Kemp@DOCKMASTER.NCSC.MIL (08/06/89)
Wolfgang Rupprecht writes: > There are much more "open" platforms than Suns. > > A 386 PC clone has a well defined *and* documented register layouts. > [ . . . ] > > Wouldn't it be ironic to spend a lot of effort to write a free kernel > for a proprietary hardware base? Yea, wasn't Berkeley foolish for spending so much effort on writing a kernel for a proprietary architecture like the VAX? :-) I don't have any proprietary knowledge about Sun's MMU, but in order to write code for my WhizBang-2000 Sparcstation clone, I don't have to. The whole reason Sun is licensing the Sparc architecture is to *encourage* clones, to provide a large enough market for shrink-wrap software vendors like Lotus to be interested. You can't have shrink-wrap software without an Application Binary Interface which is "well defined *and* documented". The kernal (GNU or otherwise) isolates the hardware from the application, and applications that muck with registers, screen memory, etc directly (like much PC software) are poor examples of software engineering. Sun itself has at least 4 Kernel Virtual Memory architectures - sun3 for the 68020, sun3x for the 68030, sun4 for the 4/100 & 200 series, and sun4c for the Sparcstation-1. The code that depends on the MMU is isolated in the kvm libraries. Presumably Solbourne, Prisma, Toshiba, the Taiwanese, Koreans, and other clone makers will have a similar arrangement. To claim that a particular MMU hardware design is a "standard" that software should know about is to constrain future development. I think an "opaque" interface as provided by the ABI is more in line with today's state of the art. Besides, I would much rather have a Sparcstation on my desk than a PC! Dave Kemp <Kemp@dockmaster.ncsc.mil>
RLN101%URIACC.BITNET@mitvma.mit.edu (Marshall Feldman) (08/07/89)
You've got to consider price. '386 systems are still much less expensive than SPARC stations. Also, while I agree many DOS programs use awful style, there's a huge and important software base out there. If GNU had a DOS-under-UNIX capability like Locus Merge or VP/ix, it would make the system much more useful for (1) persons who have or want to use low-cost DOS programs and (2) persons who need access to DOS (for file transfers, consulting with users) but aren't really interested in it for its own sake.
keith@fstohp.lynn.ge.com (Keith D Gregory) (08/07/89)
It seems that some thought should be given to history here. As I recall from
first reading "The GNU Manifesto", GNU is not to give any notice to such
"insignificant architectures" as the 8086 (as I recall, the original wording
was not so overtly inflammatory; one had to read between the lines). As a
result, it seems that most of the code is currently written for 68k or VAX.
So, now the .*86 line has significant power, has mostly shaken the constraints
of its segmented architecture, and is the most available and most inexpensive
platform in existance.
But is it going to be that easy to throw away history, such as GCC et al? If
GCC was written to produce optimal code for the VAX/68000 (dare I say PDP-11)
style architecture, will it transport well to the Intel architecture? Is a
project currently underway (or finished) to do so, or has anyone simply made
a [non-optimal] code generator using the current compiler?
********************************************************************************
I will agree, however, that a 386 (or 486) is probably the best platform, simply
due to the fact that it is such a widespread platform. Perhaps a Mac would also
be a contender, although it seems to me that one could go overboard trying to
usefully use its user-interface (multiple sessions would be nice, tho' ... it's
something that A/UX is missing, unless you want 3 pty's per session).
I think that SPARC, VAX, Sun-{2,3}, &c are not realistic. If you have a uVax
that isn't being used for anything else, more power to you. Same with a Sun.
Most people, however, aren't going to be able to convince their local MIS dept
to turn over that old 780 for use as a GNU system. Nor are most people going
to buy a SPARCStation, even at $9995, simply so that it can sit beside their
AT-386 or Mac-2, and run GNU (OK, to keep a religious war from starting, I will
simply say that _I_ wouldn't, although I might think of buying another drive
for my Mac-2).
