Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!cpk-news-hub1.bbnplanet.com!su-news-hub1.bbnplanet.com!news.bbnplanet.com!news.Stanford.EDU!kithrup.com!cygnus.com!cygnus.com!not-for-mail From: billm@cygnus.com (Bill Moyer) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Growth of space and density, and some wild-ass predictions Date: 17 Apr 1998 19:57:54 -0700 Organization: Subtle, but there. Lines: 66 Distribution: usa Message-ID: <6h94ri$av3$1@andros.cygnus.com> References: <6h0tia$iri$1@owl.slip.net> <3537832D.3BF8B968@hobbes.dtcc.edu> <3537F0FC.3489@cisco.com> NNTP-Posting-Host: andros.cygnus.com In article <3537F0FC.3489@cisco.com> John Ahlstrom writes: > >In an age when RAM was about 1$ per kilo byte, bit addressability >made a lot of sense. Today with RAM at < $1 per megabyte, it >doesn't. A minor nit to pick: DRAM has been hanging around $1.3-1.5 per MB for a while. But your main point still stands. Today, systems are even getting away from byte addressability and leaning towards word- sized granularity. I've been collecting information about the costs, sizes, and capabilities of computer systems since I was a boy (I wanted to see if reality reflected Moore's Law), and found that the rates of growth of capability-per-dollar has been more or less stable. Transistor density: 2x every 24 months. Secondary storage density: 2.5x every 24 months. There was a burble in the storage/price curve around the time that 540MB ATA disks were insufficient for home computer users' needs and ATA2 had not yet emerged. The prices of the largest (540MB) ATA drives were inflated, throwing off the curve. After ATA2 drives appeared on the scene, however, the prices of the large ATA drives dropped sharply until they were where they were "supposed" to be, then started following the 2.5x/24mo curve again. Whether or not these trends will continue is another matter, but I won't get into that here. I'd like to throw out a table for popular consumption which shows you what you could buy for a given amount of $$$ as the years go by, assuming these trends *do* continue: $100 $250 Price of Alpha Price of Intel DRAM HD 533MHz 21164 300MHz PII 1998 64 MB 6 GB $495 $400 1999 96 MB 9.5 GB $355 $285 2000 128 MB 15 GB $255 $205 2001 192 MB 24 GB $180 $145 2002 256 MB 37 GB $130 $105 2003 384 MB 60 GB 2004 512 MB 92 GB 2005 768 MB 150 GB 2006 1 GB 234 GB 2007 1.4 GB 375 GB 2008 2.0 GB 585 GB 2009 2.8 GB 937 GB 2010 4.0 GB 1.4 TB Note on CPU models: It gets really hard to find processors once they've dropped below $100/unit. Eventually they end up in surplus outlets for $10 (including motherboard), but it's seldom worth buying at that point. On the other hand, $2 microprocessors are common in the embedded systems market. Perhaps in 2010, low-power versions of the 500MHz Alpha could be had for $10 apiece, inte- grated with 64MB of RAM, 16MB of PROM, and a gigabit ethernet controller? The numbers run, and there's already some precedent in today's embedded market (qv PIC and MPC850). Anyone remember what ethernet cards cost ten years ago? Nowadays you can get an NE2K clone that'll do 600KB/s bulk transfer for $12. -- Bill Moyer ###### Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!news-peer-east.sprintlink.net!news.sprintlink.net!Sprint!ix.netcom.com!eee From: eee@netcom.com (Mark Thorson) Subject: Re: Growth of space and density, and some wild-ass predictions Message-ID: Organization: Netcom References: <6h0tia$iri$1@owl.slip.net> <3537832D.3BF8B968@hobbes.dtcc.edu> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> Distribution: usa Date: Sun, 19 Apr 1998 21:34:52 GMT Lines: 17 Sender: eee@netcom11.netcom.com In article <6h94ri$av3$1@andros.cygnus.com>, Bill Moyer wrote: > >Note on CPU models: > It gets really hard to find processors once they've dropped >below $100/unit. Eventually they end up in surplus outlets for >$10 (including motherboard), but it's seldom worth buying at >that point. On the other hand, $2 microprocessors are common >in the embedded systems market. Perhaps in 2010, low-power >versions of the 500MHz Alpha could be had for $10 apiece, inte- >grated with 64MB of RAM, 16MB of PROM, and a gigabit ethernet >controller? The numbers run, and there's already some precedent >in today's embedded market (qv PIC and MPC850). But how much will it cost to run Windows NT server on it? (And if you can't run that, what good is it?) ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!news-peer.sprintlink.net!news-backup-east.sprintlink.net!news-in-east.sprintlink.net!news.sprintlink.net!Sprint!169.197.1.22!feeder.news.azstarnet.com!reader1.news.azstarnet.com!news From: bill_h Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Sun, 19 Apr 1998 22:32:07 -0700 Organization: Starnet Lines: 9 Message-ID: <353ADDD7.63B2@azstarnet.com> References: <6h0tia$iri$1@owl.slip.net> <3537832D.3BF8B968@hobbes.dtcc.edu> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> NNTP-Posting-Host: 169.197.35.154 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 2.0 (Win16; I) Another curious thing: It generally seems to take 5-7 YEARS to actually reach a point where we're getting something approaching optimal use of the features of a given processor generation. Yet Intel, partly for profit reasons, is more or less trapped onto an 18month-24month per generation treadmill. They have been able to maintain a lead, during which they can command relatively high profits. But, for how long? I'm sure Intel would like to know, too. ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!howland.erols.net!feed2.news.erols.com!erols!winter.news.erols.com!not-for-mail From: "GMenke" Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 20 Apr 1998 00:19:32 GMT Organization: .. Lines: 28 Message-ID: <01bd6bf2$0c016e80$64646464@gregm> References: <6h0tia$iri$1@owl.slip.net> <3537832D.3BF8B968@hobbes.dtcc.edu> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> NNTP-Posting-Host: 207-172-130-96.s33.as10.col.erols.com X-Trace: winter.news.erols.com 893031572 25895 207.172.130.96 (20 Apr 1998 00:19:32 GMT) X-Complaints-To: abuse@erols.com X-Newsreader: Microsoft Internet News 4.70.1162 Why in the world would you want to run NT anyhow? Gregm -- To reply, replace nospam with menkesjg. Thanks. Mark Thorson wrote in article ... > In article <6h94ri$av3$1@andros.cygnus.com>, > Bill Moyer wrote: > > > >Note on CPU models: > > It gets really hard to find processors once they've dropped > >below $100/unit. Eventually they end up in surplus outlets for > >$10 (including motherboard), but it's seldom worth buying at > >that point. On the other hand, $2 microprocessors are common > >in the embedded systems market. Perhaps in 2010, low-power > >versions of the 500MHz Alpha could be had for $10 apiece, inte- > >grated with 64MB of RAM, 16MB of PROM, and a gigabit ethernet > >controller? The numbers run, and there's already some precedent > >in today's embedded market (qv PIC and MPC850). > > But how much will it cost to run Windows NT server on it? > (And if you can't run that, what good is it?) > > ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!enews.sgi.com!su-news-feed4.bbnplanet.com!news.bbnplanet.com!forum.apple.com!news.apple.com!handma.apple.com!user From: handleym@apple.com (Maynard Handley) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Mon, 20 Apr 1998 07:06:55 -0700 Organization: Apple Computer Lines: 38 Message-ID: References: <6h0tia$iri$1@owl.slip.net> <3537832D.3BF8B968@hobbes.dtcc.edu> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <353ADDD7.63B2@azstarnet.com> NNTP-Posting-Host: handma.apple.com X-Newsreader: MT-NewsWatcher 2.3.1 In article <353ADDD7.63B2@azstarnet.com>, bill_h wrote: > Another curious thing: It generally seems to take 5-7 YEARS to actually > reach a point where we're getting something approaching optimal use of > the features of a given processor generation. Yet Intel, partly for > profit reasons, is more or less trapped onto an 18month-24month per > generation treadmill. They have been able to maintain a lead, during > which they can command relatively high profits. > > But, for how long? I'm sure Intel would like to know, too. This is not just a feature of hardware. I remain convinced that a major reason for the success of MS is that they have persistence. Or to put it another way-a major reason for the failure of Apple is that they lack persistence. Apple has a long history of introducing software technologies, being amazed that they aren't part of every major app after 3 months, and cancelling them after 6 months. MS have had a few cancellations of system software, but for the most part they simply keep plugging away at it. One could argue the same obsession with 3 months from now is what has doomed UNIX/X---no-one in that part of the business has been prepared to make changes that will have no immediate impact---but they will make life a whole lot easier in 5 yrs. The reasons for this are of course obvious to anyone. A new idea can't have any impact on this round of the product, so it only gets a chance to be LOOKED at in once rev 3.0 is out the door. For rev 4.0 we'll try it out in small doses to see what the possible problems are, and it's only with rev 5.0 that we start writing 3 yrs from now that we actually incorporate the innovation in a serious way. Regardless of how much we may mock Mr Gate's coding skills, perhaps the fact that he DID actually code once, rather more so than any Apple CEO, puts him more in touch with reality. Maynard -- My opinion only ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!enews.sgi.com!su-news-feed4.bbnplanet.com!news.bbnplanet.com!forum.apple.com!news.apple.com!handma.apple.com!user From: handleym@apple.com (Maynard Handley) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Mon, 20 Apr 1998 07:06:55 -0700 Organization: Apple Computer Lines: 38 Message-ID: References: <6h0tia$iri$1@owl.slip.net> <3537832D.3BF8B968@hobbes.dtcc.edu> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <353ADDD7.63B2@azstarnet.com> NNTP-Posting-Host: handma.apple.com X-Newsreader: MT-NewsWatcher 2.3.1 In article <353ADDD7.63B2@azstarnet.com>, bill_h wrote: > Another curious thing: It generally seems to take 5-7 