you had a user who was working with an operating system he'd had no training with. And you were surprised you had problems? Samael ###### Newsgroups: alt.folklore.computers From: m@mbsks.franken.de (Matthias Bruestle) Subject: Re: Multics References: <365d422d.22238927@news.ricochet.net> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <366F0E61.275D@gazonk.del> <74pgse$nm0$1@sue.cc.uregina.ca> X-No-Archive: yes Organization: PDP-11/34A Date: Fri, 11 Dec 1998 17:12:02 GMT X-Newsreader: TIN [version 1.2 PL2] Message-ID: <1998Dec11.171202.12255@mbsks.franken.de> Lines: 31 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!194.165.93.117.MISMATCH!newshub.bart.net!newsfeed.wirehub.nl!btnet-peer!btnet!news-lond.gip.net!news.gsl.net!gip.net!rill.news.pipex.net!pipex!newsfeed.nacamar.de!uni-erlangen.de!chico.franken.de!mbsks.franken.de!m Mahlzeit John Bayko (bayko@pollux.cs.uregina.ca) wrote: > In article <366F0E61.275D@gazonk.del>, > Foobar T. Clown wrote: > [...] > >Don't forget that your PC has a minimum of two processors: One of > >them is the Pentium whatever, and the other one is the "accelerator" on > >your video adaptor. [...] > Probably more - your keyboard could have more computing power > than a VIC-20 did way back when (and probably a varient of the same CPU). Harddisk controllers (e.g. on an AHA1542C is a Z80) Modems (e.g. in the Zyxel 1496E+ is a 68000 with I think 16 MHz) Harddisks - most have a DSP and a microcontroller Keyboard - 8048? Mahlzeit endergone Zwiebeltuete -- PGP: SIG:C379A331 ENC:F47FA83D I LOVE MY PDP-11/34A, M70 and MicroVAXII! -- It's all *your* fault!! Yours, yours, yours!! ###### From: lucvdv@null.net (Luc Van der Veken) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Fri, 11 Dec 1998 20:14:42 GMT Organization: . Message-ID: <36716c84.909477@news.uunet.be> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <1998Dec9.194507.18383@mbsks.franken.de> <36744b53.58147201@news.uunet.be> <36708e38.95649957@news.bright.net> NNTP-Posting-Host: pool02b-194-7-80-228.uunet.be Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent 1.5/32.451 X-No-Archive: yes Lines: 10 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!masternews.telia.net!news-feed.inet.tele.dk!bofh.vszbr.cz!dispose.news.demon.net!demon!diablo.theplanet.net!join.news.pipex.net!pipex!krypton.inbe.net!INbe.net!not-for-mail Also sprach deke.spamblock@generous.net on Fri, 11 Dec 1998 03:18:36 GMT to alt.folklore.computers: > Woz's law: sooner or later, every chip costs $1.75 And inflation will make that come true even for potato chips. -- "Managing senior programmers is like herding cats." --Dave Platt ###### From: ftit@engin.umich.edu (Sergej Roytman) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 12 Dec 1998 00:36:22 GMT Organization: University of Michigan, College of Engineering Lines: 23 Message-ID: <74sdq6$jru@srvr1.engin.umich.edu> References: <365d422d.22238927@news.ricochet.net> <366d24c4.0@192.168.0.20> <74jmk1$1kvu$1@quince.news.easynet.net> <366e4f83.0@192.168.0.20> NNTP-Posting-Host: az.engin.umich.edu Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.cwix.com!128.230.129.106!news.maxwell.syr.edu!news.eecs.umich.edu!srvr1.engin.umich.edu!ftit In article <366e4f83.0@192.168.0.20>, Samael wrote: >barnacle wrote in message <74jmk1$1kvu$1@quince.news.easynet.net>... >>In article <366d24c4.0@192.168.0.20>, "Samael" >wrote: >>Fair comment Samael...but I see all this extra power and it's not used >the >>user...it's the bloody operating system (no names but it's small and >flexible) >>adding improved uselessness...do I really need dancing paperclips? >Yes, our office secretary, who finds it invaluable. >Also, my mum. >Other non-computer people who find it infinitely easier to interface >program that has an anthropomorphic front end. Barf. A guy at the lab suggested only today that a shareware game involving skeet-shooting at the little buggers would probably be very popular. Hm, I'm planning on refreshing my Java in a while, for work. Might be the thing... -- Sergej Roytman ###### From: jkatz@ultrix6.cs.csubak.edu (who?) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 12 Dec 1998 08:35:55 GMT Organization: California State University, Bakersfield Lines: 17 Message-ID: <74t9tb$35g$2@hades.csu.net> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <3670c092.2894255@news.newsguy.com> NNTP-Posting-Host: ultrix6.cs.csubak.edu X-Newsreader: TIN [version 1.2 PL2] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!howland.erols.net!usc!newshub.csu.net!ultrix6.cs.csubak.edu!jkatz Paul Guertin (pg@sff.net) wrote: : On 9 Dec 1998 01:30:25 GMT, Kin Hoong CHUNG : wrote: : >There are many things which require computing power than any game one : >could imagine. [...] There are also : >some people who think that factorising very large primes is an important : >pursuit too. : Ha! I can factor arbitrarily large primes in my head! : Paul Guertin : pg@sff.net In arbitrary time? jeremy ###### Reply-To: "Dennis J. Minette" From: "Dennis J. Minette" References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <3670c092.2894255@news.newsguy.com> <74t9tb$35g$2@hades.csu.net> Subject: Re: Multics Date: Sat, 12 Dec 1998 15:19:40 -0500 Lines: 24 Organization: Minette Data Systems, Inc. X-Newsreader: Microsoft Outlook Express 4.72.3155.0 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3155.0 Message-ID: <#CWCVzgJ#GA.119@upnetnews03> Newsgroups: alt.folklore.computers NNTP-Posting-Host: 1Cust49.tnt1.sarasota.fl.da.uu.net [153.37.162.49] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.cwix.com!207.68.152.14!upnetnews04!upnetnews03 who? wrote in message <74t9tb$35g$2@hades.csu.net>... >Paul Guertin (pg@sff.net) wrote: >: On 9 Dec 1998 01:30:25 GMT, Kin Hoong CHUNG >: wrote: > >: >There are many things which require computing power than any game one >: >could imagine. [...] There are also >: >some people who think that factorising very large primes is an important >: >pursuit too. > >: Ha! I can factor arbitrarily large primes in my head! > >: Paul Guertin >: pg@sff.net > >In arbitrary time? > >jeremy Most likely instantaneously! ###### From: genew@vip.net (Gene Wirchenko) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sat, 12 Dec 1998 22:08:51 GMT Organization: Posted via RemarQ, http://www.remarQ.com - Discussions start here! Lines: 28 Message-ID: <36721ef6.51255075@news.vip.net> References: <366e4f83.0@192.168.0.20> <1595.647T2126T6885175@sky.bus.com> <2631.648T2320T6605417@sky.bus.com> <36712536.0@192.168.0.20> <1111.649T2525T6435686@sky.bus.com> Reply-To: genew@vip.net NNTP-Posting-Host: 204.209.212.26 X-Trace: 913500378 A01OARAUVD41ACCD1C usenet57.supernews.com X-Complaints-To: newsabuse@remarQ.com X-Newsreader: Forte Free Agent 1.1/32.230 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.berkeley.edu!Supernews73!supernews.com!Supernews69!not-for-mail "Charlie Gibbs" wrote: >In article <36712536.0@192.168.0.20> Samael@dial.pipex.com (Samael) >writes: [snip] >>And you were surprised you had problems? > >Entertained, perhaps. Not suprised. After all, this is the >Real World [tm], where software packages wind up in random >mom-and-pop motels and get used by people who barely know >what a computer is. But the marketroids say that Windoze >is so easy to use that anyone can use it without training... >and there are millions of people out there who are more than >willing to believe them. Yes, it's SO intuitive: [shift-click], [control-click] obviously cast confusion spells at the user. Sincerely, Gene Wirchenko Computerese Irregular Verb Conjugation: I have preferences. You have biases. He/She has prejudices. ###### From: Kin Hoong CHUNG Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 12 Dec 1998 05:28:38 GMT Organization: University of New South Wales Lines: 15 Message-ID: <74suu6$fu$1@mirv.unsw.edu.au> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> NNTP-Posting-Host: alpha.maths.unsw.edu.au User-Agent: tin/pre-1.4-980618 (UNIX) (OSF1/V4.0 (alpha)) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news.mel.connect.com.au!news.syd.connect.com.au!nsw.nntp.telstra.net!news1.mpx.com.au!news1.optus.net.au!optus!news.usyd.edu.au!unsw.edu.au!not-for-mail Dann Lunsford wrote: : In article <74kjrh$jbn$2@mirv.unsw.edu.au>, Kin Hoong CHUNG wrote: :>There are also :>some people who think that factorising very large primes is an important :>pursuit too. : ^^^^^^^^^^^ ^^^^^^ : Hmmm... Bill Gates, posting under a pseudonym, via a .au relay? : Whoda thought? :-) Doh! Not Bill Gates, just me being careless. I meant factorising large numbers into primes. Cheers, Kin Hoong ###### From: "Joel C. Ewing" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Sun, 13 Dec 1998 20:48:24 -0600 Content-Transfer-Encoding: 7bit References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <74suu6$fu$1@mirv.unsw.edu.au> X-Posted-Path-Was: not-for-mail Content-Type: text/plain; charset=us-ascii X-ELN-Date: 14 Dec 1998 05:36:58 GMT X-ELN-Insert-Date: Sun Dec 13 21:45:03 1998 Organization: EarthLink Network, Inc. Lines: 25 Mime-Version: 1.0 NNTP-Posting-Host: 1cust207.tnt8.dfw5.da.uu.net Message-ID: <36747C78.3D32FDBA@acm.org> X-Mailer: Mozilla 4.05 [en] (Win95; U) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!isdnet!logbridge.uoregon.edu!newsfeed.berkeley.edu!newsfeed1.earthlink.net!nntp.earthlink.net!posted-from-earthlink!not-for-mail Kin Hoong CHUNG wrote: > Dann Lunsford wrote: > : In article <74kjrh$jbn$2@mirv.unsw.edu.au>, Kin Hoong CHUNG wrote: > :>There are also > :>some people who think that factorising very large primes is an important > :>pursuit too. > : ^^^^^^^^^^^ ^^^^^^ > : Hmmm... Bill Gates, posting under a pseudonym, via a .au relay? > : Whoda thought? :-) > > Doh! Not Bill Gates, just me being careless. I meant factorising large > numbers into primes. > Cheers, > Kin Hoong The utility of factoring large primes depends on your pursuit. From your remarks I judge you are unaware that tasks of "cracking" the typical public-key encryption schemes is closely interrelated with the task of factoring large primes. If your job involves espionage or national security, could be a highly important capability to possess. -- Joel C. Ewing, Fort Smith, AR jcewing@acm.org ###### From: mwilson@interlog.com (Mel Wilson) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 14 Dec 1998 11:56:41 -0500 Organization: Interlog Internet Services -Voice (416) 975-2655 -Data 515-1414 Lines: 13 Message-ID: References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <74suu6$fu$1@mirv.unsw.edu.au> <36747C78.3D32FDBA@acm.org> NNTP-Posting-Host: shell1.interlog.com NNTP-Posting-Time: 14 Dec 1998 16:56:42 GMT Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!howland.erols.net!news-peer1.sprintlink.net!news.sprintlink.net!news.maxwell.syr.edu!rill.news.pipex.net!pipex!newsfeed.interlog.com!news.interlog.com!not-for-mail In article <36747C78.3D32FDBA@acm.org>, "Joel C. Ewing" wrote: >The utility of factoring large primes depends on your pursuit. From >your remarks I judge you are unaware that tasks of "cracking" the >typical public-key encryption schemes is closely interrelated with the >task of factoring large primes. If your job involves espionage or >national security, could be a highly important capability to possess. Perhaps, but there are known algorithms for it. I can factor a large prime in logarithmic time if I have to write it down, or constant time if it's written down for me. Regards. Mel. ###### From: slanning@buphy.bu.edu (Scott Lanning) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 14 Dec 1998 15:37:28 GMT Organization: Boston Univ. Physics Lines: 13 Message-ID: <753bbo$hrv$1@news1.bu.edu> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <74suu6$fu$1@mirv.unsw.edu.au> <36747C78.3D32FDBA@acm.org> NNTP-Posting-Host: buphy.bu.edu X-Newsreader: TIN [version 1.2 PL2] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news.eecs.umich.edu!news.bu.edu!slanning Joel C. Ewing (jcewing@acm.org) wrote: : Kin Hoong CHUNG wrote: : > Doh! Not Bill Gates, just me being careless. I meant factorising : > large numbers into primes. : The utility of factoring large primes depends on your pursuit. From : your remarks I judge you are unaware that tasks of "cracking" the : typical public-key encryption schemes is closely interrelated with the : task of factoring large primes. If your job involves espionage or : national security, could be a highly