dmocsny@minerva.che.uc.edu (Daniel Mocsny) (11/21/90)
In article <ROLAND.90Nov19160821@lolly.Stanford.EDU> roland@CS.Stanford.EDU (Roland Conybeare) writes: >I believe Sun hopes to enlarge the market for SPARC machines by inviting >competition (this seems like an extraordinary action to me!). Sun's move is extraordinary, in that it shows insight and courage almost unheard of in the computer industry. The alternative, however, can be observed failing just about everywhere people are still trying it. Attempting to sustain high profit margins on obsolete architectures via proprietary lock-ins just doesn't work in the long term. Eventually your customers are losing so much money that they are willing to undergo the agony of porting to somebody else's box, no matter how hard you try to make it for them. >Sun wins if it keeps market share & SPARCs become commonplace. On the other >hand Sun may lose to superior manufacturing. I suspect that Sun realizes the only way to insure that it maintains its superior manufacturing is to expose itself deliberately to competition. Top management at a company can't simply order the whole company to become more competitive. Everybody in the enterprise has their own personal agenda, and if line managers don't feel a pressure to run a tight ship, they will always find something else to do. By insuring that competition will exist, Sun sends a convincing message to every employee: produce, or the party is over. While putting the company in a sink-or-swim position might seem insane, it is probably the only strategy that has a chance in the medium-term. Because every computer company is in a sink-or-swim position if you look down the road a few years. Victories are won and lost by the cumulative result of seemingly insignificant differences in growth rates. For example, during the 19th century, the USA had an industrial growth rate 1% higher than that of Great Britain. As a result, after 100 years the USA went from a banana country to replacing Great Britain as the world's leading industrial power. Things happen faster in the computer industry. No company can afford to become complacent simply because it can sell something today that nobody else has. Within a couple of years, everybody else will be selling something better. Look at the proprietary minicomputer lines showing flat sales today. You are seeing what happens when everybody at the company believes proprietary lock-ins will protect them from having to compete. Lock-ins can work in the short run. But they only buy you a few years in which to dig your nice, comfortable grave. Sun also understands that customers don't buy computers to decorate their desks, but to run software. Therefore, the more software a computer will run, the more valuable it is to the customer, period. Since the vast majority of computer buyers do not write their own code, the price and availability of software for a particular computer is at least as important as the computer's underlying technical merit. And the price and availability of software depends almost entirely on one thing, directly or indirectly: the size of the computer's installed base. By inviting in the cloners, Sun is insuring that the total number of SPARC systems installed will be higher than if Sun marketed them exclusively. The cloners won't merely cannibalize Sun sales; they will also penetrate new markets that Sun would not have reached on its own. As a direct consequence, more software developers will target this market. That will increase the value of the architecture to the consumer, further stimulating sales. The overall size of the market will then be much higher. Sun will have to compete harder for market share, but at least it has insured that 10 years from now it will still have a market to compete in. Consider where Apple computer would be right now if it had made cloning the Mac easy. Consider where DEC would be if it had made cloning the VAX easy. Maybe in chapter 11 :-), or maybe fighting to hang onto a share of a still-thriving market. Instead, the handwriting is on the wall for the proprietary lines. Eventually they will simply get swamped in a flood of cheap PC clones, SPARC clones, etc., which will be doing everything they can do, for less. Proprietary lock-ins don't work in the long run. The whole organization adapts to easy sales and high margins, and starts getting fat and slow. Not to mention spending more money on its legal activities than on creating value through superior engineering. But even the best legal talent in existence can't persuade customers to spend more than they have to. -- Dan Mocsny Snail: Internet: dmocsny@minerva.che.uc.edu Dept. of Chemical Engng. M.L. 171 dmocsny@uceng.uc.edu University of Cincinnati 513/751-6824 (home) 513/556-2007 (lab) Cincinnati, Ohio 45221-0171
henry@zoo.toronto.edu (Henry Spencer) (11/22/90)
In article <6749@uceng.UC.EDU> dmocsny@minerva.che.uc.edu (Daniel Mocsny) writes: >I suspect that Sun realizes the only way to insure that it maintains >its superior manufacturing is to expose itself deliberately to >competition... To be plausible, such theories should explain why Sun's earlier products remain cloaked in secrecy. Consideration of that suggests that this is a marketing ploy rather than a grand scheme to flog Sun's own people into being competitive. (Either that, or Sun's management is simply irrational, a definite possibility.) The most obvious sinister possibility is that this is an attempt to do for hardware what many people thought the AT&T/Sun alliance was meant to do for software: to give Sun a guaranteed lead on the competition by giving it control of an "industry standard", so that Sun can change the rules of the game regularly, forcing everyone else into a constant game of catch-up. As witness the uproar that followed the AT&T/Sun business, the trick is to do this without getting the victims upset in advance. To accomplish that, you have to give them the illusion that they have a say and some modicum of control. One way to do this is to let them spend their time arguing over "standardizing" the previous generation, making it clear that the "standard" must not invalidate any of your designs, while you quietly get the next generation ready. My prediction: if/when SPARC (or S-bus) becomes a non-Sun-controlled standard, Sun will promptly announce SPARC II (S-bus II), with loud claims that it renders all the old stuff obsolete and is the obvious new standard. Who, me, cynical? :-) -- "I don't *want* to be normal!" | Henry Spencer at U of Toronto Zoology "Not to worry." | henry@zoo.toronto.edu utzoo!henry
dmocsny@minerva.che.uc.edu (Daniel Mocsny) (11/22/90)
In article <1990Nov21.174938.7861@zoo.toronto.edu> henry@zoo.toronto.edu (Henry Spencer) writes: >In article <6749@uceng.UC.EDU> dmocsny@minerva.che.uc.edu (Daniel Mocsny) writes: >>I suspect that Sun realizes the only way to insure that it maintains >>its superior manufacturing is to expose itself deliberately to >>competition... >To be plausible, such theories should explain why Sun's earlier products >remain cloaked in secrecy. While I hardly claim inside knowledge, I have read commentary suggesting that Sun did not see its earlier systems competing against PC's, but rather as addressing a separate market niche. Two factors have changed this: (1) Sun's rapid growth threatens to saturate the technical workstation market, prompting Sun to target the commercial sector; and (2) the emergence of cheap 386 PC's threatened to start chipping away at Sun from the low end. Nobody can succeed in the commercial market by talking about MIPS, SPECmarks, etc. Success in the commercial market requires being able to run lots of cheap, powerful, slick software right out of a box, and without a Ph.D. standing by to provide everything the vendor forgot. I.e., just like the suits know they can do with Mac's and PC's. The only way to do this, in turn, is to pull in the big and small software houses with the magic phrase: i n s t a l l e d b a s e. I don't think Sun worried too much about going head to head with the PC's when they were selling Sun-1's and Sun-2's. The overlap wasn't there like we see today. Today, an 80386 can do everything a SPARC can do, only slower and cheaper. Sun got to its first billion selling a pretty specialized product. To get to its first 10 billion, Sun is going to have to sell something that suits can understand and operate. Apple has carved a substantial back alley of the PC market for itself. But imagine where the Mac would be today if Apple hadn't closed up the box and nailed the corners down. The Mac might have trampled and squashed the PC clones if Apple hadn't kept the price artificially high and scared off the suits by being the sole supplier. Apple would be competing for a share of a much larger market. I'm sure they'd still be in it. After all, IBM still sells a lot of PC's. > Consideration of that suggests that this is >a marketing ploy rather than a grand scheme to flog Sun's own people into >being competitive. (Either that, or Sun's management is simply irrational, >a definite possibility.) Well, I did once read a story of one of the 19th century robber barons, perhaps it was Andrew Carnegie (but don't quote me). As a young man, he was hired to manage a foundry. He came in the first day, met the troops, and asked the foreman how many batches his men had created that shift. The foreman replied "seven" or some such. So Andrew took a piece of chalk and wrote a large "7" on the floor. He stuck around for the next shift, watched them turn out nine batches, and he wrote a "9" on the floor. No cajoling or threats, just the chalked number. Pretty soon the shifts were engaged in a contest with each other, and plant production rose substantially. (Remember, this was in the days before quality control was discovered.) The computer industry is extraordinarily competitive, and this competition proceeds on different time scales. I'm sure that Sun's licensing efforts are worth more to its customers than to its employees. Nonetheless, Sun's management can't be blind. They must have some idea what to expect when they invite in the cloners. It's just like the battles between the ancient Persians and the Spartans. The Persians invaded with a 10:1 numerical advantage, or some such, but the Spartans crushed them. The Persians were just used to guarding a sleepy empire, while the Spartans didn't do much else besides fight with their neighbors all the time. (My apologies to historians if I'm garbling this too badly, or even fabricating a story outright to support my general argument.) What's going to happen when a few lean, mean companies emerge from the smoke and ruins of the Clone Wars, and then take on the somnolent Persian empire? >My prediction: if/when SPARC (or S-bus) becomes a non-Sun-controlled >standard, Sun will promptly announce SPARC II (S-bus II), with loud claims >that it renders all the old stuff obsolete and is the obvious new standard. Interestingly, I read in _UNIX Today_, albeit yesterday :-) that Sun's new 40 MHz SPARCstation II is generating heavy criticism from competitors about having to go to a high clock rate to match the performance of the other workstations with substantially slower clocks. The HP spokesman was saying Sun is now hurt by having a standardized SPARC that is preventing Sun from innovating. I will be surprised if Sun tries to junk the existing standard, since it is trying very hard to build up an applications library for the SPARC. Since software lags hardware by 3--5 years, it seems impossible to deliver, simultaneously, leading-edge hardware and a broad applications base. All the more reason to push for greater software portability. -- Dan Mocsny Snail: Internet: dmocsny@minerva.che.uc.edu Dept. of Chemical Engng. M.L. 171 dmocsny@uceng.uc.edu University of Cincinnati 513/751-6824 (home) 513/556-2007 (lab) Cincinnati, Ohio 45221-0171