-kdgguy@auspex.auspex.com (Guy Harris) (08/08/89)
>I don't have any proprietary knowledge about Sun's MMU, but in order to >write code for my WhizBang-2000 Sparcstation clone, I don't have to. Yes, but the intent is that GNU ultimately include a kernel, and to write VM code for your WhizBang-2000 SPARCStation clone, you *do* have to know how the WhizBang-2000's MMU works....
james@raid.dell.com (James Van Artsdalen) (08/08/89)
In <20528@adm.BRL.MIL>, keith@fstohp.lynn.ge.com (Keith D Gregory) wrote: > But is it going to be that easy to throw away history, such as GCC et al? If > GCC was written to produce optimal code for the VAX/68000 (dare I say PDP-11) > style architecture, will it transport well to the Intel architecture? Sigh. Look before you leap... gcc works fine on the 386. It is competitive with the commercial compilers I have examined. The only weak spot is in floating point support: the FP half of the *86 architecture is stack (as opposed to the conventional integer half). gcc works with 386 FP, it's just not fast. Fortunately this affects very few people. -- James R. Van Artsdalen james@raid.dell.com "Live Free or Die" DCC Corporation 9505 Arboretum Blvd Austin TX 78759 512-338-8789
jeffrey@algor2.uu.net (Jeffrey Kegler) (08/08/89)
In article <20519@adm.BRL.MIL> Kemp@DOCKMASTER.NCSC.MIL writes: >Wolfgang Rupprecht writes: > > There are much more "open" platforms than Suns. > > A 386 PC clone has a well defined *and* documented register layouts. > > Wouldn't it be ironic to spend a lot of effort to write a free kernel > > for a proprietary hardware base? > >Yea, wasn't Berkeley foolish for spending so much effort on writing a >kernel for a proprietary architecture like the VAX? :-) Yes, it would have been foolish if there were non-proprietary supermini's at that point. As it was, they choose the most widely cloned supermini architecture. >Besides, I would much rather have a Sparcstation on my desk than a PC! You obviously do not buy hardware with your own money. What is the real killer about proprietary workstation prices is adding-on, which is ridiculously priced. For example, in the ISA 386 workstation market, a file server is just another clone with a large fast disk and a cheap monitor. The price difference is close to a factor of ten. The basic unit is "under-priced"--that is, almost competitive with the ISA clones. Simply comparing the technology, the ISA is clearly inferior to just about everything else on the market. The 80386 has got to have one of the ugliest and least elegant instruction set ever used in a chip. "Workable" is about the nicest thing you can say. But given the relative pricing, workable is more than good enough. Even if RMS does not make the 80386 ISA one of the first platforms, someone soon will do it. Ten people will use this version of GNU for every one using the most popular proprietary architecture--since your vendor has you by the sensitive parts, why not use the UNIX he provides? You cannot get more dependent. Price is important because the better funded the industry segment, the less creative it is. If the opposite were true, all the excitement would be with COBOL and MVS. I want to detail all the humiliations and expenses vendors put you through once they lock you into a technology, but this is long enough so I will leave that to others. Disclaimer: I have no connection with any of the vendors or technologies described, except that I do have what for me is a substantial investment in ISA boxes. I do own the stupid little company below, so I guess that means I speak for it. -- Jeffrey Kegler, Independent UNIX Consultant, Algorists, Inc. jeffrey@algor2.ALGORISTS.COM or uunet!algor2!jeffrey 1762 Wainwright DR, Reston VA 22090
trost@crl.labs.tek.com (Bill Trost) (08/09/89)
Probably the wrong place for this (see the Followup-to), but hey, follow the bouncing thread.... (Wizards know about kill files anyhow, right?) I've seen lots of postings recently about GNU OS hardware and the like, and it seems there's a certain amount of misunderstanding and lack of information on the topic, so I'll upgrade it to my level of misunderstanding and lack of information. One individual suggests that 386 systems would provide an advantage because you could run Mess-DOS under Unix to gain access to "a huge and important software base." Somehow, I doubt that the _Free_ Software Foundation has any interest in supporting the use of software made by software hoarders like Microsoft and Lotus on its operating system. The foundation has the goal of creating a