johnm@cup.portal.com (John - Madison) (12/14/90)
Assertion: Sun is retreating from the idea of Open Systems. Consider: - Sun has decided *not* to provide future versions of SunOS to clone makers. They will instead be forced to buy System V from ATT. This will allow Sun to produce proprietary systems and claim that the clone manufacturers are not compatible. - Sun has been aggressively unbundling software like the C compiler, OS improvements, and such. Further, they *prohibit* you from buying the unbundled Sun products and running them on a clone. Not supporting such products is one thing, but a blanket prohibition is clearly aimed at the clone makers. - Sun is providing marginal support for SBus card manufacturers. While publishing specs and claiming that it is fully open, their technical support leaves a lot to be desired. In addition, Sun has recently been pretty aggressive in competing with third party SBus card manufacturers and has been developing products to skim the cream by claiming the best selling cards for themselves. Most SBus developers are small companies and need the revenue from high volume cards to plow back into r&d. Taking this market away from them will insure that no SBus developers grows to be very large, and may well discourage many third party developers. After all, if you had an idea for a card, would you develop and market it, if you thought that Sun would jump into the market as soon as you demonstrated the concept? According to (unnamed for obvious reasons) marketing people at Sun, they are very concerned about clone makers, both at the low and high end, and are trying to figure out ways to frustrate them. Instead of standard approaches like faster machines, lower prices, and better support, they are using legal strategies like jerking the vendors around on licensing. If you want to learn the novel forms of competitive advantage that Sun is seeking, talk to Solborne or Auspex. In conclusion: when Sun was small, they supported open systems since it was a good tool to compete against larger, closed systems. Now that they are much larger and more able to set the standards, they have no more incentive to support truly open systems than any other company. Open Systems has gone from a strategy to a vapid marketing slogan.
chuck@trantor.harris-atd.com (Chuck Musciano) (12/14/90)
In article <36911@cup.portal.com>, johnm@cup.portal.com (John - Madison) writes: > Assertion: Sun is retreating from the idea of Open Systems. > > Consider: > - Sun has decided *not* to provide future versions of SunOS to clone > makers. They will instead be forced to buy System V from ATT. This > will allow Sun to produce proprietary systems and claim that the clone > manufacturers are not compatible. But Sun will be running SVr4 themselves. They are pretty much locked into SVr4, and in fact, SunOS 5.0 will be SVr4. So why shouldn't the clones get SVr4 from the real source, USL, instead of Sun? Sun has no obligation to support clone makers. All they did was create a market where cloners could compete. They don't have to make it easy for them. > - Sun has been aggressively unbundling software like the C compiler, > OS improvements, and such. Further, they *prohibit* you from buying > the unbundled Sun products and running them on a clone. Not supporting > such products is one thing, but a blanket prohibition is clearly aimed > at the clone makers. I hate to see the compiler completely unbundled, which is a mistake. But again, what obligation does Sun have to a clone maker? How technical support does IBM give Compaq? Also, can you cite the licensing agreement which restricts Sun products to Sun hardware? Perhaps there is a market for clones of unbundled Sun software! > - Sun is providing marginal support for SBus card manufacturers. While > publishing specs and claiming that it is fully open, their technical > support leaves a lot to be desired. In addition, Sun has recently been > pretty aggressive in competing with third party SBus card manufacturers > and has been developing products to skim the cream by claiming the best > selling cards for themselves. Most SBus developers are small companies > and need the revenue from high volume cards to plow back into r&d. > Taking this market away from them will insure that no SBus developers > grows to be very large, and may well discourage many third party > developers. After all, if you had an idea for a card, would you > develop and market it, if you thought that Sun would jump into the > market as soon as you demonstrated the concept? If you can't stand the heat, get out of the kitchen. Sun has no obligation to not compete in the SBus market, nor do they have an obligation to support all SBus vendors. The fact that more than 100 SBus cards are already available seems to show that small companies can compete in the SBus market. > According to (unnamed for obvious reasons) marketing people at Sun, > they are very concerned about clone makers, both at the low and high > end, and are trying to figure out ways to frustrate them. Instead of > standard approaches like faster machines, lower prices, and better > support, they are using legal strategies like jerking the vendors > around on licensing. If you want to learn the novel forms of > competitive advantage that Sun is seeking, talk to Solborne or Auspex. Sun should be concerned, and if I were Sun, I'd do everything I can to compete