YEARS to actually > reach a point where we're getting something approaching optimal use of > the features of a given processor generation. Yet Intel, partly for > profit reasons, is more or less trapped onto an 18month-24month per > generation treadmill. They have been able to maintain a lead, during > which they can command relatively high profits. > > But, for how long? I'm sure Intel would like to know, too. This is not just a feature of hardware. I remain convinced that a major reason for the success of MS is that they have persistence. Or to put it another way-a major reason for the failure of Apple is that they lack persistence. Apple has a long history of introducing software technologies, being amazed that they aren't part of every major app after 3 months, and cancelling them after 6 months. MS have had a few cancellations of system software, but for the most part they simply keep plugging away at it. One could argue the same obsession with 3 months from now is what has doomed UNIX/X---no-one in that part of the business has been prepared to make changes that will have no immediate impact---but they will make life a whole lot easier in 5 yrs. The reasons for this are of course obvious to anyone. A new idea can't have any impact on this round of the product, so it only gets a chance to be LOOKED at in once rev 3.0 is out the door. For rev 4.0 we'll try it out in small doses to see what the possible problems are, and it's only with rev 5.0 that we start writing 3 yrs from now that we actually incorporate the innovation in a serious way. Regardless of how much we may mock Mr Gate's coding skills, perhaps the fact that he DID actually code once, rather more so than any Apple CEO, puts him more in touch with reality. Maynard -- My opinion only ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!enews.sgi.com!news.sgi.com!cygnus.com!cygnus.com!not-for-mail From: billm@cygnus.com (Bill Moyer) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 20 Apr 1998 14:01:44 -0700 Organization: Subtle, but there. Lines: 38 Distribution: usa Message-ID: <6hgd3o$pvu$1@andros.cygnus.com> References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> NNTP-Posting-Host: andros.cygnus.com In article eee@netcom.com (Mark Thorson) writes: >In article <6h94ri$av3$1@andros.cygnus.com>, >Bill Moyer wrote: >> >>Note on CPU models: >> It gets really hard to find processors once they've dropped >>below $100/unit. Eventually they end up in surplus outlets for >>$10 (including motherboard), but it's seldom worth buying at >>that point. On the other hand, $2 microprocessors are common >>in the embedded systems market. Perhaps in 2010, low-power >>versions of the 500MHz Alpha could be had for $10 apiece, inte- >>grated with 64MB of RAM, 16MB of PROM, and a gigabit ethernet >>controller? The numbers run, and there's already some precedent >>in today's embedded market (qv PIC and MPC850). > >But how much will it cost to run Windows NT server on it? >(And if you can't run that, what good is it?) No more than it costs to run WNT now: your money, your sanity, some functionality and flexibility, and a blackened bit of your eternal soul. To more objectively address your question: * Since Alpha runs WNT now, it should continue to run it in the future (even if it ceases to be supported under newer versions of the OS). * The low-cost processors tend to be for embedded applications. If it's an embedded processor, then you would probably want to run an embedded system OS on it, not WNT. As an aside: you are joking, right? I know that there are people who think that AOL invented the internet, but are there also people who think that WNT is the only useful OS? -- Bill Moyer ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!cam-news-hub1.bbnplanet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!nntp2.dejanews.com!nnrp1.dejanews.com!not-for-mail From: spaik@simagic.com Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Mon, 20 Apr 1998 21:51:27 -0600 Organization: Deja News - The Leader in Internet Discussion Lines: 13 Message-ID: <6hh1jf$4ks$1@nnrp1.dejanews.com> References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <6hgd3o$pvu$1@andros.cygnus.com> NNTP-Posting-Host: 208.1.239.152 X-Article-Creation-Date: Tue Apr 21 02:51:27 1998 GMT X-Http-User-Agent: Mozilla/3.01 (WinNT; I) In article <6hgd3o$pvu$1@andros.cygnus.com>, billm@cygnus.com (Bill Moyer) wrote: > * The low-cost processors tend to be for embedded applications. > If it's an embedded processor, then you would probably want > to run an embedded system OS on it, not WNT. Of course, Microsoft is trying to push Windows NT as an embedded (and real time) OS. Sam -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!newspump.monmouth.com!newspeer.monmouth.com!woodstock.news.demon.net!demon!nntprelay.mathworks.com!howland.erols.net!newsxfer.itd.umich.edu!gumby!newspump.wustl.edu!rice!snowy.owlnet.rice.edu!maliyekk From: maliyekk@rice.edu (Anil Thomas Maliyekkel) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Followup-To: alt.folklore.computers,comp.sys.unisys,comp.arch Date: 20 Apr 1998 22:18:13 GMT Organization: Rice University, Houston, Texas Distribution: usa Message-ID: <6hghj5$6gl$1@joe.rice.edu> References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <6hgd3o$pvu$1@andros.cygnus.com> NNTP-Posting-Host: snowy.owlnet.rice.edu X-Newsreader: TIN [version 1.2 PL2] Lines: 12 Bill Moyer (billm@cygnus.com) wrote: : As an aside: you are joking, right? I know that there are : people who think that AOL invented the internet, but are there : also people who think that WNT is the only useful OS? : -- Bill Moyer Unfortunately for some people, the definition of a useful OS is that it runs all the latest and greatest Microsoft applications. ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!news.voicenet.com!news.idt.net!newsfeed.internetmci.com!141.211.144.13!newsxfer3.itd.umich.edu!news.eecs.umich.edu!newshub.tc.umn.edu!skypoint.com!not-for-mail From: rsteiner@skypoint.com (Richard Steiner) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Mon, 20 Apr 1998 22:18:31 -0500 Organization: FIELDATA FORTRAN ENTHUSIASTS CLUB Lines: 16 Message-ID: References: <6h0tia$iri$1@owl.slip.net> <3537832D.3BF8B968@hobbes.dtcc.edu> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <01bd6bf2$0c016e80$64646464@gregm> Reply-To: rsteiner@skypoint.com NNTP-Posting-Host: dial546.skypoint.net X-Newsreader: Yarn for OS/2 v0.92 X-Stuff-Running: There are 42 Processes with 171 Threads. X-HomePage: http://www.skypoint.com/~rsteiner Here in comp.sys.unisys, Craig Burley spake unto us, saying: >I know this because I saw it on CNBC a few weeks ago during their >financial coverage. One of the talking heads, in explaining what >a workstation was, said something basically like "a workstation >is a computer that runs Windows NT". > >So, if you want a workstation, you gotta run Windows NT. Wow. I learn something new every day. :-) :-) -- -Rich Steiner >>>---> rsteiner@skypoint.com >>>---> Bloomington, MN OS/2 Warp 4 + Linux + Executor = PC Hobbyist Heaven! PHASE 5 SYMBOL TABLE OVERFLOW - REDUCE PROGRAM SIZE ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!howland.erols.net!feed2.news.erols.com!erols!winter.news.erols.com!not-for-mail From: "GMenke" Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 21 Apr 1998 03:48:46 GMT Organization: .. Lines: 53 Message-ID: <01bd6cd8$756b6470$64646464@gregm> References: <6h0tia$iri$1@owl.slip.net> <3537832D.3BF8B968@hobbes.dtcc.edu><3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <01bd6bf2$0c016e80$64646464@gregm> NNTP-Posting-Host: 207-172-129-106.s43.as6.col.erols.com X-Trace: winter.news.erols.com 893130526 15894 207.172.129.106 (21 Apr 1998 03:48:46 GMT) X-Complaints-To: abuse@erols.com X-Newsreader: Microsoft Internet News 4.70.1162 Well, I use NT every day- thats why I'd never want to use it if given the choice..... More reliable than Win 95, but then again, my Atari 800 is more reliable. What I want is DOS 7- without Win 95. Seems to me I got everything done I needed to do under Win 3.1- plus I could make the damn thing go away when appropriate. Gregm -- To reply, replace nospam with menkesjg. Thanks. Craig Burley wrote in article ... > "GMenke" writes: > > > Why in the world would you want to run NT anyhow? > > Because you might need a workstation instead of just a PC. > > I know this because I saw it on CNBC a few weeks ago during their > financial coverage. One of the talking heads, in explaining what > a workstation was, said something basically like "a workstation > is a computer that runs Windows NT". > > So, if you want a workstation, you gotta run Windows NT. > > After all, the guy couldn't be wrong -- he's on TV. And he > couldn't just be promoting Microsoft products as if they're all that > matter -- he's on CNBC, which has nothing to do with MSNBC, the > Microsoft channel. (They've got similar initials, but if you > look at what they stand for, they're entirely separate National > Broadcasting Companies. Trust me.) > > > As far as what to call machines that run Solaris, HP-UX, Digital UNIX, > and so on? I guess they're called "servers" now. > > And, for years, I've known what to call my Linux systems (whether > running on 486, Pentium II, or Alpha chips): > > Hackstations > > ;-) > > -- > > "Practice random senselessness and act kind of beautiful." > James Craig Burley, Software Craftsperson burley@gnu.org > ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!howland.erols.net!feed2.news.erols.com!erols!winter.news.erols.com!not-for-mail From: "GMenke" Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 21 Apr 1998 03:50:17 GMT Organization: .. Lines: 28 Message-ID: <01bd6cd8$abc2fba0$64646464@gregm> References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <6hgd3o$pvu$1@andros.cygnus.com> <6hh1jf$4ks$1@nnrp1.dejanews.com> NNTP-Posting-Host: 207-172-129-106.s43.as6.col.erols.com X-Trace: winter.news.erols.com 893130617 15894 207.172.129.106 (21 Apr 1998 03:50:17 GMT) X-Complaints-To: abuse@erols.com X-Newsreader: Microsoft Internet News 4.70.1162 And next they'll say Win 32 is good for real time? God I hope I never fly on that airplane.... Sheesh. MFC on an embedded PC? It gives me shivers...! Gregm -- To reply, replace nospam with menkesjg. Thanks. spaik@simagic.com wrote in article <6hh1jf$4ks$1@nnrp1.dejanews.com>... > In article <6hgd3o$pvu$1@andros.cygnus.com>, billm@cygnus.com (Bill Moyer) > wrote: > > * The low-cost processors tend to be for embedded applications. > > If it's an embedded processor, then you would probably want > > to run an embedded system OS on it, not WNT. > > Of course, Microsoft is trying to push Windows NT as an embedded > (and real time) OS. > > Sam > > -----== Posted via Deja News, The Leader in Internet Discussion ==----- > http://www.dejanews.com/ Now offering spam-free web-based newsreading > ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!ubnnews.unisource.ch!News.Amsterdam.UnisourceCS!newshunter.cosy.sbg.ac.at!cosy.sbg.ac.at!Cabal.CESspool!bofh.vszbr.cz!howland.erols.net!