henry@zoo.toronto.edu (Henry Spencer) (11/24/90)
In article <6769@uceng.UC.EDU> dmocsny@minerva.che.uc.edu (Daniel Mocsny) writes: >>>I suspect that Sun realizes the only way to insure that it maintains >>>its superior manufacturing is to expose itself deliberately to >>>competition... > >>To be plausible, such theories should explain why Sun's earlier products >>remain cloaked in secrecy. > >While I hardly claim inside knowledge, I have read commentary suggesting >that Sun did not see its earlier systems competing against PC's, but >rather as addressing a separate market niche. Two factors have changed >this... So why haven't they changed Sun's policies regarding its older systems? As I've asked (rhetorically) before: what possible end can be served by the continuing secrecy regarding the insides of the Sun 2? One would expect secrecy about new systems but a more relaxed attitude about old ones. From Sun, if anything we are getting the reverse! This is why I suggest devious scheming or sheer irrationality as an explanation. >>My prediction: if/when SPARC (or S-bus) becomes a non-Sun-controlled >>standard, Sun will promptly announce SPARC II (S-bus II), with loud claims >>that it renders all the old stuff obsolete and is the obvious new standard. > >...I will be surprised if Sun tries to junk the existing standard, since >it is trying very hard to build up an applications library for the >SPARC... They won't junk it; they will pull an Intel, announcing a new standard that is a superset of the old one, rendering all the competitors' machines out of date without requiring a total software rewrite. -- "I'm not sure it's possible | Henry Spencer at U of Toronto Zoology to explain how X works." | henry@zoo.toronto.edu utzoo!henry
agn@unh.cs.cmu.edu (Andreas Nowatzyk) (11/24/90)
In article <1990Nov23.181851.26588@zoo.toronto.edu> (Henry Spencer) writes: > So why haven't they changed Sun's policies regarding its older systems? > As I've asked (rhetorically) before: what possible end can be served by > the continuing secrecy regarding the insides of the Sun 2? One would > expect secrecy about new systems but a more relaxed attitude about old > ones. From Sun, if anything we are getting the reverse! This is why > I suggest devious scheming or sheer irrationality as an explanation. This is plain BS. The Arcons project here at CMU had SUN's schematics for the SUN-2 boards. This was necessary to modify the boards for real-time OS experiments. You just have to talk to the right people. This is much less an issue of secrecy rather an issue of documentation cost. It takes time and money to document things to the point that some outsider can make use of it. Since there aren't too many customers that are interested in this, SUN has better things to do than to produce document kits for obsolete hardware (just like most other companies). On Sbus and Sparc: >They won't junk it; they will pull an Intel, announcing a new standard that >is a superset of the old one, rendering all the competitors' machines out >of date without requiring a total software rewrite. Sparc International's reference machine is not build by SUN... -- -- Andreas Nowatzyk (DC5ZV) Carnegie-Mellon University agn@unh.cs.cmu.edu Computer Science Department (412) 268-3617
rick@ameristar (Rick Spanbauer) (11/25/90)
In article <1990Nov23.181851.26588@zoo.toronto.edu> henry@zoo.toronto.edu (Henry Spencer) writes: >In article <6769@uceng.UC.EDU> dmocsny@minerva.che.uc.edu (Daniel Mocsny) writes: [edited] >>>To be plausible, such theories should explain why Sun's earlier products >>>remain cloaked in secrecy. >> >So why haven't they changed Sun's policies regarding its older systems? >As I've asked (rhetorically) before: what possible end can be served by >the continuing secrecy regarding the insides of the Sun 2? One would >expect secrecy about new systems but a more relaxed attitude about old >ones. From Sun, if anything we are getting the reverse! This is why >I suggest devious scheming or sheer irrationality as an explanation. Well, Henry, it is pretty clear that when it comes to Sun you prefer elaborate conspiracy theories ;-) I find Sun's reluctance to release eg schematics of motherboards, power supplies, etc an inconvenience, too. Under the "dubious scheming" heading, we might find that they don't make such information available to limit choices in arranging for third party repair of modules. >"I'm not sure it's possible | Henry Spencer at U of Toronto Zoology >to explain how X works." | henry@zoo.toronto.edu utzoo!henry Rick Spanbauer Ameristar
rick@ameristar (Rick Spanbauer) (11/25/90)
In article <11182@pt.cs.cmu.edu> agn@unh.cs.cmu.edu (Andreas Nowatzyk) writes: > >In article <1990Nov23.181851.26588@zoo.toronto.edu> (Henry Spencer) writes: >> So why haven't they changed Sun's policies regarding its older systems? >> As I've asked (rhetorically) before: what possible end can be served by >> the continuing secrecy regarding the insides of the Sun 2? One would >> expect secrecy about new systems but a more relaxed attitude about old >> ones. From Sun, if anything we are getting the reverse! This is why >> I suggest devious scheming or sheer irrationality as an explanation. > >This is plain BS. The Arcons project here at CMU had SUN's schematics for >the SUN-2 boards. This was necessary to modify the boards for real-time OS >experiments. You just have to talk to the right people. This is much less an Actually, it is not BS, Andreas. Sun has never been forthcoming with schematics or *any* low level information about their systems. At another place that I work for, one with both university credentials and more than 100 Suns, we attempted to get such documentation. We were told alternatively that either the information was "not available" or provided only on a "need to know" basis. Doing repairs in house did not consitute need to know. Note that we could not get schematics to 3/75 power supplies either, even though they are manufactured to Sun spec by a third party. Sun uses ECAD like the rest of use, and surely has all their ECO trails and everything else on line - I submit that it is more than a matter of not having the information sitting pretty somewhere that people do not have access to the stuff. As a counter example, these days I do most of my hardware work on Commodore Amigas - CBM provides a 380 page reference manual (published by Addison-Wesley, no less) on programming their full custom ASIC parts in the Amiga and they ship full schematics with *every* box they sell. Remember that CBM is a smaller company than Sun and that eg Amiga 3000 is a higher parts count machine than the SS-1. Sun is wrong in this matter. They ought to provide full low level docs + schematics on one of those nice quarterly CD-ROMs they publish - if done right, they might get Valid to ship their schematics editor (stripped, of course) as a viewer. I don't have high hopes of this ever happening, though. > -- Andreas Nowatzyk (DC5ZV) Rick Spanbauer Ameristar
henry@zoo.toronto.edu (Henry Spencer) (11/25/90)
In article <11182@pt.cs.cmu.edu> agn@unh.cs.cmu.edu (Andreas Nowatzyk) writes: >This is plain BS. The Arcons project here at CMU had SUN's schematics for >the SUN-2 boards... I'd be extremely surprised if this wasn't under a nondisclosure agreement, one that took some effort to get. >... You just have to talk to the right people... And have a convincing project, and be lucky. >This is much less an >issue of secrecy rather an issue of documentation cost. It takes time and >money to document things to the point that some outsider can make use of it. Except that in many cases, this stuff is already documented for internal use. (Often the documentation is a bit raw, but that's a different issue.) I can't speak for the Sun 2, actually, but near-complete programming docs for the early Sun 3s *do* exist... and very few people can get them, even today. -- "I'm not sure it's possible | Henry Spencer at U of Toronto Zoology to explain how X works." | henry@zoo.toronto.edu utzoo!henry
myhui@bnr.ca (Michael Hui) (11/25/90)
In article <1990Nov24.162427.10023@ameristar> rick@ameristar (Rick Spanbauer) writes: > somewhere that people do not have access to the stuff. As a counter > example, these days I do most of my hardware work on Commodore Amigas > - CBM provides a 380 page reference manual (published by Addison-Wesley, > no less) on programming their full custom ASIC parts in the Amiga and > they ship full schematics with *every* box they sell. In the other hardware sphere of oscilloscopes, Tektronix also ships full schematic, _trouble shooting_, and _calibration_ information with each and every scope. Legend has it that a smaller scope maker's VP of engineering was too proud to copy Tek's horizontal sweep generator circuit, and went on to design his own for his company, and got a patent for it in the process. In the old days, scopes were made of discrete transistors, hand soldered to boards, wired by hand. Copying would have been trivial. Michael Hui myhui@bnr.ca
rds@SYRAH.NECTAR.CS.CMU.EDU (Robert Sansom) (11/26/90)
In article <11182@pt.cs.cmu.edu>, agn@unh.cs.cmu.edu (Andreas Nowatzyk) writes: |> Organization: Carnegie-Mellon University, CS/RI |> |> In article <1990Nov23.181851.26588@zoo.toronto.edu> (Henry Spencer) writes: |> > So why haven't they changed Sun's policies regarding its older systems? |> > As I've asked (rhetorically) before: what possible end can be served by |> > the continuing secrecy regarding the insides of the Sun 2? One would |> > expect secrecy about new systems but a more relaxed attitude about old |> > ones. From Sun, if anything we are getting the reverse! This is why |> > I suggest devious scheming or sheer irrationality as an explanation. |> |> This is plain BS. The Arcons project here at CMU had SUN's schematics for |> the SUN-2 boards. This was necessary to modify the boards for real-time OS |> experiments. You just have to talk to the right people. This is much less an |> issue of secrecy rather an issue of documentation cost. It takes time and |> money to document things to the point that some outsider can make use of it. |> Since there aren't too many customers that are interested in this, SUN has |> better things to do than to produce document kits for obsolete hardware |> (just like most other companies). |> |> On Sbus and Sparc: |> >They won't junk it; they will pull an Intel, announcing a new standard that |> >is a superset of the old one, rendering all the competitors' machines out |> >of date without requiring a total software rewrite. |> |> Sparc International's reference machine is not build by SUN... |> |> -- |> -- Andreas Nowatzyk (DC5ZV) |> |> Carnegie-Mellon University agn@unh.cs.cmu.edu |> Computer Science Department (412) 268-3617 Actually Andreas works for Sun Microsystems these days, not Carnegie Mellon. Robert Sansom, School of Computer Science Carnegie Mellon University, Pittsburgh, PA 15213 INTERNET: sansom@cs.cmu.edu CSNET: sansom%cs.cmu.edu@relay.cs.net BITNET: sansom%cs.cmu.edu@cmuccvma UUCP: ...!seismo!cs.cmu.edu!sansom