important capability to possess. Duhhhhmmm, say I want to factor a large prime, call it p. Its factors are p and 1. Do you think NSA knows...? ###### From: tjmaxal@aol.com (TJMAXAL) Newsgroups: alt.folklore.computers Subject: Re: Multics Lines: 4 NNTP-Posting-Host: ladder03.news.aol.com X-Admin: news@aol.com Date: 15 Dec 1998 04:26:52 GMT Organization: AOL http://www.aol.com References: Message-ID: <19981214232652.01029.00002753@ng-ca1.aol.com> Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news-peer.gip.net!news.gsl.net!gip.net!portc01.blue.aol.com!audrey03.news.aol.com!not-for-mail Apple did come out with a basic It's called APPLE BASIC I have a manual for it and have used it before. TJMAXAL@AOL.COM "consider the source!" ###### From: Kin Hoong CHUNG Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 15 Dec 1998 04:44:27 GMT Organization: University of New South Wales Lines: 46 Message-ID: <754pfb$9nr$1@mirv.unsw.edu.au> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <74suu6$fu$1@mirv.unsw.edu.au> <36747C78.3D32FDBA@acm.org> NNTP-Posting-Host: alpha.maths.unsw.edu.au User-Agent: tin/pre-1.4-980618 (UNIX) (OSF1/V4.0 (alpha)) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!newspeer1.nac.net!news.maxwell.syr.edu!sunqbc.risq.qc.ca!news.uow.edu.au!news.usyd.edu.au!unsw.edu.au!not-for-mail Joel C. Ewing wrote: : Kin Hoong CHUNG wrote: :> :> Doh! Not Bill Gates, just me being careless. I meant factorising large :> numbers into primes. : The utility of factoring large primes depends on your pursuit. From : your remarks I judge you are unaware that tasks of "cracking" the : typical public-key encryption schemes is closely interrelated with the : task of factoring large primes. If your job involves espionage or : national security, could be a highly important capability to possess. I can't tell if this is a joke or not... If you meant prime factorisation, then because my email address is khchung@maths.unsw.edu.au, probability would dictate that I would know something of the interest in cracking public key encryption schemes. Not enough to know the details though. I consider this to be a curious** pursuit, as anybody who knows anything would use one-time secret, private key codes for the secure data (can be guaranteed uncrackable without the key)*** and RSA for the low security stuff only, as these are crackable with probability >0 without the key. The problem of finding suitably large, unexpected* (ie. not arising from a famous sequence like the Mersenne primes) primes is in itself hard enough (especially now that you really want primes bigger than 2^10) to be almost not worth the trouble. *Choosing a prime from a known series is just plain silly, as those are going to be the first ones anybody searches through. **Number theorists tend to find such things interesting, but I don't at the moment. This is not to say that these things are not interesting, just that I don't have much of an interest in it at the moment. ***One method of attack is to exhaust the key by intercepting all transmissions, then follow the person around, and try steal the new key. ObThread: It is interesting to note that in another thread in afc. there is some discussion about VMS and UNIX, both inspired, in part, by Multics, yet are so completely different. VMS has security rings, but presumably a very different filesystem structure, and UNIX has a more similar filesystem structure (right down to command names), but lacks the interesting features like security rings. What else have we inherited from Multics? Cheers, Kin Hoong ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 15 Dec 1998 15:14:05 GMT Lines: 19 Message-ID: <755ubt$quj$4@irk.zetnet.co.uk> References: <366F0E61.275D@gazonk.del> NNTP-Posting-Host: man-047.dialup.zetnet.co.uk X-Trace: irk.zetnet.co.uk 913734845 27603 194.247.41.58 (15 Dec 1998 15:14:05 GMT) NNTP-Posting-Date: 15 Dec 1998 15:14:05 GMT X-Everything: Net-Tamer V 1.08X Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!195.200.0.51.MISMATCH!newshub.bart.net!ayres.ftech.net!news.ftech.net!