huge and important _free_ software base, and would like to encourage others to do the same. (I keep mumbling things to myself about writing GNUcalc, but....) And, support for the 386 is readily available in most GNU tools *now*. The 386 is now being used in real computers (like the Sequent Symmetry), and certainly cannot be ignored. Someone even bravely ventured to suggest that the Mac be used as a base for GNU. FSF is boycotting Apple and encourages you to do likewise. Their boycott centers around Apple's lawsuit against HP and Microsoft. I can send you details if you ask, but I *do not* wish to seed another flame war on this topic --- that's what gnu.misc.discuss is for :-). And, I could go on about Vaxes in schools and the like, but I think the most interesting point of this whole thing is that the front runner for the GNU OS, Mach, a message-passing kernel written at CMU, already runs on Suns and Vaxes and Sequents and Multimaxes and...well, lots of things, and it probably wouldn't be too hard to port. My guess is that the AT port could be done in, at worst, 4 months by people familiar with the hardware. Or maybe it's already been done -- venture into comp.os.mach for details on the OS. Bill Trost, Computer Research Labs, Tektronix trost@crl.labs.tek.com / tektronix!crl.labs!trost (trost@reed.bitnet, but probably tektronix!reed!trost)
RUSOFFMH@ctrvx1.vanderbilt.edu (T B A L) (08/11/89)
There are really several issues involved in the selection of a platform. Part of the point of writing a kernel in a high level language is that it really should not matter what platform you develop on. One of the things that everyone forgets is that in "cloning" UNIX, it is not necessary to duplicate its structure. Many operating systems have been produced that have a more rational structure. Among them is Mach. Another is Amoeba. Both are very portable, and both support virtual memory. While GNU OS has to "look" like UNIX, it does not have to "be" like UNIX. Perhaps, if RS could send out his basic ideas on what the kernel structure should be and we all could comment... Anyway, if you have to pick a platform, select systems that are based on chips with MMUs, select systems that have been around long enough for programmers to really understand them. Choose cheap systems that have inexpensive compilers. This essentially means 68030, 803/4/86 or 88000. One forward looking choice might be the 860. It holds the promise of inexpensive workstations and shares the 386 MMU (kind of). Probably, the 386 family offers the best immediate choice. It is, sadly, brain damaged, but then no one ever said anything was perfect... ------------------------------------------------------------------------ Martin H. Rusoff Vanderbilt University My opinions are not that of my employer or any other organization with which I may or may not be affiliated...
madd@bu-cs.BU.EDU (Jim Frost) (08/13/89)
In article <20519@adm.BRL.MIL> Kemp@DOCKMASTER.NCSC.MIL writes: |Sun itself has at least 4 Kernel Virtual Memory |architectures - sun3 for the 68020, sun3x for the 68030, sun4 for the |4/100 & 200 series, and sun4c for the Sparcstation-1. Let us not forget 68000, 68010, and 80386. While the 68k architectures probably aren't all that different (don't really know much about VM on them), the sparc and 80386 are quite different. |Besides, I would much rather have a Sparcstation on my desk than a PC! And I'd like a Cray. But some of us have to be realistic. As a sidebar, it's interesting to see what people call "PC"'s. The Sun 3/50, for instance, boasts 1.5mips performance. The Sun 3/60 is a bit more than twice that. They are "workstations". In contrast, the 25MHz 80386 is a bit more than 5 and the 33MHz is something like 7. The 80486 is rumored to be in the mid to high teens -- right about in the domain of RISC right now -- and these are "PC"'s. The line is not blurred, as many people suggest, but downright nonexistant. The only thing that differentiates a "PC" from a "workstation", so far as I can tell, is the display and the configurations that come from the dealer (eg PC's tend to come with 1Mb RAM, workstations with 4Mb or more and increasingly 8Mb or more). Oh, and the fact that PC's don't tend to need service contracts. If it gets the work done at a lower cost and higher reliability, it's a better tool. If it's a status symbol, however, rationality doesn't count. jim frost software tool & die madd@std.com
ari@kolmogorov.physics.uiuc.edu (08/15/89)
Sure, workstations and PC's can get blurred in terms of sheer CPU performance.