uad1077@dircon.uucp (11/26/90)
Dan Mocsny writes: > Consider where Apple computer would be right now if it had made > cloning the Mac easy. Consider where DEC would be if it had made While I largely agree with your analysis of Sun's move, Apple are perhaps the crwoning example that shows there is another way. Apple often seem to be downplayed by netters, perhaps because they sell so many machines to non-computer users... It's hard to imagine that the market for Macs could be many times bigger if they were clonable. In a sense the Mac follows the Japanese consumer electronics thing of making sure that everyone who sees one WANTS one. -- Ian D. Kemmish Tel. +44 767 601 361 18 Durham Close uad1077@dircon.UUCP Biggleswade ukc!dircon!uad1077 Beds SG18 8HZ United Kingd uad1077%dircon@ukc.ac.uk
sgolson@pyrite.East.Sun.COM (Steve Golson) (11/27/90)
In article <6769@uceng.UC.EDU> dmocsny@minerva.che.uc.edu (Daniel Mocsny) writes: >Interestingly, I read in _UNIX Today_, albeit yesterday :-) that >Sun's new 40 MHz SPARCstation II is generating heavy criticism from >competitors about having to go to a high clock rate to match the >performance of the other workstations with substantially slower >clocks. The HP spokesman was saying Sun is now hurt by having a >standardized SPARC that is preventing Sun from innovating. How is this different from HP's "standardized" Precision Architecture? Does this likewise prevent HP from innovating? Steve Golson -- Trilobyte Systems -- Carlisle MA -- sgolson@east.sun.com (consultant for, but not employed by, Sun Microsystems) "As the people here grow colder, I turn to my computer..." -- Kate Bush
dtynan@unix386.Convergent.COM (Dermot Tynan) (12/01/90)
In article <1990Nov25.194404.3376@dircon.uucp>, uad1077@dircon.uucp writes: > > It's hard to imagine > that the market for Macs could be many times bigger if they were > clonable. With all due respect, I think this has more to do with your imagination, than the size of the market. The Marketplace has accepted the Mac very readily, due to its software advances. Let's face it, you can only go so far with the MS-DOS prompt, C>, which is of course, based on the CP/M prompt, A>, which is in turn based on the RT-11 prompt, \n. . I'm not saying I prefer the Mac interface, just that the majority of the computer-buying public do. That majority would be a lot larger, if the machine sold, say, for $600.00! It would be a prerequisite in every new home in America (3 bedrooms, washer, dryer, Mac II). AT&T were very embarrassed, when the Federal Government demanded they lower their charges, and they started making *more* money. A lesson lost on the "Pepsi" generation at Apple. > In a sense the Mac follows the Japanese consumer electronics > thing of making sure that everyone who sees one WANTS one. How about an analogy (I know, I know, everyone hates analogies). Ferrari have done a good job of "making sure that everyone who sees one WANTS one". Honda, however, sell more cars and make more money. - Der -- Dermot Tynan dtynan@zorba.Tynan.COM {altos,apple,mips,pyramid}!zorba!dtynan "Five to one, baby, one in five. No-one here gets out alive."
rballard@isis.cs.du.edu (Rexford E Ballard) (12/01/90)
Sun, or any other vendor for that matter, has a simple competitive strategy. Be "open" enough tbe the beneficiary of "commodity" support software and peripherals, "unique" enough to warrant a higher mark-up, and "closed" enough toensure growth and migration to new standards as they are developed. There have been many "cold steel" manufacturers of "standard" "cold iron" boxes such as 68020 running BSD-4.2, CP/IP, and SCSI. Sun survived by adding valued features such as windows (SunView) back when X-10 was still an experimental project and Macs came with 128K. The migration to Xview was somewhat traumatic at first, but as applications are ported to the commodity environments (Common subset of Xview, Motif, ICCM), applications previously only available on PCs are showing up on UNIX. There have been many attempts to turn an MS-DOS PC into a more "unix-like" box, with simple multitasking, windows, device independent graphics, and interprocess communication. Unfortunately, each attempt (including early versions of windows) failed, primarily because of applications such as Lotus 1-2-3 that literally took over the machine. Each upgrade to IBM graphics mandates an upgrade to Lotus. When the IBM-PC first came out, IBM published quite a bit of technical documentation. In addition, they had a "hacker's contest", offering incentives to those who could write softwarehat could take advantage of the unique features of the PC. The intent was to motivate businesses to to switch away from CP/M and MP/M to the PC. This "openness" was a double edged sword. While it did render CP/M obsolete, and it did prevent excessive cloning, it also made it difficult to enhance the basic PC. To be fair to Sun, they have adopted some very open standards including SCSI, TCP/, X-11, and UNIX. They also also provided SysV and BSD 4.2 support in the same machine. They have also been good about defining and opening their standards. NFS, Yellow Pages (NIS), and RPC are examples of standards defined, liscenced cheaply (note -ot free), and well documented through well known sources (Internet RFCs). It would be nice if Sun would ease up on the "Smoke and Mirrors" that make Xview applications seem incompatible with other environments. It would be nice if they could "play nice" with other window managers (XView initialization is sooooo slooooow). On the other hand, look at the trouble they have had to go through to maintain backward compatibility with their previous window systems. XView still proves support for SunView and NeWS. This is real important when you have to wait for 3rd party vendors to publish and release their XView compatible software. Unfortunately, there was so much hype over the Motif vs Xview that vendors had to "wait and see". It was interesting that as the smoke cleared, there was an equivalent to "STDIO" that worked on both platforms, along with many others. I remember when MIT students passed around copies of the UNIX version 6 source code for PDP11/23s. Back then the entire kernal was 128K and were swapped by the context switcher. You could hang 30 users on a machine with less power than a 8Mhz PC/XT. The original AT&T UNIX was almost public domain because antitrust laws prevented them from selling the OS. BSD improved it (to 4 Meg), commercialized it, and liscenced it back to AT&T for SysV.3+. The irony here is that a 3rd party vendor will write software to the "lowest common denominator" (BSD, Curses, Plot, XWidgets, EPS...) to be sure that he can "catch the next wave".'
philip@pescadero.Stanford.EDU (Philip Machanick) (12/02/90)
In article <1635@unix386.Convergent.COM>, dtynan@unix386.Convergent.COM (Dermot Tynan) writes: |> In article <1990Nov25.194404.3376@dircon.uucp>, uad1077@dircon.uucp writes: |> > It's hard to imagine |> > that the market for Macs could be many times bigger if they were |> > clonable. |> [...] I'm not saying |> I prefer the Mac interface, just that the majority of the computer-buying |> public do. That majority would be a lot larger, if the machine sold, say, |> for $600.00! It would be a prerequisite in every new home in America (3 |> bedrooms, washer, dryer, Mac II). AT&T were very embarrassed, when the |> Federal Government demanded they lower their charges, and they started |> making *more* money. A lesson lost on the "Pepsi" generation at Apple. Maybe they've learnt at last. The new Classic costs $750 at the Stanford Bookstore, and the waiting list for them is 2 months long. -- Philip Machanick philip@pescadero.stanford.edu
dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) (12/03/90)
In <1990Dec2.014554.3491@Neon.Stanford.EDU> philip@pescadero.Stanford.EDU (Philip Machanick) writes: >Maybe they've learnt at last. The new Classic costs $750 at the Stanford >Bookstore, and the waiting list for them is 2 months long. Nope. "Classic" here means "obsolete". In this day and age of 28-MIPS machines, what Apple does with an 8 MHz 68000 is utterly irrelevant. -- Rahul Dhesi <dhesi%cirrusl@oliveb.ATC.olivetti.com> UUCP: oliveb!cirrusl!dhesi
sysmgr@KING.ENG.UMD.EDU (Doug Mohney) (12/03/90)
In article <2760@cirrusl.UUCP>, dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) writes: >Nope. "Classic" here means "obsolete". In this day and age of 28-MIPS >machines, what Apple does with an 8 MHz 68000 is utterly irrelevant. That's humorous, for two reasons: A) Apple's "cash crop" is the Mac Classic line (replacing the Apple //e, another irrelevant, obsolete cash cow which managed to give Apple lots of money to develop the Mac line) B) The people who are going to purchase the Mac are highly unlikely to be able to afford 28-MIPS machines. Personally, I'd love to get a QUARTER of the profits from irrelevant 8 MHz 68000 machines which are sold by Apple, Atari, and Commodore. %%%%% Signature v1.1 %%%%% Doug Mohney, Operations Manager, CAD Lab/ME, Univ. of Maryland College Park * Why do VMS system managers get more sleep and less ulcers than their * * UNIX(TM) counterparts, despite the sophistication of UNIX? *
grover@brahmand.Eng.Sun.COM (Vinod Grover) (12/04/90)
In article <2760@cirrusl.UUCP> dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) writes: >In <1990Dec2.014554.3491@Neon.Stanford.EDU> >philip@pescadero.Stanford.EDU (Philip Machanick) writes: > >>Maybe they've learnt at last. The new Classic costs $750 at the Stanford >>Bookstore, and the waiting list for them is 2 months long. > >Nope. "Classic" here means "obsolete". In this day and age of 28-MIPS >machines, what Apple does with an 8 MHz 68000 is utterly irrelevant. According to a recent WSJ article, the demand for the "utterly irrelevant" Mac Classic has been so high that Apple is having trouble meeting it. Vinod Grover
lovejoy@alc.com (Alan Lovejoy) (12/04/90)
In article <2760@cirrusl.UUCP> dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) writes: >In <1990Dec2.014554.3491@Neon.Stanford.EDU> >philip@pescadero.Stanford.EDU (Philip Machanick) writes: >>Maybe they've learnt at last. The new Classic costs $750 at the Stanford >>Bookstore, and the waiting list for them is 2 months long. >Nope. "Classic" here means "obsolete". In this day and age of 28-MIPS >machines, what Apple does with an 8 MHz 68000 is utterly irrelevant. Oh! So THAT's why the 28-MIPS machines are breaking sales records and have months-long waiting lists, while the 8-MHz 68k and x86 machines are sitting in wharehouses with few buyers in sight? I knew there had to be a reason! :-) "Better" is the enemy of "good enough." Once a device does its job acceptably, price becomes the only issue. For most users, an 8 MHz cpu is sufficient. That is, it is PERCEIVED TO BE sufficient. When the public becomes aware that there is this [currently hypothetical] whizbang software program that can double their income every year--but which requires a 28 MIPS machine for acceptable usability, THEN 8 MHz cpus will become irrelevant. We aren't there yet, guys. -- %%%% Alan Lovejoy %%%% | "Do not go gentle into that good night, % Ascent Logic Corp. % | Old age should burn and rave at the close of the day; UUCP: lovejoy@alc.com | Rage, rage at the dying of the light!" -- Dylan Thomas __Disclaimer: I do not speak for Ascent Logic Corp.; they do not speak for me!
dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) (12/04/90)
In <2760@cirrusl.UUCP> I wrote: >>The new Classic costs $750 at the Stanford >>Bookstore, and the waiting list for them is 2 months long. >Nope. "Classic" here means "obsolete". In this day and age of 28-MIPS >machines, what Apple does with an 8 MHz 68000 is utterly irrelevant. A number of people have protested in email and in follow-up postings. Some of them implicitly assumed that any discussion of MIPS must be about a Sun running UNIX. This is not so. The state-of-the-art today in single-user machines is anywhere from 10 to 30 MIPS depending upon your budget. People buying 80x86-based systems are mostly buying machines running 80286 and 80386 CPUs at 16 MHz or higher. The Apple Classic *does not* cost $750 in any usable configuration. (And educational discounts only benefit a small minority of users.) Add a decent amount of mass storage and other peripherals and you can easily pay $2,000 or more. There's just no comparison with equivalent 80386SX-based systems. -- Rahul Dhesi <dhesi%cirrusl@oliveb.ATC.olivetti.com> UUCP: oliveb!cirrusl!dhesi
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (12/04/90)
In article <2760@cirrusl.UUCP> dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) writes: | Nope. "Classic" here means "obsolete". In this day and age of 28-MIPS | machines, what Apple does with an 8 MHz 68000 is utterly irrelevant. Except in technical, social, or ecconomic terms. All the dcomputing the average home user ever needs to do could be done in CP/M, with a Z80 and 64k. People did (and do) run 100 employee businesses on them. All the extra power, memory, and disk space goes to provide the user interface. I'm only slightly kidding. -- bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen) VMS is a text-only adventure game. If you win you can use unix.
philip@pescadero.Stanford.EDU (Philip Machanick) (12/04/90)
In article <2764@cirrusl.UUCP>, dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) writes: |> The state-of-the-art today in single-user machines is anywhere from 10 |> to 30 MIPS depending upon your budget. People buying 80x86-based |> systems are mostly buying machines running 80286 and 80386 CPUs at 16 |> MHz or higher. The Apple Classic *does not* cost $750 in any usable |> configuration. (And educational discounts only benefit a small |> minority of users.) Add a decent amount of mass storage and other |> peripherals and you can easily pay $2,000 or more. There's just no |> comparison with equivalent 80386SX-based systems. The street price of the basic Classic is not much more than $750. The _list_ price of the next model up, with 2M of RAM and a 40M hard disk is $1500. The street price is probably much lower. This is certainly a usable system, if only just. I'm not claiming this is a fantastic deal. All I said was people are queueing up to buy them at this price. This doesn't mean they are "state of the art", "cheaper than an 80386SUX (*) system" or whatever else you want to read into this. Thank you for your indulgence; can we get back to talking about architecture now? -- Philip Machanick philip@pescadero.stanford.edu (*) Meaningless name to ensure I don't launch another "my computer is better than yours" war.
mash@mips.COM (John Mashey) (12/04/90)
In article <1990Dec1.030450.28687@isis.cs.du.edu> rballard@isis.UUCP (Rexford E Ballard) writes: >Sun, or any other vendor for that matter, has a simple competitive >strategy. Be "open" enough tbe the beneficiary of "commodity" support >software and peripherals, "unique" enough to warrant a higher mark-up, and >"closed" enough toensure growth and migration to new standards as they are >developed. Good summary.... Needless to say, this is an interesting discussion. :-) Here are a few more attribiutes that people might discuss in this one, particularly to compare the IBM PC evolution, and SPARCclone stuff: 1) Consider following model of industry structure, and try to figure out which companies control which pieces betweeen the two model:, and to what degree they control them: a) Clear monopoly b) Power of market %, and hence to determine evolution and do designs that othersmers must follow, or to do software also that others must track c) Buying power Note that a) is usually obvious, whereas b) and c) are more subtle. Microprocessor Architecture+Logic Design 1 Circuit design 2 Fab & sell 3 Support chips Design 4 Fab and sell 5 H/W System Design 6 Fab and sell 7 Software O/S (not just who licenses, but who controls by volume) 8 Key compilers 9 Interface Standards 10 Application software 11 Distribution End-user vs dealer vs OEM vs VAR, support, compatibility 12 Interesting thing: for IBM PC, IBM "controlled" hardly any of these. Also, note that if a single buyer buys a huge percentage of a seller's product, they have more influence over what's happening, and of course they get better prices. Sun has been qute open about wanting multiple sources of supply (as does any systems company). (I have my own charts of these things, but I might be biased; it would be interesting to hear other people's views in this model.) 2) Another way to slice it is to draw the chart that shows the largest vendor's % of a "compatible" market, by year. In this case, one would compare IBM: 1981, 1982, .... with Sun: 1987, 1988..... here are the IBM numbers, from IDC (or Dataquest; I'm in Japan right now and don't have it handy): year IBM % year Sun share of SPARC-based systems 1981 100% 1987 100% 1982 98% 1988 100% 1983 85% 1989 97.5% 1984 79% 1985 63% 1986 39% and these days, I think it's around 20% for IBM. I'd suggest, that rather than philosophical arguments, that the percentage of market is at least one realistic metric regarding the "open-ness" of what's really going on. The other useful one is to to see if there is any difference between marketing and sales, i.e., if you call up your Sun salesperson and tell them you're about to buy a lot of clones, do they say: a) Well, OK, I guess it's an open market b) Well, we can beat them on price or performance, or other "clonable" attributes. c) Well, you should stick with Sun because: support compatibility migration strategy earlier compilers or OS releases software that DOESN'T run on the clones (a lotoesn't; for example, about 350 of the items listed in the Summer 1990 SPARCware list require specific hardware or at least I/O busses and drivers, which are NOT necessarily guaranteed to work, even on a clone with same I/O bus. (oops, time to run to press interview). -- -john mashey DISCLAIMER: <generic disclaimer, I speak for me only, etc> UUCP: mash@mips.com OR {ames,decwrl,prls,pyramid}!mips!mash DDD: 408-524-7015, 524-8253 or (main number) 408-720-1700 USPS: MIPS Computer Systems, 930 E. Arques, Sunnyvale, CA 94086
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (12/04/90)
In article <303@mesa.alc.com> lovejoy@alc.com (Alan Lovejoy) writes: | "Better" is the enemy of "good enough." Once a device does its job acceptably, | price becomes the only issue. For most users, an 8 MHz cpu is sufficient. | That is, it is PERCEIVED TO BE sufficient. Truth. The reason that many people are buying SX's instead of 486's, or Mac Classic instead of Mac IIFX, is that they get the job done just as fast. Consider that most of the work done on small computers is just not CPU bound. Home and business users typically run spreadsheets, with leters, design fancy party invitations, and generally don't use much CPU. And the big problems which are CPU intensive may well require something really large, and not be practical on 28 MIPS, either. -- bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen) VMS is a text-only adventure game. If you win you can use unix.