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!nntp2.dejanews.com!nnrp1.dejanews.com!not-for-mail From: spaik@simagic.com Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Tue, 21 Apr 1998 13:37:49 -0600 Organization: Deja News - The Leader in Internet Discussion Lines: 19 Message-ID: <6hip1t$9ri$1@nnrp1.dejanews.com> References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <6hgd3o$pvu$1@andros.cygnus.com> <6hh1jf$4ks$1@nnrp1.dejanews.com> <01bd6cd8$abc2fba0$64646464@gregm> NNTP-Posting-Host: 208.1.239.152 X-Article-Creation-Date: Tue Apr 21 18:37:49 1998 GMT X-Http-User-Agent: Mozilla/3.01 (WinNT; I) In article <01bd6cd8$abc2fba0$64646464@gregm>, "GMenke" wrote: > And next they'll say Win 32 is good for real time? God I hope I never fly > on that airplane.... Clearly neither the Windows 95 core nor Windows NT kernel is real time currently. I understand Microsoft is working with some other companies to add real time extensions to the NT kernel. As for embedded systems, I think the NT kernel is rather heavyweight for most uses. It looks like Microsoft is now pushing the Windows CE for embedded/set top applications. Sam who thinks neither Windows NT nor Windows CE is a very good choice for real time or embedded systems, currently. -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!newsfeed.internetmci.com!141.211.144.13!newsxfer3.itd.umich.edu!jobone!dailyplanet.srl.ford.com!eccws1.dearborn.ford.com!longhorn!tph From: tph@longhorn.uucp (Tom Harrington) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Followup-To: alt.folklore.computers,comp.sys.unisys,comp.arch Date: 21 Apr 1998 15:22:51 GMT Organization: Mechanist Industries Lines: 21 Message-ID: <6hidkb$moa2@eccws1.dearborn.ford.com> References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <6hgd3o$pvu$1@andros.cygnus.com> <6hh1jf$4ks$1@nnrp1.dejanews.com> Reply-To: tph@rmi.net NNTP-Posting-Host: cs0053.eld.ford.com X-Newsreader: TIN [version 1.2 PL2] spaik@simagic.com wrote: : In article <6hgd3o$pvu$1@andros.cygnus.com>, billm@cygnus.com (Bill Moyer) : wrote: : > * The low-cost processors tend to be for embedded applications. : > If it's an embedded processor, then you would probably want : > to run an embedded system OS on it, not WNT. : Of course, Microsoft is trying to push Windows NT as an embedded : (and real time) OS. NT? You sure? I've heard about their grand plan to push WinCE as a real-time OS. But NT? Isn't that more than a bit bloated for an embedded system? I'd think that even MS would have trouble convincing people to cram NT into most embedded applications, unless they can really strip it down a lot. -- Tom Harrington --------- tph@rmii.com -------- http://rainbow.rmii.com/~tph "Moses parted the red sea. Oppenheimer split the atom. But "Bob" cut the crap." -Rev. Steve Antczak, Revelation X: The "Bob" Apocryphon Cookie's Revenge: ftp://ftp.rmi.net/pub2/tph/cookie/cookies-revenge.sit.hqx ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsfeed.internetmci.com!128.174.5.49!vixen.cso.uiuc.edu!newsrelay.iastate.edu!news.iastate.edu!rhawkins From: rhawkins@iastate.edu (Rick Hawkins) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 21 Apr 1998 16:38:50 GMT Organization: Iowa State University, Ames, Iowa USA Lines: 19 Message-ID: <6hii2q$auv$1@news.iastate.edu> References: <6h0tia$iri$1@owl.slip.net> <01bd6bf2$0c016e80$64646464@gregm> <01bd6cd8$756b6470$64646464@gregm> NNTP-Posting-Host: pv2086.vincent.iastate.edu In article <01bd6cd8$756b6470$64646464@gregm>, GMenke wrote: > >Well, I use NT every day- thats why I'd never want to use it if given the >choice..... More reliable than Win 95, but then again, my Atari 800 is >more reliable. > >What I want is DOS 7- without Win 95. >Seems to me I got everything done I needed to do under Win 3.1- plus I >could make the damn thing go away when appropriate. what about caldera dos? that should do it, shouldn't it? or opendos, or whatever it's called this week. -- R E HAWKINS rhawkins@iastate.edu These opinions will not be those of ISU until they pay my retainer. ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!newspump.monmouth.com!newspeer.monmouth.com!newsfeed1.earthlink.net!uwm.edu!uwvax!news From: Andy Glew Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Tue, 21 Apr 1998 17:37:45 -0500 Organization: U Wisc CS (& Intel) Lines: 12 Message-ID: <353D1FB9.EB21AD79@cs.wisc.edu> References: <6h0tia$iri$1@owl.slip.net> <01bd6cd8$abc2fba0$64646464@gregm> <6hip1t$9ri$1@nnrp1.dejanews.com> <6hir9n$ero$1@lyra.csx.cam.ac.uk> <6hj6fv$c2e$1@news.iastate.edu> NNTP-Posting-Host: helga.cs.wisc.edu Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 4.03 [en] (WinNT; I) To: John Hascall > This raises the counter question -- is it possible > to do the opposite -- start with a real time kernel > and build a general O.S. on top of it and end up with > something useful. (1) QNX. Started as a RT kernel with limited functionality. Now has pretty much anything from UNIX I want to use. (2) As for Non-RT -> RT --- I always thought that Gould Real Time UNIX was a pleasant system to work on. :-( ###### Message-ID: <353C5BA5.16A0FE68@Japan.NCR.COM> Date: Tue, 21 Apr 1998 17:41:09 +0900 From: Eric Hildum Organization: NCR Japan X-Mailer: Mozilla 4.04 [en] (Win95; I) MIME-Version: 1.0 Subject: Re: Growth of space and density, and some wild-ass predictions References: <6h0tia$iri$1@owl.slip.net> <3537832D.3BF8B968@hobbes.dtcc.edu> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> Content-Type: text/plain; charset=iso-2022-jp Content-Transfer-Encoding: 7bit Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Lines: 27 Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsfeed.internetmci.com!204.71.0.51!spamkiller.internetmci.com!pull-feed.internetmci.com!rpc1283.ncr.com!rpc1286!rpc1284.daytonoh.ncr.com Mark Thorson wrote: > In article <6h94ri$av3$1@andros.cygnus.com>, > Bill Moyer wrote: > > > >Note on CPU models: > > It gets really hard to find processors once they've dropped > >below $100/unit. Eventually they end up in surplus outlets for > >$10 (including motherboard), but it's seldom worth buying at > >that point. On the other hand, $2 microprocessors are common > >in the embedded systems market. Perhaps in 2010, low-power > >versions of the 500MHz Alpha could be had for $10 apiece, inte- > >grated with 64MB of RAM, 16MB of PROM, and a gigabit ethernet > >controller? The numbers run, and there's already some precedent > >in today's embedded market (qv PIC and MPC850). > > But how much will it cost to run Windows NT server on it? > (And if you can't run that, what good is it?) Is that really the question, or whether the processor is so fast that the (short) mean time between crashes becomes a significant issue... ;-) -- --------------------------- Eric Hildum Eric.Hildum@Japan.NCR.COM ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-dc.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsfeed.nacamar.de!fu-berlin.de!Cabal.CESspool!bofh.vszbr.cz!pegasus.csx.cam.ac.uk!lyra.csx.cam.ac.uk!nmm1 From: nmm1@cus.cam.ac.uk (Nick Maclaren) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 21 Apr 1998 19:16:07 GMT Organization: University of Cambridge, England Lines: 31 Message-ID: <6hir9n$ero$1@lyra.csx.cam.ac.uk> References: <6h0tia$iri$1@owl.slip.net> <6hh1jf$4ks$1@nnrp1.dejanews.com> <01bd6cd8$abc2fba0$64646464@gregm> <6hip1t$9ri$1@nnrp1.dejanews.com> NNTP-Posting-Host: ursa.cus.cam.ac.uk In article <6hip1t$9ri$1@nnrp1.dejanews.com>, wrote: >In article <01bd6cd8$abc2fba0$64646464@gregm>, "GMenke" >wrote: >> And next they'll say Win 32 is good for real time? God I hope I never fly >> on that airplane.... > >Clearly neither the Windows 95 core nor Windows NT kernel is real time >currently. I understand Microsoft is working with some other >companies to add real time extensions to the NT kernel. 40 years of experience has shown that adding "real-time extensions" to a general-purpose operating system merely adds complexity, and doesn't do anything useful. There are some fairly convincing theoretical arguments that this is an immutable law of nature. Unfortunately, the combination of experience and theory is insufficient to prevent idiots from repeating the mistakes of the past. >As for embedded systems, I think the NT kernel is rather heavyweight for >most uses. It looks like Microsoft is now pushing the Windows CE for >embedded/set top applications. It really gives me a warm, cosy feeling of safety to think that my life and finances are going to be under the control of a component with the proven reliability record of the Windows CE :-( Nick Maclaren, University of Cambridge Computing Service, New Museums Site, Pembroke Street, Cambridge CB2 3QG, England. Email: nmm1@cam.ac.uk Tel.: +44 1223 334761 Fax: +44 1223 334679 ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!howland.erols.net!newsxfer.itd.umich.edu!newsrelay.iastate.edu!news.iastate.edu!john From: john@iastate.edu (John Hascall) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 21 Apr 1998 22:27:11 GMT Organization: Iowa State University, Ames, Iowa USA Lines: 24 Message-ID: <6hj6fv$c2e$1@news.iastate.edu> References: <6h0tia$iri$1@owl.slip.net> <01bd6cd8$abc2fba0$64646464@gregm> <6hip1t$9ri$1@nnrp1.dejanews.com> <6hir9n$ero$1@lyra.csx.cam.ac.uk> NNTP-Posting-Host: pvtest.cc.iastate.edu Nick Maclaren wrote: } wrote: }>Clearly neither the Windows 95 core nor Windows NT kernel is real time }>currently. I understand Microsoft is working with some other }>companies to add real time extensions to the NT kernel. }40 years of experience has shown that adding "real-time extensions" to a }general-purpose operating system merely adds complexity, and doesn't do }anything useful. There are some fairly convincing theoretical arguments }that this is an immutable law of nature. Unfortunately, the combination }of experience and theory is insufficient to prevent idiots from repeating }the mistakes of the past. This raises the counter question -- is it possible to do the opposite -- start with a real time kernel and build a general O.S. on top of it and end up with something useful. John -- John Hascall, Software Engr. Shut up, be happy. The conveniences you ISU Computation Center demanded are now mandatory. -Jello Biafra mailto:john@iastate.edu http://www.cc.iastate.edu/staff/systems/john/welcome.html <-- the usual crud ###### Path: ccw.ch!usenet From: Neil Franklin Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Wed, 22 Apr 1998 00:46:42 +0200 Organization: My own Private Self Lines: 20 Message-ID: <353D21D2.2A894685@ccw.ch> References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <6hgd3o$pvu$1@andros.cygnus.com> <6hh1jf$4ks$1@nnrp1.dejanews.