and succeed. You seem to think that Sun should roll over and give up market share to clone vendors. To be honest, I'm seeing faster machines, lower prices, and better support from Sun. And in this day and age, I can't blame them from using legal tactics, too. > In conclusion: when Sun was small, they supported open systems since it > was a good tool to compete against larger, closed systems. Now that > they are much larger and more able to set the standards, they have no > more incentive to support truly open systems than any other company. > Open Systems has gone from a strategy to a vapid marketing slogan. The point of open systems is to foster a market where software runs everywhere, and people buy systems because of perceived advantages from their vendor. A company that produces a SPARC 1+ knockoff with no distinct differences from existing Sun (or other) products deserves to die, because they are not differentiating themselves in the market place. Sun doesn't owe anyone a free ride. Only truly innovative companies will win a market niche. There may be 42 SPARC clone vendors right now, but you can be sure there won't be in two years. Only companies like Solbourne (multiprocessing), Toshiba (laptops), Matsushita (multi-media), FPS (supercomputing) and maybe Mars or RDI (PC and Mac compatibility) will survive this very cut-throat marketplace. Of course, if you don't like the way Sun does business, switch to any other vendor who has over 150,000 installed RISC systems (and expects 250,000 by June '91 and 1,000,000 by Dec '92), a thriving clone market, 3,000 available applications, and a 40% market share. :-) -- Chuck Musciano ARPA : chuck@trantor.harris-atd.com Harris Corporation Usenet: ...!uunet!x102a!trantor!chuck PO Box 37, MS 3A/1912 AT&T : (407) 727-6131 Melbourne, FL 32902 FAX : (407) 729-2537 A good newspaper is never good enough, but a lousy newspaper is a joy forever. -- Garrison Keillor
croft@csusac.csus.edu (Steve Croft) (12/14/90)
The other points may be valid, but I don't think the fact that Sun will not license their OS anymore is not a sign of "un-openess". Starting with SVR4, Sun will be basically "adding value" to the AT&T product, just like any other UNIX vendor. So Sun will now require that clone makers buy UNIX from AT&T and add their own value (which several are doing). And just because clone makers won't have access to the features that Sun adds, doesn't mean they can't compete. There are several areas in the OS that can be done better than the way Sun did it. As for the general feeling that Sun is getting scared of the clone makers... well, they *better* be! :) Steve stevec@water.ca.gov
lamson@sierra.crd.ge.com (scott h lamson) (12/14/90)
>From: chuck@trantor.harris-atd.com (Chuck Musciano) > Also, can you cite the licensing agreement >which restricts Sun products to Sun hardware? Perhaps there is a market for Last I heard, Sun had exclusive marketing rights to Lucid Lisp for SPARC architectures, but refused to sell it to Solbourne customers. They seem to want it both ways here. Lisp is a third party software product. Also Sun refused to sell SunPhigs, a Sun product, to Solbourne customers. The license agreement is VERY specific to SUN hardware, not SPARC hardware. -- Scott| ARPA: lamson@crd.ge.com Lamson| UUCP: uunet!crd.ge.com!lamson (518)387-5795| UUCP: uunet!sierra.crd.ge.com!lamson General Electric Corporate Research and Development
henry@zoo.toronto.edu (Henry Spencer) (12/15/90)
In article <5089@trantor.harris-atd.com> chuck@trantor.harris-atd.com (Chuck Musciano) writes: >... A company that produces a SPARC 1+ knockoff with no distinct >differences from existing Sun (or other) products deserves to die, because >they are not differentiating themselves in the market place... Nonsense. Differentiation can be done on things like price, support, quality, bundled software, and the like rather than on the shape of the box. That's how the free market works in other areas. It could really stand to work this way in SPARCs, too. -- "The average pointer, statistically, |Henry Spencer at U of Toronto Zoology points somewhere in X." -Hugh Redelmeier| henry@zoo.toronto.edu utzoo!henry
prc@erbe.se (Robert Claeson) (12/17/90)
In article <5089@trantor.harris-atd.com> chuck@trantor.harris-atd.com (Chuck Musciano) writes: >In article <36911@cup.portal.com>, johnm@cup.portal.com (John - Madison) writes: >> - Sun has decided *not* to provide future versions of SunOS to clone >> makers. They will instead be forced to buy System V from ATT. This >> will allow Sun to produce proprietary systems and claim that the clone >> manufacturers are not compatible. > But Sun will be running SVr4 themselves. They are pretty much locked >into SVr4, and in fact, SunOS 5.0 will be SVr4. As Sun will tell you if you ask, their SVR4 o/s will be *based* on SVR4, but will also contain numerous proprietary extensions "to ensure our leadership". Also, the SPARC version of SVR4 that anyone can license from AT&T will not have the SunOS 4.x compatibility support built-in. Clone vendors will have to construct that part themselves, if they want to be able to run programs from SunOS. -- Robert Claeson |Reasonable mailers: rclaeson@erbe.se ERBE DATA AB | Dumb mailers: rclaeson%erbe.se@sunet.se Jakobsberg, Sweden | Perverse mailers: rclaeson%erbe.se@encore.com Any opinions expressed herein definitely belongs to me and not to my employer.