kitchel@iuvax.cs.indiana.edu (Sid Kitchel) (12/04/90)
dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) writes: |In <2760@cirrusl.UUCP> I wrote: |||The new Classic costs $750 at the Stanford |||Bookstore, and the waiting list for them is 2 months long. ||Nope. "Classic" here means "obsolete". In this day and age of 28-MIPS ||machines, what Apple does with an 8 MHz 68000 is utterly irrelevant. |A number of people have protested in email and in follow-up postings. |Some of them implicitly assumed that any discussion of MIPS must be |about a Sun running UNIX. This is not so. |The state-of-the-art today in single-user machines is anywhere from 10 |to 30 MIPS depending upon your budget. People buying 80x86-based |systems are mostly buying machines running 80286 and 80386 CPUs at 16 |MHz or higher. The Apple Classic *does not* cost $750 in any usable |configuration. (And educational discounts only benefit a small |minority of users.) Add a decent amount of mass storage and other |peripherals and you can easily pay $2,000 or more. There's just no |comparison with equivalent 80386SX-based systems. I'm typing this reply on my 80386-SX Austin at work. (I've been told that "SX" stands for "sucks", but I cannot confirm this.) For a turkey, this is a nice, cute little machine. But I must agree with you completely in your: "There's just no comparison..." You are absolutely correct. For productivity, getting it just the way I want it, for crisp graphics, desktop publishing, fonts I can read all day, etc. etc. I much prefer my Mac at home. It has better displays than any Super VGA I've ever seen. The software available for the Mac blows away anything I have at work. I sure am glad that I don't have to hold my breath waiting for the OS/2, Windows 3.XX vs. DOS 6 war to end with a clear winner!! And as a computer architect, I cringe every time I think of the design of the 80x86 machines. Give me a break and give me a 68xxx machine with a reasonable user interface NOW!! --Sid -- Sid Kitchel...............WARNING: allergic to smileys and hearts.... Computer Science Dept. kitchel@cs.indiana.edu Indiana University kitchel@iubacs.BITNET Bloomington, Indiana 47405-4101........................(812)855-9226
sysmgr@KING.ENG.UMD.EDU (Doug Mohney) (12/04/90)
In article <2764@cirrusl.UUCP>, dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) writes: >A number of people have protested in email and in follow-up postings. >Some of them implicitly assumed that any discussion of MIPS must be >about a Sun running UNIX. This is not so. > >The state-of-the-art today in single-user machines is anywhere from 10 >to 30 MIPS depending upon your budget. People buying 80x86-based >systems are mostly buying machines running 80286 and 80386 CPUs at 16 >MHz or higher. Euh, dunno how to tell you this, but a 16Mhz '386 BARELY clocks 1 MIPS. I think the 20 MHz maybe clocks 4. Maybe. So the castrated-16, er '386SX barely runs at 1-2 MIPS (VERY generous assumption). That's for $2K. If you are talking '486 speeds, on a sunny day, with all things in harmony in the Universe, and Intel not tinkering with the benchmarks (;-), you probably can pull 12-16 MIPS. Average mileage may vary depending on amount of cache available and clock speed. Of course, you are talking $3500-4000, with actual vendor configurations averaging $5K. >The Apple Classic *does not* cost $750 in any usable >configuration. (And educational discounts only benefit a small >minority of users.) >Add a decent amount of mass storage and other >peripherals and you can easily pay $2,000 or more. There's just no >comparison with equivalent 80386SX-based systems. So add a hard disk and an extra MB of RAM, and you now have a usable configure for $1250, or thereabouts. That's street price, not educational. And you haven't spent $750. ************************************** This all begs the question of where we can all purchase 28-30 MIPS workstations at $750. Or $2,000. TODAY. Three years from now, when the '486 & and the '040 "trickle down" to us unwashed masses, you may claim that the paltry 8MHz 68000 is "irrelevant." However, today is a far different story, and I would once again like to have some of the profits from these irrelevant machines show up in my bank account. %%%%% Signature v1.1 %%%%% Doug Mohney, Operations Manager, CAD Lab/ME, Univ. of Maryland College Park * Why do VMS system managers get more sleep and less ulcers than their * * UNIX(TM) counterparts, despite the sophistication of UNIX? *
peter@ficc.ferranti.com (Peter da Silva) (12/05/90)
In article <2764@cirrusl.UUCP> dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) writes: > [users] are mostly buying machines running 80286 and 80386 CPUs at 16 > MHz or higher. But running DOS this horsepower is pointless, and under OS/2 or windows too much horsepower is wasted maintaining compatibility with DOS or fighting an inadequate graphics subsystem. A plain jane 68000 at 7.16 MHz provides far better user response time and more functionality if it's not compensating for the mistakes of the past. > there's no comparison with 80x86-based systems That's for sure. -- Peter da Silva. `-_-' +1 713 274 5180. 'U` peter@ferranti.com
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (12/05/90)
In article <3+_7KS1@xds13.ferranti.com> peter@ficc.ferranti.com (Peter da Silva) writes: | But running DOS this horsepower is pointless, and under OS/2 or windows too | much horsepower is wasted maintaining compatibility with DOS or fighting | an inadequate graphics subsystem. When I can get V.4 with compilers, full TCP/NFS, and X-windows for less than $1k, why would I run any of these things? Oh, and that includes DOS under UNIX, so I can run all those MB of cheap/free software, and enough Berkeley compatibility so I can run almost everything I have on a Sun at work, at least if it uses X rather than SunView. | A plain jane 68000 at 7.16 MHz provides far better user response time and | more functionality if it's not compensating for the mistakes of the past. Depends on what you're running. People confuse the CPU with the software, and say they love the computer when all they see is the interface. Running a multitasking o/s on an SX or 68010 makes them look about the same, because they are: both have 32 bits internally, 16 externally. And the [34]86 and 680[23]0. This is a bit blurry, because the chips don't have as good a match as the SX/68010, but the performance is comparable, running the same type of task. A much more interesting question (to me) is "what can be done with the Mac interface to support subdirectories?" The command (diddle slashes for UNIX or DOS) local/bin/graphics/myplotter maps to click/pop hard disk scroll and read, find local click and pop local scroll and read, find bin click and pop bin scroll and read, find graphics click and pop graphics scroll and read, find myplot click and run myplot I just can't feature doing this, and the Mac users all offer solutions which effectively defeat having a tree structure, by putting everything in some low level directory (yes I know these things are all called something else and some clicks above are double clicks). There must be a better way. The pop and click stuff worked well for floppies, it wears a bit thin with a pair of 1.2GB hard drives (which look like two drives rather than one namespace with mounted partitions). -- bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen) VMS is a text-only adventure game. If you win you can use unix.
csd29@seq1.keele.ac.uk (A. Herath) (12/05/90)
From article <1990Nov25.194404.3376@dircon.uucp>, by uad1077@dircon.uucp: > > > Dan Mocsny writes: >> Consider where Apple computer would be right now if it had made >> cloning the Mac easy. Consider where DEC would be if it had made > > While I largely agree with your analysis of Sun's move, Apple are > perhaps the crwoning example that shows there is another way. Apple > often seem to be downplayed by netters, perhaps because they sell > so many machines to non-computer users... It's hard to imagine ^ You mean as show pieces rather than the machines that can be used (for the intended purpose) ? :-?
kls30@duts.ccc.amdahl.com (Kent L Shephard) (12/06/90)
In article <76095@iuvax.cs.indiana.edu> kitchel@iuvax.cs.indiana.edu (Sid Kitchel) writes: -Stuff Deleted > >||Nope. "Classic" here means "obsolete". In this day and age of 28-MIPS >||machines, what Apple does with an 8 MHz 68000 is utterly irrelevant. > -Stuff Deleted > > I'm typing this reply on my 80386-SX Austin at work. (I've been >told that "SX" stands for "sucks", but I cannot confirm this.) For a A 386dx is comprable to a 68030 and a 386sx blows away the 68020 and lesser members of that family. sx does not stand for sucks. I have an sx at home w/4 megs of RAM, 800x600 display, 100 megs of disk storage 3 floppies, midi interface, modem, mouse etc. >turkey, this is a nice, cute little machine. But I must agree with you ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Well since I'm a System Design Engineer, have written simulators for a vector processor of my own design, designed and built couple of complete microprocessor systems, have a BSEE and am currently working on an MSEE, I'm definitely no turkey. My machine is more than just cute, see above. >completely in your: "There's just no comparison..." You are absolutely >correct. For productivity, getting it just the way I want it, for crisp >graphics, desktop publishing, fonts I can read all day, etc. etc. I >much prefer my Mac at home. It has better displays than any Super VGA ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ You obviously haven't seen many Super VGA displays. Remember a Classic only has a 9" screen and something like 540x480 display (????) I know I wouldn't call a Classic's screen high resolution. You have to get a large monitor to get better than the resolution avail. on Super VGA 800x600 and 1024x768. >I've ever seen. The software available for the Mac blows away anything Depends on what you do. A lot of software for the Mac has a IBM PC product that is almost identical, and the opposite is also true. I do know that the engineering packages for design entry, circuit analysis on the Mac are not as mature as products available on PCs. >I have at work. I sure am glad that I don't have to hold my breath >waiting for the OS/2, Windows 3.XX vs. DOS 6 war to end with a clear I don't hold my breath, Windows is a good product and so are a number of other operating system enhancements. If you want something more powerful get Unix on a PC. At least on 80x86 platforms there is a binary standard so that anything compiled under SCO will run under any other Unix. Not so for the 680x0. >winner!! And as a computer architect, I cringe every time I think of ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Then you also cringe when you think of the design of 680x0 machines right ???? >the design of the 80x86 machines. Give me a break and give me a 68xxx ~~~~~~~ >machine with a reasonable user interface NOW!! ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ What is NeXTstep, Motif, Openlook, MS Window????????? That's right you. You like the Mac and well, good for you. Diff. strokes for diff. folks. I like the openess of the PC arch. I also like the NeXT but I don't go around bashing Atari's, Amiga's and Mac's. > --Sid > >-- >Sid Kitchel...............WARNING: allergic to smileys and hearts.... >Computer Science Dept. kitchel@cs.indiana.edu >Indiana University kitchel@iubacs.BITNET >Bloomington, Indiana 47405-4101........................(812)855-9226 -- /* -The opinions expressed are my own, not my employers. */ /* For I can only express my own opinions. */ /* */ /* Kent L. Shephard : email - kls30@DUTS.ccc.amdahl.com */
jerry@TALOS.UUCP (Jerry Gitomer) (12/06/90)
csd29@seq1.keele.ac.uk (A. Herath) writes: >From article <1990Nov25.194404.3376@dircon.uucp>, by uad1077@dircon.uucp: >> >> >> Dan Mocsny writes: >> >> While I largely agree with your analysis of Sun's move, Apple are >> perhaps the crwoning example that shows there is another way. Apple >> often seem to be downplayed by netters, perhaps because they sell >> so many machines to non-computer users... It's hard to imagine > ^ > You mean as show pieces rather than the machines > that can be used (for the intended purpose) ? :-? From what I understand the percentage of unused Macs is no higher than the percentage of unused PCs and clones. In both cases it is reported that about 1/3 of the units shipped are _not_ used. Managers in academia, business, and government like to have a computer sitting on the credenza -- and more often than we like never even turn it on! -- Jerry Gitomer at National Political Resources Inc, Alexandria, VA USA I am apolitical, have no resources, and speak only for myself. Ma Bell (703)683-9090 (UUCP: ...{uupsi,vrdxhq}!pbs!npri6!jerry
coolidge@cs.uiuc.edu (John Coolidge) (12/06/90)
kls30@duts.ccc.amdahl.com (Kent L Shephard) writes: >A 386dx is comprable to a 68030 and a 386sx blows away the 68020 and >lesser members of that family. sx does not stand for sucks. I have >an sx at home w/4 megs of RAM, 800x600 display, 100 megs of disk storage >3 floppies, midi interface, modem, mouse etc. 386sx's are pretty much the same, operationally, as 68020's. I've got a Mac II at home w/8 megs of RAM, 640x480 display (so I'm cheap :-)), 470 megs of disk storage, 2 floppies, 2 modems, mouse, printer etc. Peripherals don't prove very much. Computationally, it's pretty hard to tell the difference. >You obviously haven't seen many Super VGA displays. Remember a Classic >only has a 9" screen and something like 540x480 display (????) I know >I wouldn't call a Classic's screen high resolution. You have to get >a large monitor to get better than the resolution avail. on Super VGA >800x600 and 1024x768. You're confusing resolution with pixels. The resolution of a Macintosh monitor (any Macintosh monitor from Apple and almost all third party monitors) is 72+/- 1 dpi. >>winner!! And as a computer architect, I cringe every time I think of > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ >Then you also cringe when you think of the design of 680x0 machines >right ???? The 680x0 is a much cleaner architecture than the 80x86, if for no other reason than that the backward compatibility is limited to a reasonable processor (8086! gimme a break!). --John -------------------------------------------------------------------------- John L. Coolidge Internet:coolidge@cs.uiuc.edu UUCP:uiucdcs!coolidge Of course I don't speak for the U of I (or anyone else except myself) Copyright 1990 John L. Coolidge. Copying allowed if (and only if) attributed. You may redistribute this article if and only if your recipients may as well.