But there are a lot of things that separate the two for my personal
use (not based on marketing etc.)
1) Operating system supported by vendor on machine -
Certainly any IBM/PS-2 clone could support UNIX from a
third party. For my personal computer this is fine (it is cheaper)
However for my workstation, I would like the two to be
optimized for each other.
2) Multi-window environment supported by operating system -
I could get X to run on my personal computer with UNIX.
But again, I'd rather it was developed or at least optimized
by the vendor for my workstation. Also, high quality
large monochrome or color displays cost money, and
require good device to memory bandwidth.
3) High speed I/O and memory -
I'll settle for micro-bus XYZ for my personal computer
But I'd like a high speed VME bus, DMA etc. for my
workstation. I'd like >327MB, <18ms disks too.
4) Networking -
I'll need a modem and RS-232 line for my personal computer.
But I'd need a Ethernet connection, network support software,
etc. on my workstation.
Let's assume I start with "your average Joe" PC with a nice fast 386, and add
all the things it needs to be my workstation. Well, guess what?
It costs as much or more than a decent workstation from SUN etc.
And I'll also be the only one in the world with this monster.
I wouldn't disagree that companies jack the price up on some of these machines,
but when you need the entire package, there are no real cheap routes.
Work for a company that gets volume discounts!:-).
ari
Aritomo Shinozaki ari@kolmogorov.physics.uiuc.edu
Loomis Laboratory of Physics (217)-244-1744
University of Illinois, Urbana-Champaign
Urbana IL, 61801
jeffrey@algor2.uu.net (Jeffrey Kegler) (08/16/89)
Aritomo Shinozaki (ari@kolmogorov.physics.uiuc.edu) writes: > >Let's assume I start with "your average Joe" PC with a nice fast 386, >and add all the things it needs to be my workstation. Well, guess >what? It costs as much or more than a decent workstation from SUN >etc. And I'll also be the only one in the world with this monster. >I wouldn't disagree that companies jack the price up on some of these >machines, but when you need the entire package, there are no real >cheap routes. Work for a company that gets volume discounts!:-). I have a 386 PC with a Mono BLIT, IWS card, Intel S5R3.2, 380MB disk, 5MB memory (not enough, I know). By the time I got all this, yes, I could have had a Sun. So why didn't I get a Sun? 1) Since this is my own money, I would like some resale value for when I upgrade. Sun is going to come out with a fancy new box next year and then where will you be? Someone will buy my 386 to run Lotus 1-2-3 and this creates a strong resale market. There are weekly price quotes in Computerworld for 386 boxes, whereas last year's Sun machine can be hard to unload at any price. Anyone want last year's Sun box? 2) Service alternatives. You can get hardware for a 386 serviced all over the place. You know how much clout the owner of one machine has with Sun? 3) Prices. When I want to find the cost of an add-on, I check the ads in PC Magazine, etc. Sun's price list is secret. You wonder why? (I hereby challenge Sun to allow its current price list to be posted to the net. No unauthorized posting please. Let's argue facts here. Show me I'm wrong.) For example, I have a 60M cartridge drive. You cannot back up this size metal on floppies! The equivalent for your Sun will require an expansion chassis which alone costs more than the drive. The basic workstation setup is priced to look competitive with the PC equivalent, but move beyond it and you get the sort of stuff that gives monopoly pricing a bad name. 4) Futures. Sun controls your upgrade path if you buy from them. They want you to buy a brand new box. In the PC market hundreds of vendors are falling over each other to enable you to upgrade. => The real difference, at this point, is the distribution => channel. If the predominant distribution channel is through => retailers, it's a personal computer. If the predominant => distribution channel is through the manufacturer's sales reps, it's => a workstation. The technology is irrelevant. => John Nagle This above is the real definition of a workstation. All the other "definitions" are descriptions of what you'd like in a workstation, what you mainly see in thing called PC's as opposed