pjg@acsu.buffalo.edu (Paul Graham) (12/17/90)
johnm@cup.portal.com (John - Madison) writes: |Assertion: Sun is retreating from the idea of Open Systems. |- Sun has decided *not* to provide future versions of SunOS to clone |makers. sun asserts that the sparc compliance definition solves this problem. if this is true (i.e. scd is as good as posix) why buy unix v rel. 4 from sun? |- Sun has been aggressively unbundling software like the C compiler, |OS improvements, and such. Further, they *prohibit* you from buying |the unbundled Sun products and running them on a clone. this is unpleasant (i still have yet to carefully read a recent software license). since they've suggested they'll soon completely unbundle the compiler the question is: should cloners try to license the compiler or use gnu? [sun is using legalisms rather than technical superiority to get sales] |If you want to learn the novel forms of |competitive advantage that Sun is seeking, talk to Solborne or Auspex. solbourne negotiated a contract with sun. if they failed to do a good enough job one can hardly blame sun. (sun's current motto is: we enable compatibles, we don't support them.) i can't speak on the hardware issues but i'll check back in a year. however one could compare sun with dec. i could go on at length about this but i feel it more properly belongs in alt.sys.sun or the info-solbourne mailing list (sorry no info-tatung et.al.). -- pjg@acsu.buffalo.edu / rutgers!ub!pjg / pjg@ubvms (Bitnet) opinions found above are mine unless marked otherwise.
dlp@zule.EBay.Sun.COM (Dan Pritchett) (12/18/90)
In article <1990Dec16.222542.28236@erbe.se> prc@erbe.se (Robert Claeson) writes: > But Sun will be running SVr4 themselves. They are pretty much locked >into SVr4, and in fact, SunOS 5.0 will be SVr4. As Sun will tell you if you ask, their SVR4 o/s will be *based* on SVR4, but will also contain numerous proprietary extensions "to ensure our leadership". Also, the SPARC version of SVR4 that anyone can license from AT&T will not have the SunOS 4.x compatibility support built-in. Clone vendors will have to construct that part themselves, if they want to be able to run programs from SunOS. Few of our users buy our systems to run only the software available in the Sun Price List. Most buy additional software from ISV's that are available on SPARC. These ISV's will have no incentive to target their applications to the compatibility mode of SunOS/SVR4 when they can use the SPARC ABI (SVR4 + SCD) and hit all of the SPARC platforms including ours. The compatibility with SunOS 4.x is like the OpenWindows compatibility with SunView. It lets users get on the new platform immediately and bring along their applications from the prior release as until they are available on the new platform. This is of course my opinion and does not represent an official statement of policy from my employer. -- Dan Pritchett | ARPA/Internet: dlp@zule.EBay.Sun.COM Sun Federal System Engineer | UUCP: ...!sun!dlp -------------------------------------------------------------------------- I've read plenty of books, about heros and crooks. And I've learned much from both of their styles... --Jimmy Buffet
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (12/18/90)
In article <1990Dec16.222542.28236@erbe.se> prc@erbe.se (Robert Claeson) writes: | Also, the SPARC version of SVR4 that anyone can license from | AT&T will not have the SunOS 4.x compatibility support built-in. Clone | vendors will have to construct that part themselves, if they want to be | able to run programs from SunOS. I'm led to believe that's correct. However, while total binary compatibility may not be there for programs using SunOS extensions, I am told that Sun will use the SPARC ABI, and that applications written for generic SPARC will run on SunOS as well as V.4 from other vendors. I've been able to take a reasonable number of programs from a Sun to 386 V.4 in source, so the prudent vendor will probably pass on Sun extensions unless they provide a large performance bonus. -- 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.