mcdonald@aries.scs.uiuc.edu (Doug McDonald) (12/06/90)
In article <3+_7KS1@xds13.ferranti.com> peter@ficc.ferranti.com (Peter da Silva) writes: >In article <2764@cirrusl.UUCP> dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) writes: >> [users] are mostly buying machines running 80286 and 80386 CPUs at 16 >> MHz or higher. > >But running DOS this horsepower is pointless That is simply false. Try running Autocad or TeXing my book, or previewing the TeX output. Or integrating 24 simultaneous linear ODEs for a million steps. For those tasks a 20 MHz 386 is enough power, but most certainly not too much. Doug McDonald
fargo@iear.arts.rpi.edu (Irwin M. Fargo) (12/06/90)
In article <1990Dec6.035617.4873@ux1.cso.uiuc.edu> mcdonald@aries.scs.uiuc.edu (Doug McDonald) writes: >In article <3+_7KS1@xds13.ferranti.com> peter@ficc.ferranti.com (Peter da Silva) writes: >>In article <2764@cirrusl.UUCP> dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) writes: >>> [users] are mostly buying machines running 80286 and 80386 CPUs at 16 >>> MHz or higher. >> >>But running DOS this horsepower is pointless > > >That is simply false. Try running Autocad or TeXing my book, >or previewing the TeX output. Or integrating 24 simultaneous linear >ODEs for a million steps. For those tasks a 20 MHz 386 is enough >power, but most certainly not too much. > >Doug McDonald Although I'm not a big MuShy-DOS fan, I do feel MS-DOS has one advantage over most other OS's for the Intel based PC machines: it was poorly designed. This has basically allowed developers to mix design modes. Most programs will do screen I/O by writing directly to video memory while file I/O is done using MS-DOS system calls. Being able to bypass the OS, of course, is de- pendent on compatibility between hardware systems, but it's there, and it allows a developer to take full advantage of the available computing power, albeit, with a possible trade-off in portability. My point, simply, is that in most cases, MS-DOS, by being poorly designed, has provoked developers to bypass the OS and give more horsepower to the user. Mind you, I'll be getting myself an 80386 machine soon and hopefully I'll be running UNIX with X-Windows. Ooh! So scarey! Thank you and happy hunting! Actually: Ethan M. Young "If Linus looked like Worf, would you Internet: fargo@iear.arts.rpi.edu try to take his blanket away?" Bitnet (??): usergac0@rpitsmts.bitnet - dorsai@pawl.rpi.edu Disclaimer: Who said what?
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (12/06/90)
In article <c82A02WA03kX01@JUTS.ccc.amdahl.com> kls30@DUTS.ccc.amdahl.com (Kent L. Shephard) writes: | That's right you. You like the Mac and well, good for you. Diff. strokes | for diff. folks. I like the openess of the PC arch. I also like the | NeXT but I don't go around bashing Atari's, Amiga's and Mac's. The Mac is an example os "central planning," while X-windows is an example of "free market" design. On the Mac all applications look the same, and having learned one you have a start on learning others. In X you choose a window manager which makes the universe look the way you want it. There are good and bad points to both approaches, and a belief that either is superior in all cases will not give a useful view of the situation. I'm going to assume that user interface design is a valid topic, even if a lot of stuff before has just been flamage. The Mac approach presnts a consistant interface, with reduced user training. Unfortunately, if the user doesn't *like* that single unchangable interface, s/he is just out of luck. The idea is that the designers know best what's good for the user, and the user should adapt or die. The X approach assumes the user will find an interface which enhances productivity. In at least some cases this isn't true. The user will spend a lot of time with various interfaces, never learning any one enough to be productive. The up side is that the typical user will choose an interface, and use that window manager which matches the way the user's mind works. That's a big boost to productivity. The power user may use various interfaces for certain tasks, and will benefit from the best parts of all of them. Let's stop worrying about which is better, because neither is perfect and they reflect two approaches to delivering resources, and do not compete, in that they solve diferent problems, even though they may look like the same problem to the user. I really believe that if Mac, X, NeWS, and Windows3 had all started even, at the same time with no installed base, and portable, that a clear winner would have emerged, and neither X nor Mac would not be the winner. X because the promise of cheap dumb terminals is false, the CPU and memory to run X will also run NeWS at a lower bandwidth between client and server, and Mac because it lacks a command interface for those cases where there are too many choices for a comfortable menu. This is my opinion, feel free to discuss it, but don't flame me about it, I never claimed it was the only true truth, just that it looks that way to me. If Mac hadn't been first, they would have had another Lisa on their hands. -- bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen) VMS is a text-only adventure game. If you win you can use unix.
peter@ficc.ferranti.com (Peter da Silva) (12/07/90)
In article <c82A02WA03kX01@JUTS.ccc.amdahl.com> kls30@DUTS.ccc.amdahl.com (Kent L. Shephard) writes: > At least on 80x86 platforms there is a binary standard so that anything > compiled under SCO will run under any other Unix. You haven't been following the "SCO isn't UNIX" thread on c.u.sysv386, have you? Seriously, there is a 68000 Sys V ABI. You just need to find a machine that follows it... But, really, for UNIX the day of the CISC processor is pretty much over. Neither the 80x86 nor the 68000 are significantly price-competitive with the low end Sparcs, and they're looking less exciting all the time. -- Peter da Silva. `-_-' +1 713 274 5180. 'U` peter@ferranti.com
peter@ficc.ferranti.com (Peter da Silva) (12/07/90)
In article <1990Dec6.035617.4873@ux1.cso.uiuc.edu> mcdonald@aries.scs.uiuc.edu (Doug McDonald) writes: > In article <3+_7KS1@xds13.ferranti.com> peter@ficc.ferranti.com (Peter da Silva) writes: > >In article <2764@cirrusl.UUCP> dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) writes: > >> [users] are mostly buying machines running 80286 and 80386 CPUs at 16 > >> MHz or higher. > >But running DOS this horsepower is pointless > That is simply false. Try [ eminently high-power applications ] OK, I'll amend that. Running DOS this horsepower is mostly pointless. We have DOS machines with 80386es in them here and all they run is Procomm and word processors. I.e, for the users who are "mostly buying" hot boxes they would be as well served with an XT class machine. I have at home a Xerox 820. 64K, Z80, CP/M. It runs kermit and Select and does as good a job at terminal emulation and light editing as the Compaq Deskpro in the office next door. -- Peter da Silva. `-_-' +1 713 274 5180. 'U` peter@ferranti.com
kls30@duts.ccc.amdahl.com (Kent L Shephard) (12/07/90)
In article <1990Dec6.005211.7490@julius.cs.uiuc.edu> coolidge@cs.uiuc.edu (John Coolidge) writes: > >386sx's are pretty much the same, operationally, as 68020's. I've got I don't agree because you have to get a PMMU on a 68020 which adds wait states to memory. But besides that they are equivalent. > >>You obviously haven't seen many Super VGA displays. Remember a Classic >>only has a 9" screen and something like 540x480 display (????) I know >>I wouldn't call a Classic's screen high resolution. You have to get >>a large monitor to get better than the resolution avail. on Super VGA >>800x600 and 1024x768. > >You're confusing resolution with pixels. The resolution of a Macintosh >monitor (any Macintosh monitor from Apple and almost all third party >monitors) is 72+/- 1 dpi. No I'm not, resolution is the number of pixels, dot pitch determines dpi. Look at any number of publications and manufacturer's literature and resolution is listed as 800x600, 1024x768, etc. Almost all SVGA screens have dpi greater than 72 +/- 1. The dot pitches for SVGA class monitors starts at .31mm (80 dpi) and the other standards are .28mm(90 dpi) and .25mm (100 dpi). BTW .31mm is considered not good for a SVGA since standard EGA and VGA had .31mm and .28mm dot pitches. By todays standard a monitor w/ a dot pitch of .35mm (72 dpi) is considered corse and grainy. 72 dpi IS corse dot pitch. The standard Mac IS low resolution and just looking at specs is definitly not better SVGA or VGA for that matter. > >>>winner!! And as a computer architect, I cringe every time I think of >> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ >>Then you also cringe when you think of the design of 680x0 machines >>right ???? > >The 680x0 is a much cleaner architecture than the 80x86, if for >no other reason than that the backward compatibility is limited to >a reasonable processor (8086! gimme a break!). If I'm not mistaken the 68k is circa 8086. BTW what is wrong w/ backward compatibility in a processor. If you don't have backward compatibility you've got a different processor family. The 68040 is a kluge just like the i486 both use a RISC core wrapped up in a CISC instruction set. The 68k arch. is not better that the 80x86 they're just different. > >--John Before you start blasting away at arch. and screens check you sources. Mine are industry standard for what is called resolution and dot pitch. KeNT >-------------------------------------------------------------------------- >John L. Coolidge Internet:coolidge@cs.uiuc.edu UUCP:uiucdcs!coolidge >Of course I don't speak for the U of I (or anyone else except myself) >Copyright 1990 John L. Coolidge. Copying allowed if (and only if) attributed. >You may redistribute this article if and only if your recipients may as well. -- /* -The opinions expressed are my own, not my employers. */ /* For I can only express my own opinions. */ /* */ /* Kent L. Shephard : email - kls30@DUTS.ccc.amdahl.com */
renglish@hplabsz.HPL.HP.COM (Bob English) (12/07/90)
In article <FNA^KA@rpi.edu> fargo@iear.arts.rpi.edu (Irwin M. Fargo) writes: >My point, simply, is that in most cases, MS-DOS, by being poorly designed, >has provoked developers to bypass the OS and give more horsepower to the >user. On the other hand, one could make a strong case a poorly designed OS provokes developers to bypass the OS and get more horsepower from the hardware, thus making it difficult to take advantage of future advances in hardware technology, discouraging innovation, and slowing progress for all concerned. --bob-- renglish@hplabs "If I owned the company, even the door handles would be cool."