to what you usually see in things called workstations--interesting, but not definitions. The workstation distribution channel is much less competitive and much more expensive than the PC one. Those overheads have to be paid for somehow, and noone should be fooled by the fact that one specific configuration is a few bucks cheaper. If you buy them one at a time, Sun really does not want your business. They cannot support the sales force on onesy-twosies. It is just a fact of life and not just a matter of attitude. Better to realize that before you buy than after. -- Jeffrey Kegler, Independent UNIX Consultant, Algorists, Inc. jeffrey@algor2.ALGORISTS.COM or uunet!algor2!jeffrey 1762 Wainwright DR, Reston VA 22090
wyle@inf.ethz.ch (Mitchel Wyle) (08/18/89)
jeffrey@algor2.UUCP (Jeffrey Kegler) writes: >Aritomo Shinozaki (ari@kolmogorov.physics.uiuc.edu) writes: >>Let's assume I start with "your average Joe" PC with a nice fast 386, >I have a 386 PC with a Mono BLIT, IWS card, Intel S5R3.2, 380MB disk, >5MB memory (not enough, I know). By the time I got all this, yes, I >could have had a Sun. So why didn't I get a Sun? >1) Since this is my own money, I would like some resale value for >when I upgrade. Sun is going to come out with a fancy new box next >year and then where will you be? Someone will buy my 386 to run Lotus Suns keep their resale value better than AT clones. Look at the prices in "The Sun Observer" or the other Sun rags. People who use Suns at work want one at home; since prices are dropping this lust results in a very brisk resale market. >2) Service alternatives. You can get hardware for a 386 serviced all >over the place. You know how much clout the owner of one machine has >with Sun? No one ever buys Sun service for his machine at home. There are plenty of 24-Hour board turn-around services and 3rd party Sun services. >3) Prices. When I want to find the cost of an add-on, I check the ads >in PC Magazine, etc. Sun's price list is secret. You wonder why? Again, no one in his right mind buys Sun add-ons. You buy 3rd party. You can admittedly get better deals on generic Intel boxes. Decisions about which machine to buy are religious; however, you have to admit that dbxtool is neato. On the other hand, you could place a value (how much you're willing to pay) to run the same software at home as you run at work, and how much a premium you're willing to pay for SunOS instead of 386ix or Xenix-386, or whatever. >Show me I'm wrong.) For example, I have a 60M cartridge drive. I always thought people who run Suns at home backup via SLIP on Exabyte 2.3 Gig 8 mm cartridges. :-) 60M is out; 2.3G is in. >The basic workstation setup is priced to look competitive with the PC >equivalent, but move beyond it and you get the sort of stuff that >gives monopoly pricing a bad name. This statement is probably true, but again you're ignoring 3rd parties in one situation (Sun) while expounding on them in the other situation (generic 386 box). >4) Futures. Sun controls your upgrade path if you buy from them. >They want you to buy a brand new box. In the PC market hundreds of >vendors are falling over each other to enable you to upgrade. Upgrading, add-ons, software, etc. are better in the PC market. You are absolutely correct. However (last time I'll mention it, really!) there are 3rd party Sun vendors. >=> The real difference, at this point, is the distribution >=> channel. If the predominant distribution channel is through >=> retailers, it's a personal computer. If the predominant >=> distribution channel is through the manufacturer's sales reps, it's >=> a workstation. The technology is irrelevant. Correct. >The workstation distribution channel is much less competitive and much >more expensive than the PC one. Those overheads have to be paid for >somehow, and noone should be fooled by the fact that one specific >configuration is a few bucks cheaper. Prepare yourself for a little shock. Workstations are going to have a price war. HP bought Apollo. Sun slashed prices of the 386i and may do so for the Sparcstation 1. NeXT is going to be on the ropes soon. Other (biggies) are coming. Let's not mentioni DEC, Iris, Titan, MIPS, or DG (oh well, I did :-). They are all going to be dropping prices when big (300 or more station) customers start buying low-bidder to run