root@lingua.cltr.uq.OZ.AU (Hulk Hogan) (12/19/90)
chuck@trantor.harris-atd.com (Chuck Musciano) writes: >In article <36911@cup.portal.com>, johnm@cup.portal.com (John - Madison) writes: >> Assertion: Sun is retreating from the idea of Open Systems. >> Consider: >> - Sun has been aggressively unbundling software like the C compiler, >> OS improvements, and such. Further, they *prohibit* you from buying >> the unbundled Sun products and running them on a clone. Not supporting >> such products is one thing, but a blanket prohibition is clearly aimed >> at the clone makers. > I hate to see the compiler completely unbundled, which is a mistake. >But again, what obligation does Sun have to a clone maker? How technical >support does IBM give Compaq? Also, can you cite the licensing agreement >which restricts Sun products to Sun hardware? Perhaps there is a market for >clones of unbundled Sun software! Our local Sun salesman said a couple of weeks ago that we would not be able to run the unbundled Sun products (C compiler etc etc) on our Solbourne. Period. To do some would violate the legal stuff. Even if we bought and located the binaries on a SPARCserver 2 and NFS mounted the binaries on the Solbourne. The only way would be if we suggested to our dealers [to get Solbourne] to license the software from Sun... I believe that Solbourne have a five year agreement with Sun which allows Solbourne to obtain SunOS's and be 100% compatible. Our Solbourne salesman indicated that the Solbourne version of SunOS (called OS/MP) would be approximately 90 days behind Suns. In recent times, this seems to have blown out to a *huge* delay, but probably has been compounded by the Symmetric Multiprocessing additions to SunOS that Solbourne are working on. I have heard whispers that this is because Sun has been delaying the handing over of the OS to Solbourne, but this is *only* rumour and if I am captured, I will disavow any knowledge of it. Regardless, the unbundling of the OS is effectively creating two SunOS's, a "devalued" SunOS for clone users, and the "real" SunOS for genuine Sun users. This is a *really* nice way for Sun to treat companies who license their OS. Really makes them want to license the unbundled tools I'd say. It was funny how at the SPARCstation 2 announcement, the speaker said how Sun was a single OS, single architecture (SPARC) company. Yeah, right. Funny how our 386i and 3/60 don't run the same binaries... /\ndy -- Andrew M. Jones, Systems Programmer, Internet: andy@lingua.cltr.uq.oz.au Centre for Lang. Teaching & Research, Phone (Australia): (07) 365 6915 University of Queensland, St. Lucia, Phone (World): +61 7 365 6915 Brisbane, Qld. AUSTRALIA 4072 Fax: +61 7 365 7077
croft@csusac.csus.edu (Steve Croft) (12/19/90)
In article <1990Dec19.024945.12335@lingua.cltr.uq.OZ.AU> root@lingua.cltr.uq.OZ.AU (Hulk Hogan) writes: >It was funny how at the SPARCstation 2 announcement, the speaker said >how Sun was a single OS, single architecture (SPARC) company. Yeah, >right. Funny how our 386i and 3/60 don't run the same binaries... Sun has discontinued the 386i and the Sun3 lines. The last day they accepted orders was 12/14/90. Steve stevec@water.ca.gov
prc@erbe.se (Robert Claeson) (12/29/90)
In article <1990Dec19.024945.12335@lingua.cltr.uq.OZ.AU> root@lingua.cltr.uq.OZ.AU (Hulk Hogan) writes: >It was funny how at the SPARCstation 2 announcement, the speaker said >how Sun was a single OS, single architecture (SPARC) company. Yeah, >right. Funny how our 386i and 3/60 don't run the same binaries... Heh. Remeber the good 'ole days when Sun told you that a company that did offer only one architecture just wanted to lock customers in? Funny how fast things change... -- Robert Claeson |Reasonable mailers: rclaeson@erbe.se ERBE DATA AB | Dumb mailers: rclaeson%erbe.se@sunet.se Jakobsberg, Sweden | Perverse mailers: rclaeson%erbe.se@encore.com Any opinions expressed herein definitely belongs to me and not to my employer.
don@zl2tnm.gp.co.nz (Don Stokes) (12/30/90)
prc@erbe.se (Robert Claeson) writes: > Heh. Remeber the good 'ole days when Sun told you that a company that > did offer only one architecture just wanted to lock customers in? Funny > how fast things change... Hmmm. Funny too how the company that offered one 32-bit architecture has been selling more 16 bitters, even introducing the odd new model, the whole time, and is now selling RISC boxes....... Don Stokes, ZL2TNM / / don@zl2tnm.gp.co.nz (home) Systems Programmer /GP/ GP PRINT LIMITED Wellington, don@gp.co.nz (work) __________________/ / ---------------- New_Zealand__________________________
gerard@tscs.uucp (Stephen M. Gerard) (01/05/91)