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 4.04 [en] (X11; I; Linux 2.0.27 i486) spaik@simagic.com wrote: > > In article <6hgd3o$pvu$1@andros.cygnus.com>, billm@cygnus.com (Bill Moyer) > wrote: > > * The low-cost processors tend to be for embedded applications. > > If it's an embedded processor, then you would probably want > > to run an embedded system OS on it, not WNT. > > Of course, Microsoft is trying to push Windows NT as an embedded > (and real time) OS. Windows for Powerstations? Is that what they used at Tchernobyl? Windows for Airplanes? Crashing near you soon! SCHUDDER!! -- Neil.Franklin.remove@ccw.ch http://www.ccw.ch/Neil.Franklin/ franklin.remove@arch.ethz.ch http://caad.arch.ethz.ch/~franklin/ This computer has 2 cats, one in /bin, one on top purring. ###### Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch From: stevenss@freenet.msp.mn.us (Scott Stevens) Subject: Re: Growth of space and density, and some wild-ass predictions References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <6hgd3o$pvu$1@andros.cygnus.com> <6hh1jf$4ks$1@nnrp1.dejanews.com> <6hidkb$moa2@eccws1.dearborn.ford.com> X-Newsreader: News Xpress 2.0 Beta #0 Date: Wed, 22 Apr 98 02:05:39 GMT NNTP-Posting-Host: usr-401-5-156.isd.net Message-ID: <353d506e.0@aedes.isd.net> Lines: 36 Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!newsfeed.internetmci.com!209.150.160.22!newsfeed.wli.net!news.he.net!katana!uunet!in4.uu.net!aedes.isd.net!msp00006 In article <6hidkb$moa2@eccws1.dearborn.ford.com>, tph@rmi.net wrote: >spaik@simagic.com wrote: >: In article <6hgd3o$pvu$1@andros.cygnus.com>, billm@cygnus.com (Bill Moyer) >: wrote: >: > * The low-cost processors tend to be for embedded applications. >: > If it's an embedded processor, then you would probably want >: > to run an embedded system OS on it, not WNT. > >: Of course, Microsoft is trying to push Windows NT as an embedded >: (and real time) OS. > >NT? You sure? I've heard about their grand plan to push WinCE as >a real-time OS. But NT? Isn't that more than a bit bloated for an >embedded system? I'd think that even MS would have trouble convincing >people to cram NT into most embedded applications, unless they can >really strip it down a lot. > NT real time is a bit of a stretch. But remember, lots of different people mean different things when they say "embedded." A place that I worked at two years ago was using "embedded" Windows NT as a printer server for high end color printing. What they meant by "embedded" was that the machine with NT on it had a stripped down version of NT. There was no regular GUI in place, just a control panel application that allowed manupulation of graphics, printers, etc. It was an embedded application because the OS was built into the application and not available for general use. An information Kiosk (like an electronic "where you are" map at a shopping maul) is another possible embedded application for Windows NT. And I believe one of the stock-broker type information services (Bridge?) embeds NT into their terminals. It isn't the kind of embedding that I do everyday (4 and 8 bit embedded controllers in single-chip applications) but it's what some people by embedded. >-- >Tom Harrington --------- tph@rmii.com -------- http://rainbow.rmii.com/~tph > "Moses parted the red sea. Oppenheimer split the atom. But "Bob" cut > the crap." -Rev. Steve Antczak, Revelation X: The "Bob" Apocryphon >Cookie's Revenge: ftp://ftp.rmi.net/pub2/tph/cookie/cookies-revenge.sit.hqx ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!news-peer.sprintlink.net!news-backup-west.sprintlink.net!news-in-west.sprintlink.net!news.sprintlink.net!Sprint!199.171.20.9!news.nkn.net!news.cybernews.net!news From: zorn@technologist.com (Bill Bohan) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Wed, 22 Apr 1998 04:07:53 GMT Organization: CyberNews Network Lines: 24 Message-ID: <353c337a.25723365@news.dhc.net> References: <6h0tia$iri$1@owl.slip.net> <3537832D.3BF8B968@hobbes.dtcc.edu> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> NNTP-Posting-Host: 207-55-169-30.dhc.net X-Newsreader: Forte Free Agent 1.11/32.235 On 17 Apr 1998 19:57:54 -0700, billm@cygnus.com (Bill Moyer) wrote: [Snip] > > Whether or not these trends will continue is another matter, >but I won't get into that here. I'd like to throw out a table >for popular consumption which shows you what you could buy for >a given amount of $$$ as the years go by, assuming these trends >*do* continue: > [Snip more] I think you should have done the same comparison for the processors. What speed processor will you be able to get for your $500 ? We all know that the price won't drop below a certain price, usually $100 or so. They simply discontinue the slower chip in favor of the newer, faster chips. No offense intended, just my opinion. I thought it was a good article, otherwise. Bill Bohan -- As many programmers as there are, that's how many ways there are to do it. ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!howland.erols.net!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!baron.netcom.net.uk!netcom.net.uk!server3.netnews.ja.net!news.ox.ac.uk!not-for-mail From: Thomas Womack Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 22 Apr 1998 07:09:40 GMT Organization: Oxford University, England Lines: 62 Distribution: usa Message-ID: <6hk53k$9k4$1@news.ox.ac.uk> References: <6h0tia$iri$1@owl.slip.net> <3537832D.3BF8B968@hobbes.dtcc.edu> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> NNTP-Posting-Host: sable.ox.ac.uk X-Newsreader: TIN [UNIX 1.3 unoff BETA release 970120] In comp.arch Bill Moyer wrote: : I've been collecting information about the costs, sizes, and : capabilities of computer systems since I was a boy (I wanted to : see if reality reflected Moore's Law), and found that the rates : of growth of capability-per-dollar has been more or less stable. : Whether or not these trends will continue is another matter, : but I won't get into that here. I'd like to throw out a table : for popular consumption which shows you what you could buy for : a given amount of $$$ as the years go by, assuming these trends : *do* continue: : : $100 $250 Price of Alpha Price of Intel : DRAM HD 533MHz 21164 300MHz PII : 1998 64 MB 6 GB $495 $400 : 1999 96 MB 9.5 GB $355 $285 What I would find *much* more interesting is the historical data. I've been collecting pricing on x86 processors, memory and disc for a couple of years now; I'd be very interested to see (say) the curve for the cost of a 386SX against time. Memory prices halved between Dec 95 and Apr 96, halved again between Apr 96 and Aug 96, halved *again* between Aug 96 and Feb 97, *again* between Feb 97 and Dec 97, and look perilously close to halving by about Christmas this year (32M of SIMMs suitable for low-end PC, prices in UKP) Dec 95 699 Apr 96 563 (the halving was for 4M SIMMs) Aug 96 195 May 97 95 Dec 97 47 Current 28 (P133, UKP) Sep 95 729 Jan 96 369 Aug 96 169 Jul 97 75 (withdrawn) (P200/MMX, UKP) Feb 97 368 Jul 97 175 Feb 98 99 Now 74 (PPro 200, UKP) Apr 96 899 Oct 96 419 Nov 97 265 (withdrawn) (P2/233, UKP) May 97 485 Sep 97 335 (when I bought mine) Jan 98 199 (when my roommate bought his :() Now 144, falling more slowly Tom ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!woodstock.news.demon.net!demon!news.demon.co.uk!demon!mail2news.demon.co.uk!tnglwood.demon.co.uk!unclebob From: unclebob@tnglwood.demon.co.uk (Robert Billing) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Wed, 22 Apr 98 07:46:59 GMT Message-ID: <893231219snz@tnglwood.demon.co.uk> References: <6h0tia$iri$1@owl.slip.net> <01bd6cd8$abc2fba0$64646464@gregm> <6hip1t$9ri$1@nnrp1.dejanews.com> <6hir9n$ero$1@lyra.csx.cam.ac.uk> <6hj6fv$c2e$1@news.iastate.edu> Reply-To: unclebob@tnglwood.demon.co.uk X-Mail2News-User: unclebob@tnglwood.demon.co.uk X-Mail2News-Path: post-30.mail.demon.net X-Trace: mail2news.demon.co.uk 893242804 24818 unclebob tnglwood.demon.co.uk X-Complaints-To: abuse@demon.net X-Newsreader: Demon Internet Simple News v1.29 Lines: 19 In article <6hj6fv$c2e$1@news.iastate.edu> john@iastate.edu "John Hascall" writes: > This raises the counter question -- is it possible > to do the opposite -- start with a real time kernel > and build a general O.S. on top of it and end up with > something useful. It depends. Not all real time kernels make a distinction between user and root, or between different users, and hence it is difficult to get proper protection working. -- I am Robert Billing, Christian, inventor, traveller, cook and animal lover, I live near 0:46W 51:22N. http://www.tnglwood.demon.co.uk/ "Bother," said Pooh, "Eeyore, ready two photon torpedoes and lock phasers on the Heffalump, Piglet, meet me in transporter room three" ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!newspump.monmouth.com!newspeer.monmouth.com!newsfeed.ecrc.net!newsfeed2.ecrc.net!news.siemens.de!news.HL.Siemens.DE!not-for-mail From: Bernd Paysan Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Wed, 22 Apr 1998 09:40:57 +0200 Organization: Siemens AG, Semiconductor Group Lines: 16 Message-ID: <353D9F09.D77@remove.muenchen.this.org.junk> References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <6hgd3o$pvu$1@andros.cygnus.com> <6hh1jf$4ks$1@nnrp1.dejanews.com> <353D21D2.2A894685@ccw.ch> NNTP-Posting-Host: 190.1.15.101 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 3.01 (WinNT; I) Neil Franklin wrote: > Windows for Powerstations? Is that what they used at Tchernobyl? No, they just pushed the "don't push me"-button. It said "never push this button again", pronounced "booom". > Windows for Airplanes? Crashing near you soon! A company that is not even able to create a demo without being spoiled by the Deamon of Demos should not be considered for anything that is even remote to mission critical. -- Bernd Paysan "Late answers are wrong answers!" http://www.jwdt.com/~paysan/ ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newshub.northeast.verio.net!