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-12-09 foobar@gazonk.del said: :I don't think those game systems can touch an up-to-date, multimedia :PC. Don't forget that your PC has a minimum of two processors: :One of them is the Pentium whatever, and the other one is the :"accelerator" on your video adaptor. I don't know what's in a :64-bit game console, but my guess is that any one of today's hot :video accelerators could beat it up without needing help from your :system's CPU. 3 processors; a CPU, a video rendering engine with support for textures and polygons in real time, and a DSP for sound. So I'm told. (Next door have a Net Yaroze.) -- Communa (lisard@zetnet.co.uk) -- you know soft spoken changes nothing Net-Tamer V 1.08X - Test Drive ###### From: "William \"Bill\" Carey" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Wed, 16 Dec 1998 00:17:08 -0500 Lines: 38 Message-ID: <757gau$7e1$1@winter.news.rcn.net> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <74suu6$fu$1@mirv.unsw.edu.au> <36747C78.3D32FDBA@acm.org> <754pfb$9nr$1@mirv.unsw.edu.au> X-Trace: P15LusRG4gFVK5CqH5K98zfKHP+Fqx1gXiV6gp3jQtw= X-Complaints-To: abuse@rcn.com NNTP-Posting-Date: 16 Dec 1998 05:26:54 GMT X-Newsreader: Microsoft Outlook Express 4.72.3110.1 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!news.tvd.be!newsfeed.wirehub.nl!howland.erols.net!master.news.rcn.net!not-for-mail -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Kin Hoong CHUNG wrote in message <754pfb$9nr$1@mirv.unsw.edu.au>... >Joel C. Ewing wrote: >: Kin Hoong CHUNG wrote: Another inheritance from Multics may be relational databases, since the first one available commercially was Multics Relational Data Store (MRDS) >ObThread: It is interesting to note that in another thread in afc. there >is some discussion about VMS and UNIX, both inspired, in part, by Multics, >yet are so completely different. VMS has security rings, but presumably >a very different filesystem structure, and UNIX has a more similar filesystem >structure (right down to command names), but lacks the interesting features >like security rings. What else have we inherited from Multics? > >Cheers, > >Kin Hoong -----BEGIN PGP SIGNATURE----- Version: PGPfreeware 6.0 for non-commercial use iQA/AwUBNndCVCaVPcZl+Z32EQL1UACfbq4e4+f+QNqzgNZTp+8ar/RqnvMAoJKL QckoT9PmhiwzOz4HngWwSrmm =G6Gw -----END PGP SIGNATURE----- ###### From: "Joel C. Ewing" Newsgroups: alt.folklore.computers Subject: Re: Multics Date: Wed, 16 Dec 1998 00:44:51 -0600 Content-Transfer-Encoding: 7bit References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <74suu6$fu$1@mirv.unsw.edu.au> <36747C78.3D32FDBA@acm.org> <753bbo$hrv$1@news1.bu.edu> X-Posted-Path-Was: not-for-mail Content-Type: text/plain; charset=us-ascii X-ELN-Date: 16 Dec 1998 06:45:41 GMT X-ELN-Insert-Date: Tue Dec 15 22:55:01 1998 Organization: EarthLink Network, Inc. Lines: 22 Mime-Version: 1.0 NNTP-Posting-Host: 1cust18.tnt8.dfw5.da.uu.net Message-ID: <367756E3.6B4C6A26@acm.org> X-Mailer: Mozilla 4.05 [en] (Win95; U) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!newsfeed.berkeley.edu!newsfeed1.earthlink.net!nntp.earthlink.net!posted-from-earthlink!not-for-mail Scott Lanning wrote: > > Joel C. Ewing (jcewing@acm.org) wrote: > : Kin Hoong CHUNG wrote: > > :>There are also > > :>some people who think that factorising very large primes is an important > :>pursuit too. > : > Doh! Not Bill Gates, just me being careless. I meant factorising > : > large numbers into primes. > > : The utility of factoring large primes depends on your pursuit. From ... > > Duhhhhmmm, say I want to factor a large prime, call it p. Its factors > are p and 1. Do you think NSA knows...? Proves that "mind rot" from too many hours in front of the tube is contagious. I knew Kin really meant "large numbers" but slavishly copied the original wording. I really know better! Mea culpa ... -- Joel C. Ewing, Fort Smith, AR jcewing@acm.org ###### From: slanning@buphy.bu.edu (Scott Lanning) Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 16 Dec 1998 21:56:49 GMT Organization: Boston Univ. Physics Lines: 4 Message-ID: <759ab1$jvp$1@news1.bu.edu> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <74suu6$fu$1@mirv.unsw.edu.au> <36747C78.3D32FDBA@acm.org> <753bbo$hrv$1@news1.bu.edu> <367756E3.6B4C6A26@acm.org> NNTP-Posting-Host: buphy.bu.edu X-Newsreader: TIN [version 1.2 PL2] Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!news.eecs.umich.edu!news.bu.edu!slanning Joel C. Ewing (jcewing@acm.org) wrote: : copied the original wording. I really know better! Mea culpa ... SURE, you do... ;) (just joshing, Joel....say that one quickly 5 times) ###### From: Kin Hoong CHUNG Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 17 Dec 1998 06:55:31 GMT Organization: University of New South Wales Lines: 15 Message-ID: <75a9t3$shk$1@mirv.unsw.edu.au> References: <365d422d.22238927@news.ricochet.net> <73koo5$lln$1@irk.zetnet.co.uk> <365fd57c.3293716@news.ricochet.net> <73ucrk$gg4@top.mitre.org> <366C8171.62693A07@loudoun.com> <74ilqp$fs1$1@quince.news.easynet.net> <366d2a29.12511127@news.bright.net> <74kjrh$jbn$2@mirv.unsw.edu.au> <74suu6$fu$1@mirv.unsw.edu.au> <36747C78.3D32FDBA@acm.org> <753bbo$hrv$1@news1.bu.edu> <367756E3.6B4C6A26@acm.org> NNTP-Posting-Host: alpha.maths.unsw.edu.au User-Agent: tin/pre-1.4-980618 (UNIX) (OSF1/V4.0 (alpha)) Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!news.maxwell.syr.edu!sunqbc.risq.qc.ca!news.uow.edu.au!news.usyd.edu.au!unsw.edu.au!not-for-mail Joel C. Ewing wrote: : Proves that "mind rot" from too many hours in front of the tube is : contagious. I knew Kin really meant "large numbers" but slavishly : copied the original wording. I really know better! Mea culpa ... That just about proves my worst fears: that _that_ Bill Gates speech contains a virus! It makes people replace "large numbers" with "primes" when talking about factorising large numbers... I know I was infected... so what is the antidote :-):-):-) Cheers, Kin Hoong ###### From: lisard@zetnet.co.uk Newsgroups: alt.folklore.computers Subject: Re: Multics Date: 17 Dec 1998 23:18:35 GMT Lines: 10 Message-ID: <75c3gb$cbd$6@roch.zetnet.co.uk> References: <757gau$7e1$1@winter.news.rcn.net> NNTP-Posting-Host: man-051.dialup.zetnet.co.uk X-Trace: roch.zetnet.co.uk 913936715 12653 194.247.41.63 (17 Dec 1998 23:18:35 GMT) NNTP-Posting-Date: 17 Dec 1998 23:18:35 GMT X-Everything: Net-Tamer V 1.08X Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!news.belnet.be!newsfeed.wirehub.nl!ayres.ftech.net!news.ftech.net!peer.news.zetnet.net!zetnet.co.uk!not-for-mail On 1998-12-16 wmcarey@erols.com said: :Another inheritance from Multics may be relational databases, since :the first one available commercially was Multics Relational Data :Store (MRDS) Did this precede the fellow at IBM theorising about it? (Codd?) -- Communa (lisard@zetnet.co.uk) -- you know soft spoken changes nothing ###### Date: Fri, 18 Dec 1998 12:38:23 -0800 From: thvv@multicians.org (Tom Van Vleck) Newsgroups: alt.folklore.computers Subject: Re: Multics Message-ID: References: <757gau$7e1$1@winter.news.rcn.net> <75c3gb$cbd$6@roch.zetnet.co.uk> Organization: Multicians X-Newsreader: MT-NewsWatcher 2.4.1 X-url: http://www.multicians.org/ Lines: 21 NNTP-Posting-Host: thvv.vip.best.com X-Trace: nntp1.ba.best.com 914013502 230 206.86.18.21 Path: chonsp.franklin.ch!pfaff.ethz.ch!news-zh.switch.ch!newsfeed-zh.ip-plus.net!news-nyc.telia.net!newsfeed.berkeley.edu!news2.best.com!news3.best.com!nntp1.ba.best.com!not-for-mail In article <75c3gb$cbd$6@roch.zetnet.co.uk>, lisard@zetnet.co.uk wrote: > On 1998-12-16 wmcarey@erols.com said: > :Another inheritance from Multics may be relational databases, since > :the first one available commercially was Multics Relational Data > :Store (MRDS) > > Did this precede the fellow at IBM theorising about it? (Codd?) Nope. Codd had theorized, and System R had been built. The MRDS group (which I managed for a while) read the papers and implemented a Multics relational database, and got it to market first. See http://www.mcjones.org/System_R/ for more info on IBM's early relational database systems. Note that Multics had an even earlier relational database, not a commercial product. The ADMINS system built by Jay Goldman for MIT Prof. Jim Bruce was used to run MIT's EE department in the mid 70s. But this was never a product.