In article <1990Dec28.224236.5163@erbe.se> prc@erbe.se (Robert Claeson) writes: >In article <1990Dec19.024945.12335@lingua.cltr.uq.OZ.AU> root@lingua.cltr.uq.OZ.AU (Hulk Hogan) writes: > >>It was funny how at the SPARCstation 2 announcement, the speaker said >>how Sun was a single OS, single architecture (SPARC) company. Yeah, >>right. Funny how our 386i and 3/60 don't run the same binaries... > >Heh. Remeber the good 'ole days when Sun told you that a company that >did offer only one architecture just wanted to lock customers in? Funny >how fast things change... From what I have heard 90% of Sun's business is SPARC. I doubt that any of us could afford to buy any other architecture from them unless the government starts subsidizing manufacturer's to produce architectures that the market place does not want to buy. Sun tried to continue the 680X0 product line with the 3/80 and tried to enter the 80X86 product line with the 386i. They did their part, it was us consumers who failed to maintain Sun's multi-architecture support by not purchasing sufficient quantities of their newer non-SPARC machines for them to justify making them. Just my 2 cents... -Steve -- ------------------------------------------------------------------------------ Stephen Gerard - Total Support Computer Systems - Tampa - (813) 876-5990 UUCP: gerard@tscs FAX: (813) 871-2783 US-MAIL: Post Office Box 15395 - Tampa, Florida 33684-5395
johnm@cup.portal.com (John - Madison) (01/07/91)
>Sun tried to continue the 680X0 product line with the 3/80 and tried to >enter the 80X86 product line with the 386i. They did their part, it was >us consumers who failed to maintain Sun's multi-architecture support >by not purchasing sufficient quantities of their newer non-SPARC machines >for them to justify making them. The Sun 3/80 was a miserable machine. Among its fault were: - it was slower than the machine it replaced, the 3/60. - although it used the same package as the SPARCstation, it did not use the SBUs. Instead, it used 3/60 daughterboards, and you could only install one board per machine. This was a proprietary standard, and nobody but Sun used it. - it was *very* late. Sun was about the last company to start shipping 68030 products. The 386 shipped large numbers of machines. The real problem was that Sun could not produce a 486 based machine. The problems included: - an OS that had diverged from SunOS for the other platforms, so that integration of newly developed OS features was increasingly difficult. - an inability to actually get UNIX to run on the 486. This was blamed by Sun engineers on bugs in the 486 chip, but this seems unlikely. - an unannounced, but widely leaked, plan to stop porting new OS versions to the 486, because of these reasons. Also, don't forget the 3/360 (or whatever it was numbered), the vme 68030 system. It too was late and enormously expensive. The cpu board was loaded with expensive chips and never did work right. Sun could easily have saved the 68xxx family by producing a better desktop machine and a lower cost, more highly integrated vme machine. They didn't. It is tempting to claim that this was a preplanned strategy to migrate customers to SPARC, but the actual reason is a combination of engineering blunders and management stupidity. If Sun founder Andy Bechtolsheim hadn't spent $200,000 of his own money to develop the SPARCstation, Sun would be in big trouble right now. Sun's "all the wood behind one arrow" slogan is just a recognition that they burned all the other ones.
davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (01/07/91)
In article <37681@cup.portal.com> johnm@cup.portal.com (John - Madison) writes: | Sun's "all the wood behind one arrow" slogan is just a recognition that | they burned all the other ones. I've had that discussion with Sun people on a number of topics, most recently OpenLook vs. MOTIF. I told them that it didn't matter how much wood was behind the arrowhead if they missed the target. I assume that this is a matter of religious belief to some highly placed person at Sun, since they have read the market correctly on a lot of other things. -- 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.
henry@zoo.toronto.edu (Henry Spencer) (01/08/91)
In article <1991Jan5.135651.486@tscs.uucp> gerard@tscs.UUCP (Stephen M. Gerard) writes: >Sun tried to continue the 680X0 product line with the 3/80 and tried to >enter the 80X86 product line with the 386i. They did their part, it was >us consumers who failed to maintain Sun's multi-architecture support >by not purchasing sufficient quantities ... In other contexts this is known as "blaming the victim". -- If the Space Shuttle was the answer, | Henry Spencer at U of Toronto Zoology what was the question? | henry@zoo.toronto.edu utzoo!henry
fischer@iesd.auc.dk (Lars P. Fischer) (01/08/91)
>>>>> On 7 Jan 91, davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) said:
Bill> I've had that discussion with Sun people on a number of topics, most
Bill> recently OpenLook vs. MOTIF. I told them that it didn't matter how much
Bill> wood was behind the arrowhead if they missed the target. I assume that
Bill> this is a matter of religious belief to some highly placed person at
Bill> Sun, since they have read the market correctly on a lot of other things.
That's contrary to what Bill Joy says all the time -- at the UNIX Expo
in Stockholm last year he said that Sun would deliver whatever the
market wants. If OS/2 becomes a winner (god forbid :-), we'll have
OS/2 on our SPARCstation. He even looked sincere.
But then, they're going to play there cards as long as it's not clear
it's a loosing hand. The OPEN LOOK vs. Motif thing isn't over yet.
Note that a number of companies, besides AT&T and Sun, have recently
come out with new machines running OPEN LOOK as standard.
Followups directed to comp.windows.open-look.