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!news.idt.net!kiowa.exodus.net!207.82.39.214.MISMATCH!newsread.exodus.net!not-for-mail From: Sam Yorko Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Wed, 22 Apr 1998 10:11:52 -0800 Organization: Exodus Communications Lines: 20 Message-ID: <353E32E8.6344@compuserve.com> References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <6hgd3o$pvu$1@andros.cygnus.com> <6hh1jf$4ks$1@nnrp1.dejanews.com> <6hidkb$moa2@eccws1.dearborn.ford.com> <353d506e.0@aedes.isd.net> NNTP-Posting-Host: exosecure-symbol.psd.symbol.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 3.01Gold (WinNT; I) Scott Stevens wrote: > > In article <6hidkb$moa2@eccws1.dearborn.ford.com>, tph@rmi.net wrote: > >spaik@simagic.com wrote: > NT real time is a bit of a stretch. But remember, lots of different people > mean different things when they say "embedded." A place that I worked at two > years ago was using "embedded" Windows NT as a printer server for high end > color printing. What they meant by "embedded" was that the machine with NT on > it had a stripped down version of NT. There was no regular GUI in place, just > a control panel application that allowed manupulation of graphics, printers, > etc. It was an embedded application because the OS was built into the > application and not available for general use. An information Kiosk (like > an electronic "where you are" map at a shopping maul) is another possible > embedded application for Windows NT. And I believe one of the stock-broker > type information services (Bridge?) embeds NT into their terminals. It isn't > the kind of embedding that I do everyday (4 and 8 bit embedded controllers in > single-chip applications) but it's what some people by embedded. > There are cash registers out there running NT.... ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!newspump.monmouth.com!newspeer.monmouth.com!netnews.com!europa.clark.net!208.134.241.18!newsfeed.internetmci.com!208.217.49.199!-interface!cam-news-hub1.bbnplanet.com!su-news-feed1.bbnplanet.com!news.bbnplanet.com!inet16.us.oracle.com!not-for-mail From: "Bill B." Newsgroups: alt.folklore.computers Subject: Re: Growth of space and density, and some wild-ass predictions Date: Wed, 22 Apr 1998 14:17:08 -0700 Organization: Trying, but not quite Lines: 29 Message-ID: <6hlmmb$6hq$1@inet16.us.oracle.com> References: <6h0tia$iri$1@owl.slip.net> <6hir9n$ero$1@lyra.csx.cam.ac.uk> <6hj6fv$c2e$1@news.iastate.edu> <893231219snz@tnglwood.demon.co.uk> <6hlf92$qd0$1@lyra.csx.cam.ac.uk> NNTP-Posting-Host: upsizeme.us.oracle.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 3.01 (X11; I; AIX 2) Nick Maclaren wrote: >There are certain properties of operating systems > that are almost impossible to get right unless they are designed in, > which does not mean that only one such property can be designed in. > These include: > > "Real time" operation > Continuous operation (including hot changeover) > Trustable security > Debuggability > > Very few systems have ANY of those designed in, and there aren't many > that have even two. Hi Nick, while I agree in principle with your last sentence, I think that there is a point here that needs to be made. That is, as far as the "Trusted security" and "Debuggability" issues are concerned, I think it would be nearly impossible to implement BOTH of these features in ANY operating system. Perhaps you meant it to mean debuggability at a user/programmer level? By the same token, continuous operation AND debuggability would pretty much have to go hand-in-hand. That is, you got one, you've also got the other. Just my tu-pence on the matter. Bill B. ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsfeed.internetmci.com!209.150.160.22!newsfeed.wli.net!su-news-hub1.bbnplanet.com!su-news-feed1.bbnplanet.com!news.bbnplanet.com!inet16.us.oracle.com!not-for-mail From: "Bill B." Newsgroups: alt.folklore.computers Subject: Re: Growth of space and density, and some wild-ass predictions Date: Wed, 22 Apr 1998 14:19:01 -0700 Organization: Trying, but not quite Lines: 29 Message-ID: <6hlmps$bp4$1@inet16.us.oracle.com> References: <6h0tia$iri$1@owl.slip.net> <6hir9n$ero$1@lyra.csx.cam.ac.uk> <6hj6fv$c2e$1@news.iastate.edu> <893231219snz@tnglwood.demon.co.uk> <6hlf92$qd0$1@lyra.csx.cam.ac.uk> NNTP-Posting-Host: upsizeme.us.oracle.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 3.01 (X11; I; AIX 2) Nick Maclaren wrote: >There are certain properties of operating systems > that are almost impossible to get right unless they are designed in, > which does not mean that only one such property can be designed in. > These include: > > "Real time" operation > Continuous operation (including hot changeover) > Trustable security > Debuggability > > Very few systems have ANY of those designed in, and there aren't many > that have even two. Hi Nick, while I agree in principle with your last sentence, I think that there is a point here that needs to be made. That is, as far as the "Trusted security" and "Debuggability" issues are concerned, I think it would be nearly impossible to implement BOTH of these features in ANY operating system. Perhaps you meant it to mean debuggability at a user/programmer level? By the same token, continuous operation AND debuggability would pretty much have to go hand-in-hand. That is, you got one, you've also got the other. Just my tu-pence on the matter. Bill B. ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newsfeed.uk.ibm.net!sackheads.org!ibm.net!news-feed.inet.tele.dk!bofh.vszbr.cz!masternews.telia.net!newspost.telia.com!news.lejonet.se!not-for-mail From: Ingvar Mattsson Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 22 Apr 1998 16:11:54 +0200 Organization: Foreningen Lejonet, Linkoping, Sweden Lines: 34 Message-ID: References: <6h0tia$iri$1@owl.slip.net> <01bd6cd8$abc2fba0$64646464@gregm> <6hip1t$9ri$1@nnrp1.dejanews.com> <6hir9n$ero$1@lyra.csx.cam.ac.uk> <6hj6fv$c2e$1@news.iastate.edu> NNTP-Posting-Host: ns.idasys.se X-Newsreader: Gnus v5.3/Emacs 19.34 john@iastate.edu (John Hascall) writes: > > Nick Maclaren wrote: > } wrote: > }>Clearly neither the Windows 95 core nor Windows NT kernel is real time > }>currently. I understand Microsoft is working with some other > }>companies to add real time extensions to the NT kernel. > > }40 years of experience has shown that adding "real-time extensions" to a > }general-purpose operating system merely adds complexity, and doesn't do > }anything useful. There are some fairly convincing theoretical arguments > }that this is an immutable law of nature. Unfortunately, the combination > }of experience and theory is insufficient to prevent idiots from repeating > }the mistakes of the past. > > This raises the counter question -- is it possible > to do the opposite -- start with a real time kernel > and build a general O.S. on top of it and end up with > something useful. I would think so, yes. The question is how RT the RT kernel is with the extra goo in it. ;) But I see less of a conflict in overlaying a general OS on an RT system than overlaying an RT system on a general (non-RT) OS. Diad Data tried and, sort of, succeeded in writing a real-time Sytem III (I think) unix. Diab-people, how correct are my memories? //Ingvar (another Diab shipped to The Cellar, just yesterday) -- 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 ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news.linkline.be!news-raspail.gip.net!news-peer.gip.net!news.gsl.net!gip.net!news-peer.sprintlink.net!news-backup-east.sprintlink.net!news-in-east.sprintlink.net!news.sprintlink.net!Sprint!206.172.150.11!news1.bellglobal.com!news.uunet.ca!atbowler From: atbowler@thinkage.on.ca (Alan Bowler) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 22 Apr 1998 17:02:15 GMT Organization: Thinkage Ltd. Lines: 12 Message-ID: <6hl7qn$7vi$1@nntp2.uunet.ca> References: <6hip1t$9ri$1@nnrp1.dejanews.com> <6hir9n$ero$1@lyra.csx.cam.ac.uk> <6hj6fv$c2e$1@news.iastate.edu> NNTP-Posting-Host: 192.102.11.4 In article <6hj6fv$c2e$1@news.iastate.edu> john@iastate.edu (John Hascall) writes: > > This raises the counter question -- is it possible > to do the opposite -- start with a real time kernel > and build a general O.S. on top of it and end up with > something useful. yes. Although the real time functionality sometimes gets lost after some some. VMS traces back to a real time system (rsx). Real time systems can make good single person devlopment environments also (Thoth, QNX, Wicat's MCS) ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!Cabal.CESspool!bofh.vszbr.cz!pegasus.csx.cam.ac.uk!lyra.csx.cam.ac.uk!nmm1 From: nmm1@cus.cam.ac.uk (Nick Maclaren) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 22 Apr 1998 19:09:22 GMT Organization: University of Cambridge, England Lines: 33 Message-ID: <6hlf92$qd0$1@lyra.csx.cam.ac.uk> References: <6h0tia$iri$1@owl.slip.net> <6hir9n$ero$1@lyra.csx.cam.ac.uk> <6hj6fv$c2e$1@news.iastate.edu> <893231219snz@tnglwood.demon.co.uk> NNTP-Posting-Host: ursa.cus.cam.ac.uk In article <893231219snz@tnglwood.demon.co.uk>, Robert Billing wrote: >In article <6hj6fv$c2e$1@news.iastate.edu> > john@iastate.edu "John Hascall" writes: > >> This raises the counter question -- is it possible >> to do the opposite -- start with a real time kernel >> and build a general O.S. on top of it and end up with >> something useful. > > It depends. Not all real time kernels make a distinction between user >and root, or between different users, and hence it is difficult to get >proper protection working. That is the point. There are certain properties of operating systems that are almost impossible to get right unless they are designed in, which does not mean that only one such property can be designed in. These include: "Real time" operation Continuous operation (including hot changeover) Trustable security Debuggability Very few systems have ANY of those designed in, and there aren't many that have even two. Nick Maclaren, University of Cambridge Computing Service, New Museums Site, Pembroke Street, Cambridge CB2 3QG, England. Email: nmm1@cam.ac.uk Tel.: +44 1223 334761 Fax: +44 1223 334679 ###### Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!feed1.news.luth.se!luth.se!cam-news-hub1.bbnplanet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!howland.erols.net!torn!news.trentu.ca!csess From: csess@blaze.trentu.ca (Eric S. Smith: Left-Field Marshal) Subject: Re: Growth of space and density, and some wild-ass predictions Sender: news@news.trentu.ca (Ken Brown) Message-ID: Date: Wed, 22 Apr 1998 19:51:49 GMT X-Nntp-Posting-Host: blaze.trentu.ca References: <6h0tia$iri$1@owl.slip.net> <6hir9n$ero$1@lyra.csx.cam.ac.uk> <6hj6fv$c2e$1@news.iastate.edu> <353D1FB9.EB21AD79@cs.wisc.edu> Organization: Trent University, Peterborough Canada Lines: 34 In article <353D1FB9.EB21AD79@cs.wisc.edu>, Andy Glew wrote: >(1) QNX. Started as a RT kernel with limited functionality. > Now has pretty much anything from UNIX I want to use. And has been a general-purpose OS for some time -- certainly more than ten years, though I doubt that it was Unixy enough to be truly cool at that point. I wasn't much interested in system calls when I read the manuals, so I don't recall any of them. The command-line shell I used was Unixoid -- the directory separator was / -- but the commands had different names. Hmm: Unix QNX cd .. cd ^ (pointing upward, right? Heh.) chmod chattr ("change (file) attributes") rm frel ("file release") rmdir drel ("directory release") This was running on Cemcorp (later Unisys) ICONs, diskless workstations (the file-server was called the LEXICON, haw, haw...). Cemcorp stood, I think, for Canadian Educational Microprocessor Corporation. They were neat machines, but slower than slow. Absolutely *zero* coverage of them on the Weeb. QNX is currently quite cool. If you have access to a reasonably modern PC, go visit http://www.qnx.com/iat/index.html and download their demo disk. *Very* cool. --Eric Smith ("Ha-low!") ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!news.sgi.com!cygnus.com!cygnus.com!not-for-mail From: billm@cygnus.com (Bill Moyer) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 22 Apr 1998 19:58:11 -0700 Organization: Subtle, but there. Lines: 62 Message-ID: <6hmao3$f9v$1@andros.cygnus.com> References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <353c337a.25723365@news.dhc.net> NNTP-Posting-Host: andros.cygnus.com In article <353c337a.25723365@news.dhc.net> zorn@technologist.com (Bill Bohan) writes: >On 17 Apr 1998 19:57:54 -0700, billm@cygnus.com (Bill Moyer) wrote: > >[Snip] >> >> Whether or not these trends will continue is another matter, >>but I won't get into that here. I'd like to throw out a table >>for popular consumption which shows you what you could buy for >>a given amount of $$$ as the years go by, assuming these trends >>*do* continue: >> >[Snip more] > >I think you should have done the same comparison for the >processors. What speed processor will you be able to get >for your $500 ? We all know that the price won't drop below >a certain price, usually $100 or so. They simply discontinue >the slower chip in favor of the newer, faster chips. I think about this often, actually, but it's hard to make predictions about future desktop/server processors based on Moore's Law. Things like executable units, clock frequency, instruction issue rate (or VLIW size), etc don't scale very predictably with increased transistor density (and even the folks who design new ICs on the assumption of higher future transistor densities (Intel, Digital) don't try to do so looking more than a few years into the future, and don't really know in advance how fast the design will go). I don't think I could make any predictions about high-end CPUs that would be worth the bandwidth. When I do think seriously about predicting future micro- processors, I think of modern processors shrunk down and used as embedded processors. The Z80 once powered personal computers, and now it's shrunk to near-microscopic size, clocked at a stunning 25MHz (the most powerful Z80 I ever had at home was 4MHz), and put in everything from gameboys to credit-card sized ID tags. The MPC850DC and PIC show a growing trend to integrate more "peripheral" hardware (RAM, ROM, ethernet, etc) onto the processor's die, which can at once simplify the design of embedded systems and reduce the processor's pin count (ie, if all of the RAM is on-chip, you don't need to blow ~100 pins on an external bus inter- face). So I might be able to predict the existence of a shrunk- down version of the 533MHz 21164PC Alpha with 128MB of RAM, 32MB of ROM, and an ethernet interface as its only means of communication, for about $12 in the year 2010. It might be the processor animating the images on the surfaces of your 6-year-old granddaughter's singing "Hello, Kitty" lunchpail or some other silly thing. >No offense intended, just my opinion. >I thought it was a good article, otherwise. Absolutely no offense taken! There's so much noise on usenet these days that it's flattering to be noticed at all. :-) Thanks. -- Bill Moyer ###### ws-peer.gip.net!news.gsl.net!gip.net!newshub1.home.com!news.home.com!enews.sgi.com!news.sgi.com!cygnus.com!cygnus.com!not-for-mail From: billm@cygnus.com (Bill Moyer) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 22 Apr 1998 20:06:00 -0700 Organization: Subtle, but there. Lines: 31 Message-ID: <6hmb6o$fe7$1@andros.cygnus.com> References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <6hk53k$9k4$1@news.ox.ac.uk> NNTP-Posting-Host: andros.cygnus.com In article <6hk53k$9k4$1@news.ox.ac.uk> Thomas Womack writes: >In comp.arch Bill Moyer wrote: > >: I've been collecting information about the costs, sizes, and >: capabilities of computer systems since I was a boy (I wanted to >: see if reality reflected Moore's Law), and found that the rates >: of growth of capability-per-dollar has been more or less stable. > >: Whether or not these trends will continue is another matter, >: but I won't get into that here. I'd like to throw out a table >: for popular consumption which shows you what you could buy for >: a given amount of $$$ as the years go by, assuming these trends >: *do* continue: >: >: $100 $250 Price of Alpha Price of Intel >: DRAM HD 533MHz 21164 300MHz PII > >: 1998 64 MB 6 GB $495 $400 >: 1999 96 MB 9.5 GB $355 $285 > >What I would find *much* more interesting is the historical data. I've >been collecting pricing on x86 processors, memory and disc for a >couple of years now; I'd be very interested to see (say) the curve for >the cost of a 386SX against time. Okay .. When I have the time I'll type in the data (it's currently in handwritten note form) and post the URL here, in case anyone else is interested. I'll also send you email when it's up. -- Bill Moyer ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!newsfeed.uk.ibm.net!sackheads.org!ibm.net!news-feed.inet.tele.dk!bofh.vszbr.cz!newsfeed.direct.ca!newshub1.home.com!news.home.com!enews.sgi.com!fido.asd.sgi.com!slurn From: slurn@griffin.engr.sgi.com (Scott Lurndal) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 22 Apr 1998 21:03:26 GMT Organization: Silicon Graphics, Inc. Lines: 26 Message-ID: <6hlluu$4eeja@fido.asd.sgi.com> References: <6h0tia$iri$1@owl.slip.net> <01bd6cd8$abc2fba0$64646464@gregm> <6hip1t$9ri$1@nnrp1.dejanews.com> <6hir9n$ero$1@lyra.csx.cam.ac.uk> <6hj6fv$c2e$1@news.iastate.edu> Reply-To: slurn@engr.sgi.com NNTP-Posting-Host: griffin.engr.sgi.com X-Newsreader: xrn 9.00 In article <6hj6fv$c2e$1@news.iastate.edu>, john@iastate.edu (John Hascall) writes: |> Nick Maclaren wrote: |> } wrote: |> }>Clearly neither the Windows 95 core nor Windows NT kernel is real time |> }>currently. I understand Microsoft is working with some other |> }>companies to add real time extensions to the NT kernel. |> |> }40 years of experience has shown that adding "real-time extensions" to a |> }general-purpose operating system merely adds complexity, and doesn't do |> }anything useful. There are some fairly convincing theoretical arguments |> }that this is an immutable law of nature. Unfortunately, the combination |> }of experience and theory is insufficient to prevent idiots from repeating |> }the mistakes of the past. |> |> This raises the counter question -- is it possible |> to do the opposite -- start with a real time kernel |> and build a general O.S. on top of it and end up with |> something useful. |> The unisys OPUS product was built on top of the chorus microkernel, which is a real-time kernel (used primarily in telephone switches). scott (John - say hi to Tony). ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!newspump.monmouth.com!newspeer.monmouth.com!fu-berlin.de!unlisys!news.snafu.de!jnickelsen From: jnickelsen@acm.org (Juergen Nickelsen) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Thu, 23 Apr 1998 00:53:36 +0200 Organization: Unlimited Surprise Systems, Berlin Lines: 10 Message-ID: <1d7vsiv.1mauglx3us9r6N@n33-116.berlin.snafu.de> References: <6h0tia$iri$1@owl.slip.net> <3535475B.C03684@monmouth.com> <353637D5.1BCB@cisco.com> <3536B328.ABFBEEF0@monmouth.com> <3537832D.3BF8B968@hobbes.dtcc.edu> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <01bd6bf2$0c016e80$64646464@gregm> <01bd6cd8$756b6470$64646464@gregm> <6hii2q$auv$1@news.iastate.edu> NNTP-Posting-Host: n33-116.berlin.snafu.de X-Newsreader: MacSOUP 2.3 Rick Hawkins wrote: > what about caldera dos? that should do it, shouldn't it? or opendos, > or whatever it's called this week. It's called DR-DOS again (last time I looked), but is still available from Caldera. -- Juergen Nickelsen ###### Message-ID: <353EB06B.C9DA236E@Japan.NCR.COM> Date: Thu, 23 Apr 1998 12:07:23 +0900 From: Eric Hildum Organization: NCR Japan X-Mailer: Mozilla 4.04 [en] (Win95; I) MIME-Version: 1.0 Subject: Re: Growth of space and density, and some wild-ass predictions References: <6h0tia$iri$1@owl.slip.net> <01bd6cd8$abc2fba0$64646464@gregm> <6hip1t$9ri$1@nnrp1.dejanews.com> <6hir9n$ero$1@lyra.csx.cam.ac.uk> <6hj6fv$c2e$1@news.iastate.edu> <893231219snz@tnglwood.demon.co.uk> Content-Type: text/plain; charset=iso-2022-jp Content-Transfer-Encoding: 7bit Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Lines: 24 Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!newsfeed.internetmci.com!204.71.0.51!spamkiller.internetmci.com!pull-feed.internetmci.com!rpc1283.ncr.com!rpc1286!rpc1284.daytonoh.ncr.com Robert Billing wrote: > In article <6hj6fv$c2e$1@news.iastate.edu> > john@iastate.edu "John Hascall" writes: > > > This raises the counter question -- is it possible > > to do the opposite -- start with a real time kernel > > and build a general O.S. on top of it and end up with > > something useful. > > It depends. Not all real time kernels make a distinction between user > and root, or between different users, and hence it is difficult to get > proper protection working. > > -- > I am Robert Billing, Christian, inventor, traveller, cook and animal > lover, I live near 0:46W 51:22N. http://www.tnglwood.demon.co.uk/ > "Bother," said Pooh, "Eeyore, ready two photon torpedoes and lock > phasers on the Heffalump, Piglet, meet me in transporter room three" Neither does a well know operating system... -- --------------------------- Eric Hildum Eric.Hildum@Japan.NCR.COM ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsxfer3.itd.umich.edu!jobone!dailyplanet.srl.ford.com!eccws1.dearborn.ford.com!longhorn!tph From: tph@longhorn.uucp (Tom Harrington) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Followup-To: alt.folklore.computers,comp.sys.unisys,comp.arch Date: 23 Apr 1998 16:38:34 GMT Organization: Mechanist Industries Lines: 55 Message-ID: <6hnqqa$nqd3@eccws1.dearborn.ford.com> References: <6h0tia$iri$1@owl.slip.net> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <6hgd3o$pvu$1@andros.cygnus.com> <6hh1jf$4ks$1@nnrp1.dejanews.com> <6hidkb$moa2@eccws1.dearborn.ford.com> <353d506e.0@aedes.isd.net> Reply-To: tph@rmi.net NNTP-Posting-Host: cs0053.eld.ford.com X-Newsreader: TIN [version 1.2 PL2] Scott Stevens (stevenss@freenet.msp.mn.us) wrote: : In article <6hidkb$moa2@eccws1.dearborn.ford.com>, tph@rmi.net wrote: : >spaik@simagic.com wrote: : > : >: Of course, Microsoft is trying to push Windows NT as an embedded : >: (and real time) OS. : > : >NT? You sure? I've heard about their grand plan to push WinCE as : >a real-time OS. But NT? Isn't that more than a bit bloated for an : >embedded system? I'd think that even MS would have trouble convincing : >people to cram NT into most embedded applications, unless they can : >really strip it down a lot. : > : NT real time is a bit of a stretch. But remember, lots of different people : mean different things when they say "embedded." OK, I guess it was my fault for saying "embedded" while thinking "real-time", and not considering that a system can be one of these without necessarily being the other. : A place that I worked at two : years ago was using "embedded" Windows NT as a printer server for high end : color printing. What they meant by "embedded" was that the machine with NT on : it had a stripped down version of NT. There was no regular GUI in place, just : a control panel application that allowed manupulation of graphics, printers, : etc. It was an embedded application because the OS was built into the : application and not available for general use. An information Kiosk (like : an electronic "where you are" map at a shopping maul) is another possible : embedded application for Windows NT. The real question, then, is whether there's any place where you DO need a real-time OS where NT could be shoehorned into place. Something like an information kiosk doesn't require real-time performance. Applications which do require this performance would, I think, have a hard time swallowing NT. That's why I was wondering if MS really expected that they could sell NT for real-time applications. WinCE wouldn't have the same memory requirements, and so could conceivably be sold for real-time embedded applications (though there are obviously other problems with this scenario). : And I believe one of the stock-broker : type information services (Bridge?) embeds NT into their terminals. It isn't : the kind of embedding that I do everyday (4 and 8 bit embedded controllers in : single-chip applications) but it's what some people by embedded. I'm not familiar with the requirements of such a terminal; would it require real-time performance? I guess the answer must be "no", since I don't think MS has reached the point where they claim that NT can be used for real-time applications. -- Tom Harrington --------- tph@rmii.com -------- http://rainbow.rmii.com/~tph "...one can get high scores on intelligence tests, but still be unable to think for one's self." -Revelation X: The "Bob" Apocryphon Cookie's Revenge: ftp://ftp.rmi.net/pub2/tph/cookie/cookies-revenge.sit.hqx ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!newsfeed.internetmci.com!165.87.194.248!newsm.ibm.net!ibm.net!uunet!in5.uu.net!shore!news From: dmoisan@shore.net (David Moisan) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Fri, 24 Apr 1998 01:28:55 GMT Organization: DM Productions Lines: 32 Message-ID: <353faf0f.28551667@news.shore.net> References: <6h0tia$iri$1@owl.slip.net> <6hh1jf$4ks$1@nnrp1.dejanews.com> <01bd6cd8$abc2fba0$64646464@gregm> <6hip1t$9ri$1@nnrp1.dejanews.com> <6hir9n$ero$1@lyra.csx.cam.ac.uk> Reply-To: dmoisan@shore.net NNTP-Posting-Host: tcgbos-01-83.port.shore.net X-Newsreader: Forte Free Agent 1.11/32.235 On 21 Apr 1998 19:16:07 GMT, nmm1@cus.cam.ac.uk (Nick Maclaren) wrote: > >40 years of experience has shown that adding "real-time extensions" to a >general-purpose operating system merely adds complexity, and doesn't do >anything useful. There are some fairly convincing theoretical arguments >that this is an immutable law of nature. Unfortunately, the combination >of experience and theory is insufficient to prevent idiots from repeating >the mistakes of the past. > You'll love this: The new subway cars for the T's Red Line in Boston use DOS! Yes. Someone was riding the train and noted that one of the displays in the operator's cab had a A> prompt. Sometimes programs crash with a "General Failure reading Drive...Abort, Retry, Fail." This was a thread on ne.transportation a few months ago. "Crashing the train" has an entirely new meaning here. >It really gives me a warm, cosy feeling of safety to think that my life >and finances are going to be under the control of a component with the >proven reliability record of the Windows CE :-( If you read the latest (first?) issue of Windows CE Developers Journal, there's an article therein exploring the virtues, if any, of using CE for "near-real-time" development. Dave David Moisan, N1KGH n1kgh@amsat.org http://www1.shore.net/~dmoisan Invisible Disability: http://www1.shore.net/~dmoisan/invisible_disability.html GE Superradio FAQ: http://www1.shore.net/~dmoisan/faqs/superradio/gesr_faq.html ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!newsfeed.internetmci.com!194.72.7.126!btnet-peer!btnet!dispose.news.demon.net!demon!news.demon.co.uk!demon!mail2news.demon.co.uk!tnglwood.demon.co.uk!unclebob From: unclebob@tnglwood.demon.co.uk (Robert Billing) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Fri, 24 Apr 98 08:15:24 GMT Message-ID: <893405724snz@tnglwood.demon.co.uk> References: <6hnqqa$nqd3@eccws1.dearborn.ford.com> Reply-To: unclebob@tnglwood.demon.co.uk X-Mail2News-User: unclebob@tnglwood.demon.co.uk X-Mail2News-Path: post-30.mail.demon.net X-Trace: mail2news.demon.co.uk 893418043 21031 unclebob tnglwood.demon.co.uk X-Complaints-To: abuse@demon.net X-Newsreader: Demon Internet Simple News v1.29 Lines: 20 In article <6hnqqa$nqd3@eccws1.dearborn.ford.com> tph@rmi.net "Tom Harrington" writes: > Applications which do require this performance would, I think, > have a hard time swallowing NT. That's why I was wondering if MS The sort of thing I'm doing, which involves handling TV field rate, and in extreme cases line rate, interrupts, and doing substantial calculations on each, while at the same time running a load of backgrounds, would be way beyond NT. I wouldn't consider anything where I couldn't either a) discuss timing with the authors or b) see the source. -- I am Robert Billing, Christian, inventor, traveller, cook and animal lover, I live near 0:46W 51:22N. http://www.tnglwood.demon.co.uk/ "Bother," said Pooh, "Eeyore, ready two photon torpedoes and lock phasers on the Heffalump, Piglet, meet me in transporter room three" ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news-fra1.dfn.de!news-ber1.dfn.de!fu-berlin.de!unlisys!news.snafu.de!jnickelsen From: jnickelsen@acm.org (Juergen Nickelsen) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: Fri, 24 Apr 1998 08:31:46 +0200 Organization: Unlimited Surprise Systems, Berlin Lines: 9 Message-ID: <1d7ynmg.6g2g0y112eemoN@n241-111.berlin.snafu.de> References: <6h0tia$iri$1@owl.slip.net> <3535475B.C03684@monmouth.com> <353637D5.1BCB@cisco.com> <3536B328.ABFBEEF0@monmouth.com> <3537832D.3BF8B968@hobbes.dtcc.edu> <3537F0FC.3489@cisco.com> <6h94ri$av3$1@andros.cygnus.com> <6hgd3o$pvu$1@andros.cygnus.com> <6hh1jf$4ks$1@nnrp1.dejanews.com> <6hidkb$moa2@eccws1.dearborn.ford.com> <353d506e.0@aedes.isd.net> <353E32E8.6344@compuserve.com> NNTP-Posting-Host: n241-111.berlin.snafu.de X-Newsreader: MacSOUP 2.3 Sam Yorko wrote: [embedded