coolidge@cs.uiuc.edu (John Coolidge) (12/07/90)
kls30@duts.ccc.amdahl.com (Kent L Shephard) writes: >In article <1990Dec6.005211.7490@julius.cs.uiuc.edu> coolidge@cs.uiuc.edu (John Coolidge) writes: >>386sx's are pretty much the same, operationally, as 68020's. I've got >I don't agree because you have to get a PMMU on a 68020 which adds wait >states to memory. But besides that they are equivalent. I was comparing after the PMMU, of course. I'm running Unix on my '020 most of the time. >>You're confusing resolution with pixels. The resolution of a Macintosh >>monitor (any Macintosh monitor from Apple and almost all third party >>monitors) is 72+/- 1 dpi. >No I'm not, resolution is the number of pixels, dot pitch determines >dpi. Look at any number of publications and manufacturer's literature >and resolution is listed as 800x600, 1024x768, etc. Manufacturers who don't understand English, either conversational or technical. Resolution has a very clear and precise meaning, and it isn't the number of dots on your (arbitrarily big) monitor. Resolution, in monitors as in printers, is the number of dots placed in a given area. Dot pitch and dpi are both measures of resolution, dots on the screen is not. The term is based on the idea of resolving a point --- how precisely can you display images of a given size? Clearly the number of dots on your display impacts this --- more dots on a display of the same size affects resolution. Adding more dots, but making the screen bigger, doesn't change resolution. It just gives you the ability to put more things on your screen at the same resolution. I'll concede, however, that this is a common mistake, similar to the misuse of hacker as cracker, MIPs as a measure of anything useful, and so forth. I'll also drop the point, since I don't want this to become a terminology flame. Others can pick up the discussion from here. >>The 680x0 is a much cleaner architecture than the 80x86, if for >>no other reason than that the backward compatibility is limited to >>a reasonable processor (8086! gimme a break!). >If I'm not mistaken the 68k is circa 8086. BTW what is wrong w/ backward >compatibility in a processor. If you don't have backward compatibility >you've got a different processor family. Backward compatibility is great and wonderful, as long as it doesn't cost you anything. Staying compatible with a 1980 (or earlier) design (the 8086) is clearly straining the 80x86 family of processors. The 680x0 series is based on a 1982-3 design (the 68000) (BTW, the Lisa, which uses the 68000, came out in 1983 I believe. Possibly 1982). Those two years are one of the reasons why a 68030 is only about twice as slow as a 80486 at the same clock speed and why a 68040 is about half again as fast as the 80486. The 680x0 series maintains backward compatibility with the 68000, which was two years later and based on a number of lessons learned in those two years (as well as borrowing a few more good ideas from the mainframes of yore, from whence quite a lot of microcomputer 'advances' come). >The 68040 is a kluge just like the i486 both use a RISC core wrapped up >in a CISC instruction set. The 68k arch. is not better that the 80x86 >they're just different. I agree; they're both 'kludges' in the sense you describe. They're both processor families that are popular but aging. There are clearly better processors around then either of them by a number of criteria. But, comparing only the two, I believe the 680x0 series to be a better architecture than the 80x86 series, at least from a programmer's point of view (I'm not a hardware designer, I'm an operating systems person. I know which series I'm happier working with). >Before you start blasting away at arch. and screens check you sources. >Mine are industry standard for what is called resolution and dot pitch. I have checked my sources. They indicate (you guessed it) that, on the subject of resolution, I'm correct. On the comparison between 80x86 and 680x0, there's clearly a lot of opinion involved in the comparison. On the factual questions I stand by my sources. --John -------------------------------------------------------------------------- John L. Coolidge Internet:coolidge@cs.uiuc.edu UUCP:uiucdcs!coolidge Of course I don't speak for the U of I (or anyone else except myself) Copyright 1990 John L. Coolidge. Copying allowed if (and only if) attributed. You may redistribute this article if and only if your recipients may as well.
kls30@duts.ccc.amdahl.com (Kent L Shephard) (12/07/90)
In article <1990Dec7.030303.20157@julius.cs.uiuc.edu> coolidge@cs.uiuc.edu writes: >kls30@duts.ccc.amdahl.com (Kent L Shephard) writes: >>In article <1990Dec6.005211.7490@julius.cs.uiuc.edu> coolidge@cs.uiuc.edu (John Coolidge) writes: > > > >>>You're confusing resolution with pixels. The resolution of a Macintosh >>>monitor (any Macintosh monitor from Apple and almost all third party >>>monitors) is 72+/- 1 dpi. > >>No I'm not, resolution is the number of pixels, dot pitch determines >>dpi. Look at any number of publications and manufacturer's literature >>and resolution is listed as 800x600, 1024x768, etc. > >Manufacturers who don't understand English, either conversational ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Well I guess NeXT, IBM, PC Magazine, Sony, Mitsubishi, NEC, Relisys, Accuvue, JVC, and Nanao all need to go back and learn english. Resolution in standard terms, used in manufacturers literature states resolution is (some num.)x(some num.) and dpi is specified seperately. So I'll say that resolution is a function of both dpi and number of dots on the screen. >or technical. Resolution has a very clear and precise meaning, and >it isn't the number of dots on your (arbitrarily big) monitor. >Resolution, in monitors as in printers, is the number of dots placed >in a given area. Dot pitch and dpi are both measures of resolution, >dots on the screen is not. The term is based on the idea of resolving >a point --- how precisely can you display images of a given size? >Clearly the number of dots on your display impacts this --- more dots >on a display of the same size affects resolution. Adding more dots, >but making the screen bigger, doesn't change resolution. It just >gives you the ability to put more things on your screen at the same >resolution. > But you still haven't addressed the argument you made that said a 72 dpi Mac monitor is better than a SVGA monitor on an IBM system. Which is definitely false. Once agin SVGA moitors range from 80-100dpi which translates .32mm to .25mm per pixel. > >Those two years are one of the reasons why a 68030 is only about twice >as slow as a 80486 at the same clock speed and why a 68040 is about >half again as fast as the 80486. The 680x0 series maintains backward ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I'll believe this when there are more than a handfull of '040 in systems. I know that Motorola started shipping in volume and as soon as the systems are out there to compare (running ral applications) I'll believe it. What Motorola or any manufac. claims as mips is not valid. Also speed is going to depend a lot on operating system overhead and compiler design. I'll venture to say that they are comperable processors and leave it at that. >compatibility with the 68000, which was two years later and based on >a number of lessons learned in those two years (as well as borrowing >a few more good ideas from the mainframes of yore, from whence quite a >lot of microcomputer 'advances' come). > >>The 68040 is a kluge just like the i486 both use a RISC core wrapped up >>in a CISC instruction set. The 68k arch. is not better that the 80x86 >>they're just different. > >I agree; they're both 'kludges' in the sense you describe. They're >both processor families that are popular but aging. There are clearly >better processors around then either of them by a number of criteria. >But, comparing only the two, I believe the 680x0 series to be a better >architecture than the 80x86 series, at least from a programmer's point >of view (I'm not a hardware designer, I'm an operating systems person. >I know which series I'm happier working with). > >>Before you start blasting away at arch. and screens check you sources. >>Mine are industry standard for what is called resolution and dot pitch. > >I have checked my sources. They indicate (you guessed it) that, on >the subject of resolution, I'm correct. On the comparison between ~~~~~~~~~~~ I'm looking at a bunch of stuff that people call industry standard reviews, product literature. What are your sources??? I quoted mine. See above. >80x86 and 680x0, there's clearly a lot of opinion involved in the >comparison. On the factual questions I stand by my sources. ~~~~~~~~~~~~ What sources?????? > >--John > >-------------------------------------------------------------------------- >John L. Coolidge Internet:coolidge@cs.uiuc.edu UUCP:uiucdcs!coolidge >Of course I don't speak for the U of I (or anyone else except myself) >Copyright 1990 John L. Coolidge. Copying allowed if (and only if) attributed. >You may redistribute this article if and only if your recipients may as well. -- /* -The opinions expressed are my own, not my employers. */ /* For I can only express my own opinions. */ /* */ /* Kent L. Shephard : email - kls30@DUTS.ccc.amdahl.com */
Bruce.Hoult@bbs.actrix.gen.nz (12/08/90)
In article <3005@crdos1.crd.ge.COM> davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) writes: > The Mac is an example os "central planning," while X-windows is an >example of "free market" design. On the Mac all applications look the >same, and having learned one you have a start on learning others. > > The Mac approach presnts a consistant interface, with reduced user >training. Unfortunately, if the user doesn't *like* that single >unchangable interface, s/he is just out of luck. The idea is that the >designers know best what's good for the user, and the user should adapt >or die. There seems to be a common misconception that Apple has somehow built the Mac in such a way that you *have* to use their interface. This is totally incorrect. Apple provide a set of tools for building elements of an interface, but: a) it takes a h*ll of a lot of user code to take the toolbox and produce a "standard" Mac interface. b) the programmer is free to totally or partially ignore what has been provided and do things herself. c) the user can change a lot of element of the interface by substituting different versions of the standard code for drawing menus, windows, buttons etc etc. There exist sets of such code that can be installed to give the Mac a new, but still consistant feel -- for example there are several such sets of code designed to make the Mac mimic the NeXT user interface (Black Box, for example). The only way that Apple force you to follow their interface is by gentle (and not-so-gentle) persuasion, and by making the desired interface the path of least resistance (note especilly the MacApp effort to make a standard interface easier to provide). >Mac because it lacks a command interface The Mac doesn't have a command interface supplied with it in the box, but there are CLI's available for it. MPW is the obvious example. -- Bruce.Hoult@bbs.actrix.gen.nz Twisted pair: +64 4 772 116 BIX: brucehoult Last Resort: PO Box 4145 Wellington, NZ
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (12/08/90)
In article <QK+78Z9@xds13.ferranti.com> peter@ficc.ferranti.com (Peter da Silva) writes: | But, really, for UNIX the day of the CISC processor is pretty much over. | Neither the 80x86 nor the 68000 are significantly price-competitive with | the low end Sparcs, and they're looking less exciting all the time. But... what are you comparing? If it's dickless workstations, there are virtually no vendors of them with CISC chips. You win by default. If you talk about free-standing systems, the 486 wins hands down. There are a number of vendors selling 486 systems at a price which allows the total system, with 300+MB disk, 8-12MB RAM, 800x600x256 display, UNIX, compilers, and X to come in under $6k. With a bit of very careful shopping you can come in pennies under $5k. And since you said "low end" in your posting, I have no hesitation in saying that the system is faster for integer, and about as fast for float. The 25MHz 486 SPECs between the SS and SS+ as I recall. I'm still trying to get a useful personal system with an SX for $3k with software, and I'm getting real close. Then project cheap-ix will be complete, and I can stop building, testing, and selling off systems in my spare room. -- bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen) VMS is a text-only adventure game. If you win you can use unix.