/Lars
--
Lars Fischer, fischer@iesd.auc.dk | Q: How does a project get to be one
CS Dept., Univ. of Aalborg, DENMARK. | year late? A: One day at a time.johnm@cup.portal.com (John - Madison) (01/08/91)
> I've had that discussion with Sun people on a number of topics, most >recently OpenLook vs. MOTIF. I told them that it didn't matter how much >wood was behind the arrowhead if they missed the target. I assume that >this is a matter of religious belief to some highly placed person at >Sun, since they have read the market correctly on a lot of other things. Yep, it is a matter of religion with mcnealy, the president. He has publically stated that Sun will support Motif when "hell freezes over," or words to that effect. Open Look vs. Motif is no longer a rational issue for Sun management, if indeed it ever was.
pcg@aber-cs.UUCP (Piercarlo Grandi) (01/10/91)
In article <1991Jan5.135651.486@tscs.uucp> gerard@tscs.UUCP (Stephen M. Gerard) writes:
Sun tried to continue the 680X0 product line with the 3/80 and tried to
enter the 80X86 product line with the 386i. They did their part, it was
us consumers who failed to maintain Sun's multi-architecture support by
not purchasing sufficient quantities ...
Actually (rumour mongering here) the 386/486 SUNs were killed off by SUN
because they were technically competitive with the SPARC machines, and this
was embarassing ("they would not fit with corporate strategy"). It helped
that they were the effort of the East Coast SUN operation.
Of course one must also consider that between paying $1000 for a 486 CPU or
probably something like $100 for a SPARC chipset, and sell a workstation at
the same price, SUN would rather keep the $900 "value added" that Intel
wants for themselves...
That $900 "value added" that Intel does not want to split with their OEMs is
probably a good reason behind CompuAdd and others starting to sell SPARC
based PCs. $900 out of $5000 seems a hefty premium for a CPU.
--
Piercarlo Grandi | ARPA: pcg%uk.ac.aber.cs@nsfnet-relay.ac.uk
Dept of CS, UCW Aberystwyth | UUCP: ...!mcsun!ukc!aber-cs!pcg
Penglais, Aberystwyth SY23 3BZ, UK | INET: pcg@cs.aber.ac.ukdavidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (01/10/91)
In article <2199@aber-cs.UUCP> pcg@cs.aber.ac.uk (Piercarlo Grandi) writes: | Of course one must also consider that between paying $1000 for a 486 CPU or | probably something like $100 for a SPARC chipset, and sell a workstation at | the same price, SUN would rather keep the $900 "value added" that Intel | wants for themselves... Is that the correct price? $100 sounds cheaper than any price I've heard for a 486 equivalent chipset. Who sells the CPU, FPU, cache and controller for $100? -- 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.
guy@auspex.auspex.com (Guy Harris) (01/12/91)
>Actually (rumour mongering here) the 386/486 SUNs were killed off by SUN >because they were technically competitive with the SPARC machines, "Technically competitive" in what sense? The 25 MHz 386 machine, at least on the Stanford integer benchmarks, was about as fast as the Sun-3/2xx (25 MHz 68020 machine); it was faster on the Stanford floating-point benchmarks than the 3/2xx on which I tried the benchmarks, but the 3/2xx machine had only a 68881, not an FPA. (All compilation was done with "-O", which at the time invoked just the peephole optimizer on both machines; the global optimizer wasn't available on the 386, and it sped things up by about 20% on the 68020.) I forget how well the 4/280 on which I tried them did with "-O1", but I think that even without the global optimizer, it was at least competitive, and probably better. I think the 386i also had some performance problems with its disk controller. The 486 machine might have been more competitive.
jgk@osc.COM (Joe Keane) (01/12/91)
In article <37704@cup.portal.com> johnm@cup.portal.com (John - Madison) writes: >Yep, it is a matter of religion with mcnealy, the president. He has >publically stated that Sun will support Motif when "hell freezes over," >or words to that effect. Open Look vs. Motif is no longer a rational >issue for Sun management, if indeed it ever was. Sun is behaving like a spoiled brat in this case. Part of the problem is that they missed the boat and spent a lot of time on NeWS and SunWindows, which people didn't really want it. Maybe they should learn that just because you have more engineers doesn't make people want your product. You can argue the technical merits X vs. NeWS vs. SunView, but the main point is that Sun got shut multiple times. Because of all this, they have decided to be perverse and not give its customers what they want.
pcg@cs.aber.ac.uk (Piercarlo Grandi) (01/12/91)
On 10 Jan 91 13:58:51 GMT, davidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) said:
davidsen> In article <2199@aber-cs.UUCP> pcg@cs.aber.ac.uk (Piercarlo
davidsen> Grandi) writes:
davidsen> Of course one must also consider that between paying $1000 for
davidsen> a 486 CPU or probably something like $100 for a SPARC chipset,
davidsen> and sell a workstation at the same price, SUN would rather
davidsen> keep the $900 "value added" that Intel wants for themselves...
davidsen> Is that the correct price? $100 sounds cheaper than any price
davidsen> I've heard for a 486 equivalent chipset. Who sells the CPU,
davidsen> FPU, cache and controller for $100?
SUN to themselves (my estimate, as I wrote "probably something like",
but good chance that it is fairly accurate).
I am sure that Intel do not sell themselves the 486 for $1000 either.
$100 is another IMNHO good estimate there.