vanilla SysVr4 and X. >If you buy them one at a time, Sun really does not want your business. >They cannot support the sales force on onesy-twosies. It is just a >fact of life and not just a matter of attitude. Better to realize that >before you buy than after. I won't mention 3rd parties again (oops!) :-) Suns-at-home or onesies (great word, Jeff) are Sun religeous zealots who love their machines at work and want to be "true believers" at home, too. >Jeffrey Kegler, Independent UNIX Consultant, Algorists, Inc. Great name for a band: "The Algorithmics." -Mitchell F. Wyle Institut fuer Informationssysteme wyle@inf.ethz.ch ETH Zentrum / 8092 Zurich, Switzerland +41 1 256 5237
bzs@bu-cs.BU.EDU (Barry Shein) (08/20/89)
From: jeffrey@algor2.uu.net (Jeffrey Kegler) >1) Since this is my own money, I would like some resale value for >when I upgrade. Sun is going to come out with a fancy new box next >year and then where will you be? Someone will buy my 386 to run Lotus >1-2-3 and this creates a strong resale market. There are weekly price >quotes in Computerworld for 386 boxes, whereas last year's Sun machine >can be hard to unload at any price. Anyone want last year's Sun box? I don't think this is true, have you called the several used Sun vendors for a price quote on "last year's Sun"? Make sure you're really asking about last years and not a three year old model (like the 3/50.) I think you'd find it very easy to unload any one or two year old Sun with a reasonably std configuration, just like a PC. What's the resale price of a '286? I think that's a better comparison. I had someone come to me recently with a '286 asking advice on how to get rid of it, it was only two years old and probably the height of technology at the time. He had paid $10K for the system (he brought the receipt). My estimate was that he'd be lucky to get much more than $1500 today for it ('286, 1MB, Color EGA, 40MB disk, some other knick knacks.) If you're an early buyer of technology there's always a window of high resale value. But how much will you expect for the (I'm guessing) $7500 '386 system you bought when the '486 systems come out, the disk and memory costs halve etc? Probably $1K, and that's probably only a year or so away. Do you think anyone is going to want your neato-keen 80MB hard drive when everyone else is practically giving away 300MB drives (think back to when 20MB drives were only owned by the rich.) >2) Service alternatives. You can get hardware for a 386 serviced all >over the place. You know how much clout the owner of one machine has >with Sun? This remains an absolutely valid point. One way to get around this is to buy a late-model used machine from a reputable dealer who also offers service themselves. They exist, but it's still definitely a concern for someone desiring the latest and greatest. Until serious third-party dealers show up (some are in fact appearing around the SPARCstation already, look in the latest Unix World) this will make people hesitate, and rightfully so. >3) Prices. When I want to find the cost of an add-on, I check the ads >in PC Magazine, etc. Sun's price list is secret. You wonder why? (I >hereby challenge Sun to allow its current price list to be posted to >the net. No unauthorized posting please. Let's argue facts here. >Show me I'm wrong.) For example, I have a 60M cartridge drive. You >cannot back up this size metal on floppies! The equivalent for your >Sun will require an expansion chassis which alone costs more than the >drive. I think this is out of left field. Sun has a slick, glossy price list that they hand out as door prizes. I've never had any problem getting one (the new one also has all the maintenance prices.) I got the current one by walking into a Sun sales office (in Lexington, MA) and asking the receptionist who just went to the back and came out with one. I don't understand how the example relates to the gripe. There is a healthy third-party market for things like cartridge tape drives for Suns and they'd love to quote you prices if you just called them. When was the last time you considered buying a cartridge tape drive from IBM for your PC? Call one of the dozen or so third-party vendors for Sun. >The basic workstation setup is priced to look competitive with the PC >equivalent, but move beyond it and you get the sort