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (12/08/90)
In article <d48f027P03Sc01@JUTS.ccc.amdahl.com> kls30@DUTS.ccc.amdahl.com (PUT YOUR NAME HERE) writes: | >Manufacturers who don't understand English, either conversational | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | Well I guess NeXT, IBM, PC Magazine, Sony, Mitsubishi, NEC, Relisys, | Accuvue, JVC, and Nanao all need to go back and learn english. Resolution | in standard terms, used in manufacturers literature states resolution | is (some num.)x(some num.) and dpi is specified seperately. So I'll | say that resolution is a function of both dpi and number of dots on the | screen. Manufacturers ad agencies will say anything which sells hardware. The better ones don't lie, but they all use confusing verbiage and quote totally meaningless facts. By the standard of number of pixels, the bunch of drunks at a football game holding up flash cards a meter square have more resolution than the best display you can buy. If you want 6000x500 "resulution" and a *real long* persistance, this is the display for you. I have been guilty of using this terminology, but I at least understand that it is a misnomer and would not try to defend it as correct usage. -- bill davidsen (davidsen@crdos1.crd.GE.COM -or- uunet!crdgw1!crdos1!davidsen) VMS is a text-only adventure game. If you win you can use unix.
rhealey@digibd.com (Rob Healey) (12/08/90)
In article <2760@cirrusl.UUCP> dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) writes: >>Maybe they've learnt at last. The new Classic costs $750 at the Stanford >>Bookstore, and the waiting list for them is 2 months long. >Nope. "Classic" here means "obsolete". In this day and age of 28-MIPS >machines, what Apple does with an 8 MHz 68000 is utterly irrelevant. Not to burst any bubbles but, my toaster oven and thermostat don't need 28MIPS of CPU power to function. Hell, I'll even throw in my microwave, 'fridge, VCR, Stereo, TV and dishwasher! Think about it. A Mac Classic is a household appliance, NOT a "real" computer per say. What is joe average going to DO with 28MIPS? The average person doesn't NEED that much power. My Color Computer 3 running OS/9 costs about $400.00 and can put a PC to shame in most cases. It runs an "obsolete" Motorola 6809E 8 bit processor and is a VERY useful system! For home use my CoCo is more than enough computing power and a Mac Classic has even more than that although it lacks color. We aren't taking rocket scientists here kids, we're talking Numbtendo addicts and video junkies. -Rob Speaking for self, not company.
coolidge@cs.uiuc.edu (John Coolidge) (12/08/90)
kls30@duts.ccc.amdahl.com (Kent L Shephard) writes: >But you still haven't addressed the argument you made that said a 72 dpi >Mac monitor is better than a SVGA monitor on an IBM system. Which >is definitely false. Once agin SVGA moitors range from 80-100dpi which >translates .32mm to .25mm per pixel. Please pay more attention to your attributions. I never made the original comment; my remarks were limited to the misuse of the term 'resolution'. As a matter of fact, there are a number of SVGA monitors which are better than the average Macintosh monitor. There are also a number of Mac monitors which, while having fewer dots per inch, are sharper than many SVGA monitors (because many SVGA monitors that I've seen are very fuzzy). Fuzziness turns dots per inch into spots per inch and can lower the effective dpi. However, I agree that SVGA monitors are, quite often, higher resolution than Mac monitors whether the Mac monitors are 512x384 or 1152x870. >>Those two years are one of the reasons why a 68030 is only about twice >>as slow as a 80486 at the same clock speed and why a 68040 is about >>half again as fast as the 80486. The 680x0 series maintains backward >~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ >I'll believe this when there are more than a handfull of '040 in systems. >I know that Motorola started shipping in volume and as soon as the >systems are out there to compare (running ral applications) I'll >believe it. What Motorola or any manufac. claims as mips is not valid. I agree; manufacturer claims are notoriously overrated. I'm basing my speed estimate on benchmarks I've seen run on real 68040 boxes and real 80486 boxes. The 68040 seems to be about three times as fast as the 68030 at the same clock speed; hence, it is about half again as fast as the 80486. >I'm looking at a bunch of stuff that people call industry standard >reviews, product literature. What are your sources??? I quoted mine. On 'resolution': check several of the papers done by the X Window system implementers. For that matter, run xdpyinfo on your favorite machine and see what it reports for resolution. Or check a number of papers from SIGGRAPH or other graphics journals. Or check other postings in this string... On the 680x0/80x86: personal experience with machines using them for benchmarking; processor manuals for instruction sets; opinion on which I'd rather be working with. --John -------------------------------------------------------------------------- John L. Coolidge Internet:coolidge@cs.uiuc.edu UUCP:uiucdcs!coolidge Of course I don't speak for the U of I (or anyone else except myself) Copyright 1990 John L. Coolidge. Copying allowed if (and only if) attributed. You may redistribute this article if and only if your recipients may as well.
khb@chiba.Eng.Sun.COM (Keith Bierman fpgroup) (12/08/90)
In article <1990Dec07.194407.29083@digibd.com> rhealey@digibd.com (Rob Healey) writes:
... Not to burst any bubbles but, my toaster oven and thermostat
....
We aren't taking rocket scientists here kids, we're talking
Numbtendo addicts and video junkies.
Consider the humble camera. Used to be very simple, a box with a hole.
Good enough for most purposes. OK, so we evolved lenses. Then, for
quite a time the plateau was reached (for the medium-high end) with
35mm SLR cameras. Of course, it took a bit of savvy to get good
results and lighting was critical....
Then along came folks like Minolta. 3 microprocessors, a bevy of
sensors, communication channel for outboard devices (flash) and now
nearly anyone can walk into a completely darkened room, point and
click and get a perfectly focused picture (* it really works with the
'smart' flash).
In order to make computers "simple" and "easy" to use, and accessible
to the bulk of "nintendo junkies" (read folks who have better ways to
spend their time than parsing ALGOL68 docsets) is by building vastly
more powerful systems. 28mips (whatever they are) is several hundreds
of orders of magnitude too small for a really user friendly system.
A friend of mine has a good example. He goes to his favorite pub (duke
of endinburgh, for those who know the sillycone valley area) and waves
to his favorite server. She looks him over, and picks out what he
wants to drink ... based on his mood.
He claims she's almost never wrong.
I want my computer to be at least half as good. (and it should do the
dishes ;>).
--
----------------------------------------------------------------
Keith H. Bierman kbierman@Eng.Sun.COM | khb@chiba.Eng.Sun.COM
SMI 2550 Garcia 12-33 | (415 336 2648)
Mountain View, CA 94043dhesi%cirrusl@oliveb.ATC.olivetti.com (Rahul Dhesi) (12/09/90)
In <1990Dec07.194407.29083@digibd.com> rhealey@digibd.com (Rob Healey) writes: > My Color Computer 3 running OS/9 costs about $400.00 > and can put a PC to shame in most cases. It runs an "obsolete" > Motorola 6809E 8 bit processor and is a VERY useful system! Actually the point made is a good one. The Motorola 6809 was very carefully designed to allow position-independent code. As a result, a UNIX-like OS (OS9), which can create multiple processes without memory-management hardware, was easy to write for this CPU. OS9 deserves more attention than it gets. It's a pity IBM didn't pick the 6809 (maybe it came after the 8086?) All this is, of course, irrelevant if you're talking about Mac Classics. -- Rahul Dhesi <dhesi%cirrusl@oliveb.ATC.olivetti.com> UUCP: oliveb!cirrusl!dhesi