It all has to do with whom keeps the difference between "cost" and "what
the market will bear", doesn't it?
--
Piercarlo Grandi | ARPA: pcg%uk.ac.aber.cs@nsfnet-relay.ac.uk
Dept of CS, UCW Aberystwyth | UUCP: ...!mcsun!ukc!aber-cs!pcg
Penglais, Aberystwyth SY23 3BZ, UK | INET: pcg@cs.aber.ac.ukpcg@cs.aber.ac.uk (Piercarlo Grandi) (01/13/91)
On 11 Jan 91 18:16:29 GMT, guy@auspex.auspex.com (Guy Harris) said: pcg> Actually (rumour mongering here) the 386/486 SUNs were killed off by SUN pcg> because they were technically competitive with the SPARC machines, guy> "Technically competitive" in what sense? The 25 MHz 386 machine, at guy> least on the Stanford integer benchmarks, was about as fast as the guy> Sun-3/2xx (25 MHz 68020 machine); it was faster on the Stanford guy> floating-point benchmarks than the 3/2xx on which I tried the guy> benchmarks, but the 3/2xx machine had only a 68881, not an FPA. The 386 machines had (slightly) better price/perfomance than the SUN 3s, and they sold fairly well, being also DOS compatible, and with "user-friendly" software, ... The real problem was the 25Mhz 486 vs. the original SparcStation; the performance roughly equivalent, with some edge for the 486 for certain things. Even worse, if you wanted to upgrade to a faster machine, you could either junk your Sun3 and buy a SparcStation, or instead just upgrade your SUN 386i to a SUN 486i (this upgrade was a well kept secret :-> and I think is no longer available). OOPS! Major problem for Sun marketdroids. Eventually the Sun [34]84i were doomed, whatever their technical merits; they were not Sun mainstream, and they used a microprocessor whose "value added" went into Intel's, not Sun's, bottom line. Sun had no reason to line Intel's pocket instead of their own. -- Piercarlo Grandi | ARPA: pcg%uk.ac.aber.cs@nsfnet-relay.ac.uk Dept of CS, UCW Aberystwyth | UUCP: ...!mcsun!ukc!aber-cs!pcg Penglais, Aberystwyth SY23 3BZ, UK | INET: pcg@cs.aber.ac.uk
dwf@acl.lanl.gov (Dave Forslund) (01/13/91)
In article <4139@osc.COM> jgk@osc.COM (Joe Keane) writes:
Sun is behaving like a spoiled brat in this case. Part of the problem is that
they missed the boat and spent a lot of time on NeWS and SunWindows, which
people didn't really want it. Maybe they should learn that just because you
have more engineers doesn't make people want your product. You can argue the
technical merits X vs. NeWS vs. SunView, but the main point is that Sun got
shut multiple times. Because of all this, they have decided to be perverse
and not give its customers what they want.
Sun doesn't always give its customers what they want, but, from my
experience, they do a better job than most of the other vendors. We
have almost no one around here is asking for Motif, e.g., or for that
matter, OSF/1. SVR4 is favored by a lot of vendors and it comes with
Open Look, which is the most similar to SunView where the largest
chunk of the workstation market place is coming from. Their
decisions seem pretty logical to me.
--
David Forslund
Advanced Computing Laboratory
MS B287
Los Alamos National Laboratory
Los Alamos, NM 87545
Voice:(505) 665-1907
FAX: (505) 665-4939
EMAIL: dwf@lanl.govdavidsen@crdos1.crd.ge.COM (Wm E Davidsen Jr) (01/15/91)
In article <PCG.91Jan12161251@teachk.cs.aber.ac.uk> pcg@cs.aber.ac.uk (Piercarlo Grandi) writes: | Even worse, if you wanted to upgrade to a faster machine, you could | either junk your Sun3 and buy a SparcStation, or instead just upgrade | your SUN 386i to a SUN 486i (this upgrade was a well kept secret :-> and | I think is no longer available). OOPS! Major problem for Sun | marketdroids. Were any 486i's ever delivered? I started trying to get one in early December 1989, to buy with money which went away at the end of the year, and Sun couldn't find one. Later someone tried to order one (about May 1990) and couldn't get one for one reason or another. -- 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.
guy@auspex.auspex.com (Guy Harris) (01/15/91)
>...or instead just upgrade your SUN 386i to a SUN 486i (this upgrade was >a well kept secret :-> and I think is no longer available). I don't think it ever *was* available....