OSs] > There are cash registers out there running NT.... ... and some that are running SCO Unix, with X11 on top of it. -- Juergen Nickelsen ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!nntprelay.mathworks.com!newsfeed.internetmci.com!141.211.144.13!newsxfer3.itd.umich.edu!newsxfer.itd.umich.edu!newsrelay.iastate.edu!news.iastate.edu!john From: john@iastate.edu (John Hascall) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 27 Apr 1998 18:02:05 GMT Organization: Iowa State University, Ames, Iowa USA Lines: 37 Message-ID: <6i2h6t$141$1@news.iastate.edu> References: <6h0tia$iri$1@owl.slip.net> <6h94ri$av3$1@andros.cygnus.com> <353ADDD7.63B2@azstarnet.com> NNTP-Posting-Host: pvtest.cc.iastate.edu Maynard Handley wrote: }bill_h wrote: }> Another curious thing: It generally seems to take 5-7 YEARS to actually }> reach a point where we're getting something approaching optimal use of }> the features of a given processor generation. ... [18month cycle] }This is not just a feature of hardware. }I remain convinced that a major reason for the success of MS is that they }have persistence. Or to put it another way-a major reason for the failure }of Apple is that they lack persistence. Apple has a long history of }introducing software technologies, being amazed that they aren't part of }every major app after 3 months, and cancelling them after 6 months. MS }have had a few cancellations of system software, but for the most part }they simply keep plugging away at it. One might argue that this patience is a luxury of being the far and away market share leader. People with MS Blurfl 6.0 are probably not going to abandon it in droves if it doesn't have some feature of in version 7.0 because to change would be a big deal, they can afford to wait to see if has legs and add it in V9.0. }One could argue the same obsession with 3 months from now is what has }doomed UNIX/X---no-one in that part of the business has been prepared to }make changes that will have no immediate impact---but they will make life }a whole lot easier in 5 yrs. See also the "what have you done for us THIS quarter" Wall Street Attitude.... John -- John Hascall, Software Engr. Shut up, be happy. The conveniences you ISU Computation Center demanded are now mandatory. -Jello Biafra mailto:john@iastate.edu http://www.cc.iastate.edu/staff/systems/john/welcome.html <-- the usual crud ###### From: "isd.net" Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch References: <6h0tia$iri$1@owl.slip.net> <6hgd3o$pvu$1@andros.cygnus.com> <6hghj5$6gl$1@joe.rice.edu> <6i83h2$kst@web.nmti.com> Subject: Re: Growth of space and density, and some wild-ass predictions Date: Wed, 29 Apr 1998 19:23:02 -0500 Lines: 24 X-Newsreader: Microsoft Outlook Express 4.71.1712.3 X-MimeOLE: Produced By Microsoft MimeOLE V4.71.1712.3 NNTP-Posting-Host: usr-401-2-147.isd.net Message-ID: <3547c47c.0@aedes.isd.net> Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-dc.gip.net!news-peer.gip.net!news.gsl.net!gip.net!ais.net!uunet!in1.uu.net!aedes.isd.net!usr-401-2-147.isd.net You don't run applications on the NT server. The bosses kid runs games on the NT server on the weekend when you're not around. Games he downloads from insecure FTP sites. That's the beauty of a server OS that runs the same binaries as 90% of the clients out there... Peter da Silva wrote in message <6i83h2$kst@web.nmti.com>... >In article <6hghj5$6gl$1@joe.rice.edu>, >Anil Thomas Maliyekkel wrote: >> Unfortunately for some people, the definition of a useful >> OS is that it runs all the latest and greatest Microsoft >> applications. > >Why would you want to run applications on NT server? To see how slow they >can go? > >-- >%!PS >/c{{64 sub}forall curveto}def 312 192 translate 24 24 scale 1 -2 moveto(B;>;?>) >c fill 1 0 moveto(JIJNJP)c(JRGU@U)c(9U6R6P)c(6N6I?@)c gsave stroke grestore .9 >setgray fill showpage % I'm getting serious propane withdrawal. ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsxfer3.itd.umich.edu!newsxfer.itd.umich.edu!ultraneo.neosoft.com!news-proxy.baileynm.com!web.nmti.com!peter From: peter@nmti.com (Peter da Silva) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 29 Apr 1998 20:45:22 GMT Organization: Network/development platform support, NMTI Lines: 14 Distribution: usa Message-ID: <6i83h2$kst@web.nmti.com> References: <6h0tia$iri$1@owl.slip.net> <6hgd3o$pvu$1@andros.cygnus.com> <6hghj5$6gl$1@joe.rice.edu> NNTP-Posting-Host: grendel.nmti.com X-No-Archive: yes In article <6hghj5$6gl$1@joe.rice.edu>, Anil Thomas Maliyekkel wrote: > Unfortunately for some people, the definition of a useful > OS is that it runs all the latest and greatest Microsoft > applications. Why would you want to run applications on NT server? To see how slow they can go? -- %!PS /c{{64 sub}forall curveto}def 312 192 translate 24 24 scale 1 -2 moveto(B;>;?>) c fill 1 0 moveto(JIJNJP)c(JRGU@U)c(9U6R6P)c(6N6I?@)c gsave stroke grestore .9 setgray fill showpage % I'm getting serious propane withdrawal. ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsxfer3.itd.umich.edu!newsxfer.itd.umich.edu!ultraneo.neosoft.com!news-proxy.baileynm.com!web.nmti.com!peter From: peter@nmti.com (Peter da Silva) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 29 Apr 1998 20:48:39 GMT Organization: Network/development platform support, NMTI Lines: 20 Message-ID: <6i83n7$lqi@web.nmti.com> References: <6hip1t$9ri$1@nnrp1.dejanews.com> <6hir9n$ero$1@lyra.csx.cam.ac.uk> <6hj6fv$c2e$1@news.iastate.edu> <6hl7qn$7vi$1@nntp2.uunet.ca> NNTP-Posting-Host: grendel.nmti.com X-No-Archive: yes In article <6hl7qn$7vi$1@nntp2.uunet.ca>, Alan Bowler wrote: > In article <6hj6fv$c2e$1@news.iastate.edu> john@iastate.edu (John Hascall) writes: > > This raises the counter question -- is it possible > > to do the opposite -- start with a real time kernel > > and build a general O.S. on top of it and end up with > > something useful. > yes. Although the real time functionality sometimes gets lost > after some some. VMS traces back to a real time system (rsx). Sort of. Some of us RSX users would rather disavow it. -- %!PS /c{{64 sub}forall curveto}def 312 192 translate 24 24 scale 1 -2 moveto(B;>;?>) c fill 1 0 moveto(JIJNJP)c(JRGU@U)c(9U6R6P)c(6N6I?@)c gsave stroke grestore .9 setgray fill showpage % I'm getting serious propane withdrawal. ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!newsxfer3.itd.umich.edu!newsxfer.itd.umich.edu!ultraneo.neosoft.com!news-proxy.baileynm.com!web.nmti.com!peter From: peter@nmti.com (Peter da Silva) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 29 Apr 1998 20:59:42 GMT Organization: Network/development platform support, NMTI Lines: 18 Message-ID: <6i84bu$ntm@web.nmti.com> References: <6h0tia$iri$1@owl.slip.net> <6h94ri$av3$1@andros.cygnus.com> <353c337a.25723365@news.dhc.net> <6hmao3$f9v$1@andros.cygnus.com> NNTP-Posting-Host: grendel.nmti.com In article <6hmao3$f9v$1@andros.cygnus.com>, Bill Moyer wrote: > to credit-card sized ID tags. The MPC850DC and PIC show a > growing trend to integrate more "peripheral" hardware (RAM, > ROM, ethernet, etc) onto the processor's die, which can at > once simplify the design of embedded systems and reduce the > processor's pin count (ie, if all of the RAM is on-chip, > you don't need to blow ~100 pins on an external bus inter- > face). Look at something like the Java ring, with the processor in a hearing-aid- battery case and a bit-serial interface for power and I/O. -- %!PS /c{{64 sub}forall curveto}def 312 192 translate 24 24 scale 1 -2 moveto(B;>;?>) c fill 1 0 moveto(JIJNJP)c(JRGU@U)c(9U6R6P)c(6N6I?@)c gsave stroke grestore .9 setgray fill showpage % I'm getting serious propane withdrawal. ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news.belnet.be!news-penn.gip.net!news-peer.gip.net!news.gsl.net!gip.net!dca1-hub1.news.digex.net!digex!ultraneo.neosoft.com!news-proxy.baileynm.com!web.nmti.com!peter From: peter@nmti.com (Peter da Silva) Newsgroups: alt.folklore.computers,comp.sys.unisys,comp.arch Subject: Re: Growth of space and density, and some wild-ass predictions Date: 30 Apr 1998 03:23:25 GMT Organization: Network/development platform support, NMTI Lines: 16 Message-ID: <6i8qrd$cs6@web.nmti.com> References: <6h0tia$iri$1@owl.slip.net> <6hghj5$6gl$1@joe.rice.edu> <6i83h2$kst@web.nmti.com> <3547c47c.0@aedes.isd.net> NNTP-Posting-Host: grendel.nmti.com In article <3547c47c.0@aedes.isd.net>, isd.net wrote: > You don't run applications on the NT server. The bosses kid runs games on > the NT server on the weekend when you're not around. Games he downloads > from insecure FTP sites. That's the beauty of a server OS that runs the > same binaries as 90% of the clients out there... You get to go HOME on weekends? (followup-to: the network that shall not be named) -- %!PS /c{{64 sub}forall curveto}def 312 192 translate 24 24 scale 1 -2 moveto(B;>;?>) c fill 1 0 moveto(JIJNJP)c(JRGU@U)c(9U6R6P)c(6N6I?@)c gsave stroke grestore .9 setgray fill showpage % I'm getting serious propane withdrawal. ###### Path: ccw.ch!elna.ethz.ch!news-zh.switch.ch!news-ge.switch.ch!news.maxwell.syr.edu!feed2.news.erols.com!erols!WCG!arl-news-svc-3.compuserve.com!news.informix.com!not-for-mail From: Jason Dawes Newsgroups: alt.folklore.computers Subject: Re: Growth of space and density, and some wild-ass predictions Date: Thu, 30 Apr 1998 10:13:15 -0700 Organization: Informix Software, Inc. Lines: 17 Message-ID: <3548B12B.8DFC17C2@informix.doh.com> References: <6h0tia$iri$1@owl.slip.net> <6hgd3o$pvu$1@andros.cygnus.com> <6hghj5$6gl$1@joe.rice.edu> <6i83h2$kst@web.nmti.com> <3547c47c.0@aedes.isd.net> NNTP-Posting-Host: tigereye.illustra.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 4.04 [en] (X11; I; SunOS 5.5.1 sun4u) isd.net wrote: > You don't run applications on the NT server. The bosses kid runs games on > the NT server on the weekend when you're not around. Games he downloads > from insecure FTP sites. That's the beauty of a server OS that runs the > same binaries as 90% of the clients out there... > They would be those really fun games that don't try and use graphics or the file system, right? -- ----------------------------------------------------------------------- Jason Dawes jasond@doh.informix.com