of stuff that >gives monopoly pricing a bad name. WHAT monopoly pricing??? What are you talking about? Do you need a list of third-party vendors for Sun peripherals? Are you perhaps looking in the wrong places? Join the Sun User's Group (ahem) and get README, their magazine, and look thru the vendors advertised there. Go buy copies of The Sun Observer, Digital Review, Unix [World, Review, etc], Sun Technology, etc and read the ads. Get a Catalyst catalog from Sun for a few hundred more products (admittedly mostly software in the Catalyst catalog.) Go to a Sun User's Group conference and visit the trade show which will have almost 200 vendor's booths who will be trying desparately to sell you peripherals equivalent to anything Sun sells and at prices which are much cheaper. (Caveat Emptor, of course, but the same goes for flipping thru a mail order catalog for PC clone equipment.) >4) Futures. Sun controls your upgrade path if you buy from them. >They want you to buy a brand new box. In the PC market hundreds of >vendors are falling over each other to enable you to upgrade. Again, I don't understand. For example, do the Helios, Clearpoint and other upgrades to 3/50's increasing their memory by 2 or 3 times (something which was impossible from Sun) fit the bill here? How about all the floating point solutions for VME and other buses? How about Sun's offer to upgrade most 3/50's to the equivalent of a 3/60 by replacing the board (doubling performance and increasing memory capacity)? How about all the third-party memory/disk/tape vendors? How about the used vendors who will usually take your old system in trade for a newer one if it has any resale value at all (I'll grant they're not offering a whole lot for 2/120's these days, but how much can you get for a PC/XT?) >The workstation distribution channel is much less competitive and much >more expensive than the PC one. Those overheads have to be paid for >somehow, and noone should be fooled by the fact that one specific >configuration is a few bucks cheaper. I can basically agree with this, tho it's not clear that you can't walk out right now and get what you want. This is getting a little abstract. >If you buy them one at a time, Sun really does not want your business. >They cannot support the sales force on onesy-twosies. It is just a >fact of life and not just a matter of attitude. Better to realize that >before you buy than after. I believe this is a more important point in terms of after-market items like service and has truth to it, but I've outlined some ways around it. I own a Sun3/60 as a personal computer with a third-party disk and tape. I paid about the same as I would have for a Compaq or other high-end PC that might be in the same class on the basic issues. The value to me is that it's got an excellent Unix base and a lot of the things built in that others charge for (like ethernet.) But the software is what's important. It's all a matter of what you need, I wouldn't have bought it to just run spread-sheets, but you hardly need a '386 for that either. Somehow I suspect that a lot of what you wrote is just indicating that you've missed all the marketing for third-party Sun stuff somehow and assume that therefore it doesn't exist. Honestly, try some of the pubs I've mentioned and see if it doesn't change your mind 180 degrees. -- -Barry Shein Software Tool & Die, Purveyors to the Trade 1330 Beacon Street, Brookline, MA 02146, (617) 739-0202 Internet: bzs@skuld.std.com UUCP: encore!xylogics!skuld!bzs or uunet!skuld!bzs
datri@concave.uucp (Anthony A. Datri) (08/22/89)
>I don't understand how the example relates to the gripe. There is a >healthy third-party market for things like cartridge tape drives for >Suns and they'd love to quote you prices if you just called them. Certainly, and you can often beat even their prices by buying the parts and building your own. >>If you buy them one at a time, Sun really does not want your business. I would certainly have said this in the past, but the 3/80 (and maybe the SPARCstation, I haven't seen one) makes me wonder. You can stick a floppy and two hard drives inside, and it comes with a parallel printer port. That looks suspiciously like a machine that would fit into a smaller environment decently. Oh yeah -- one of the serial ports is 9 pin, like the IBM AT.