curt@mischief.ecn.purdue.edu (Curt Freeland) (01/15/91)
guy@auspex.auspex.com (Guy Harris) writes: >>...or instead just upgrade your SUN 386i to a SUN 486i (this upgrade was >>a well kept secret :-> and I think is no longer available). >I don't think it ever *was* available.... According to the Sun-3/Sun 386i Supplement to the US Price list Effective August 1, 1990 Page 42 (Sun386i upgrades) A Sun 386i to Sun486i upgrade is part number UG-RR-350 list price $5,995 discount category A Upgrade includes: CPU board, Sun 486i POP board, and cables. Customer returns Sun 386i CPU board A Sun 486i 0Meg memory expansion board is a UG-RR0MB listing at $750, discount category A Under the ordering instructions it says: Sun 386i to 486i upgrades require SunOS 4.0.3; order USR1-03F (1/4" tape) or USR1-12F (3.5-inch diskette) at no charge if media is required. The system board includes 8 SIMM sockets for 8Mbytes of existing memory. If the Sun386i system to be upgraded has 12 or 16 Mbytes of memory, order option UG-RR0MB, listed above. --curt -- Curt Freeland Manager, Systems Engineering Purdue University Engineering Computer Network (curt@mischief.ecn.purdue.edu) (317) 494-3715
johnm@cup.portal.com (John - Madison) (01/15/91)
>A Sun 386i to Sun486i upgrade is part number UG-RR-350 list price $5,995 >discount category A .... The 486i was listed in the Sun price list. However, it was never generally available and never quite worked. Sun publically claimed that bugs in the chip prevented it from running UNIX, that they couldn't get enough chips, and other reasons for not shipping. Internally, Sun decided to stop porting new versions of the os to the x86 machines. The 386i code was based on 4.0 beta, and was extensively modified, so incorporating the new features delivered in later versions of SunOS was very difficult. Meanwhile, unsold 386i's collected in the warehouse. Eventually, Sun killed off the machine when (1) they realized that the market for the machine was pretty small (2) the people who did buy it would be *very* upset when they learned that future versions of Sunos would not available for the 486i. It is interesting to note that the (ex) Sun employee with overall responsibility for the 386/486 was the same person who brought us the DEC Rainbow. :-)
cleary@husc8.harvard.edu (Kenneth Cleary) (01/16/91)
In article <4139@osc.COM> jgk@osc.COM (Joe Keane) writes: >Sun is behaving like a spoiled brat in this case. Part of the problem is that >they missed the boat and spent a lot of time on NeWS and SunWindows, which >people didn't really want it. Maybe they should learn that just because you >have more engineers doesn't make people want your product. You can argue the >technical merits X vs. NeWS vs. SunView, but the main point is that Sun got >shut multiple times. Because of all this, they have decided to be perverse >and not give its customers what they want. Excuse me, but... there are customers who see another side to OSF. I have no intention of defending SunView or NeWS, because I don't like them. However, if I want to develop code for OpenLook, and port it to whichever machine, Sun or AT&T won't stop me. They have included everything needed to do this in the X11 distribution (XView & olwm). (Yes, I know it is not public domain...) For people wanting to develop in Motif, they have to shell out the bucks ahead of time, before they get to really look at it, and from following the discussion in comp.windows.x.motif this has not been gratifying for some. The stated mission of OSF seems to be avoidance of monopolistic control over crucial technologies (i.e. fear of AT&T). However, if you look at the companies signing into OSF (IBM, DEC, etc.) you see the old masters of the proprietary technology game, who enjoyed being able to lock customers in. On the other hand, I see Sun constantly allowing others to use its technology without charge (XView, olwm, NFS, etc). Now the folks at OSF are trying to lock technology up, so that they can charge admission, and they have the audacity to cry foul at Sun for not wanting give control to the old robber-barons of proprietary technology. Could it be that the robber-barons are concerned that Sun gives too much away for free? Are the robber-barons (with the business mentality) unable to cope with organizations like Sun (with an engineering mindset), or AT&T (with a scientific mindset)? If we are talking giving the customers what they want, how about putting the source code for everything needed to build OSF/Motif out in the public? Why put so much effort into getting proprietary code out of Mach, just so you can lock it up again?
tjo@its.bt.co.uk (Tim Oldham) (01/17/91)
In article <5353@husc6.harvard.edu> cleary@husc8.UUCP (Kenneth Cleary) writes: >The stated mission of OSF seems to be avoidance of monopolistic control over >crucial technologies (i.e. fear of AT&T). However, if you look at the >companies signing into OSF (IBM, DEC, etc.) you see the old masters of the >proprietary technology game, who enjoyed being able to lock customers in. Yes. And while the current licensing costs are relatively low, there's no guarantee what they will be in years to come. 1) Make it cheap in initial costs. 2) Make it difficult and expensive to get out and easy to stay in. 3) Push the prices up gradually, continually making it more harder and more expensive to get out. 4) while true do 3) done Tim. -- Tim Oldham, BT Applied Systems. tjo@its.bt.co.uk or ...uunet!ukc!its!tjo US Tourist: ``Say, is this college pre-war?'' Porter: ``Madam, this college is Pre-America.''