To: 386users@TWG.COM Subject: 80386 mailing list, vol 4 #19 Date: 14 Mar 89 01:07:23 EST (Tue) From: "Wm E. Davidsen" 80386 User's mailing list vol 4 #19 Mar 14, 1989 In this issue: real time operating systems for 386 Inboard 386/PC and Heath/Zenith 151 Compatability Question Re: Gateway 2000 386 PC-MOS print spooler Memory management software [ 2 msgs ] Computer Products United 386 16MHz [ 3 msgs ] Re: Inboard 386 PC Re: Inboard 386 question Loss of characters above 9600 baud... 286 vs. 386sx [ 2 msgs ] XENIX 2.2.4 and VGA The addresses for the list are now: 386users@TWG.COM - for contributions to the list or ...!uunet!TWG.COM!386users 386users-request@TWG.COM - for administrivia or ...!uunet!TWG.COM!386users-request P L E A S E N O T E If you want to get on or off the list, or change your address, please mail to the 386users-request address, or the message will be delayed by having to hand forward it (for your convenience, not mine). ---------------------------------------------------------------------- From: Ken Seefried iii Subject: real time operating systems for 386 Date: Wed, 8 Mar 89 18:28:23 EST Kenneth Salisbury asked about real time OS's for his Dell 310 80386 systems. The products that I know about are the ones from VenturCom. They sell two things that might be of use. The first is Venix, a Unix System V.2 (SVID) for the 80286 with real time extentions. Should run just fine on a 386. The second is called 386/RT, and is advertised as a real time extention to ISC 386/ix Unix System V.3. Venix is a pretty good product. I don't know anything about 386/RT other than what the ad says. Seems like these products would be a good way to go. You get the real-time stuff along with the excellent software development facilities of Unix. You can reach VenturCom at +1 617 661 1230. Disclaimer: VenturCom and I have no buisiness conection... ...ken seefried iii ken@gatech.edu ------------------------------ From: Daniel Abramovitch Subject: Inboard 386/PC and Heath/Zenith 151 Compatability Question Date: Wed, 08 Mar 89 07:44:24 PST Does anybody have any experience with putting these two together? I have the latter and am considering aquiring the former. Thanks in advance, Danny Abramovitch ------------------------------ From: "Zachary T. Smith" Subject: Re: Gateway 2000 386 Date: Thu, 9 Mar 89 13:20:45 -0500 (EST) I've seen alot of people buying the Gateway 2000 386 with the Paradise VGA card in it. Is there anything especially great about that card?? Is it any better than the VEGA 16-bit card (not the Fastwrite) that they also ship (and which I am due to receive)? Hastily, Zach Smith (zs04+@andrew.cmu.edu) ------------------------------ From: uunet!afmpc-1.arpa!Rbaker Subject: PC-MOS print spooler Date: Thu, 9 Mar 89 8:20:56 CST I am having probs with PC-MOS '386 print spooler. When any user sends a job to the print queue, everything stops until the job finishes printing. I mean _everything_. Any clues? -Dumb user here. Please respond to rbaker@afmpc-1.arpa. I might not yet be added to the list. Thanks! ------------------------------ From: uunet!cunyvm.cuny.edu!RCSTLN@HEITUE5.BITNET Subject: Memory management software Date: Thu, 9 Mar 89 08:15 N Hoi, I've a 80386 based PC. I want to buy memorymanagement software. But I can not make a choice between QEMM-386, 386 TO THE MAX and a program from AMI. Is there anyone who has any worked with any of this software. My questions are: 1. How many (in amount of ram) of TSR's can I load behind the 640 K boundary? 2. Is the expanded memory simulator fine? (Does it work always) 3. With what software does it not work? 4. Do the pakages work with VM/386? 5. How is the service when you have problems? 6. Is there better software to do these funktions? 7. Which of the software is the best? PLease repley to >>RCSTLN@HEITHE5 >>LEO. V. NIEUWENBORG ------------------------------ From: davidsen@crdos1.UUCP (Wm E. Davidsen) Subject: Re: Memory management software Date: 13 Mar 89 10:04:03 EST (Mon) I have posted you letter in the issue to be mailed tonight. Here are a few thoughts on your question. I have used Desqview and QEMM quite a bit, so here are my opinions on it. > 1. How many (in amount of ram) of TSR's can I load behind the 640 > K boundary? If I understand your question, you can set aside one whole 640k machine to run TSR's if you wish. I suppose you could even have several, although I have gotten out of TSR's since I can open another window and run a real application. > 2. Is the expanded memory simulator fine? (Does it work always) I haven't had any trouble using QEMM. That does *not* means it works always, just that it works for well behaved applications. > 3. With what software does it not work? > 4. Do the pakages work with VM/386? > 5. How is the service when you have problems? Never had one. > 6. Is there better software to do these funktions? > 7. Which of the software is the best? ------------------------------ From: bash@ihlpb.ATT.COM (Basham) AT&T Bell Laboratories - Naperville, Illinois Subject: Computer Products United 386 16MHz Date: 7 Mar 89 22:52:21 GMT Not too long ago I posted on how Computer Products United salesmen treated me in a fair and friendly manner in my quest for a 16MHz 386 box. Well, after about three weeks, the machine locked up a couple of times while playing the game "Wasteland". I attributed this to the software. A few days later the machine died. Upon power up the fan comes on, the light comes on, and that is it. No screen, no errors, not even the speaker beep. I've shipped the box back to CPU. They told me I could cound on them having it at least a week. I tried to just get a replacement, but they refused until they've diagnosed the faulty unit. As of this writing U.P.S. has not delivered the computer to CPU. The thing that bothers me about this is: the processor is a 386 16MHz chip. The motherboard is copyrighted as a 20MHz board. There is an oscillator speed jumper that sets the maximum clock speed at either 16 or 20 MHz. This jumper is missing. Possibly the clock will default to a 16MHz rate, but I doubt it. I don't have a scope at home to actually measure the clock. If they have the clock running too fast the poor 386 would try to keep up as long as possible and then give up (I've seen this trick in the world of Z80s.). Anyway, to me, missing this jumper seems to be a blarring error. I'm getting worried that this may be an indication of poor practices by their staff. Then again. maybe it was a simple mistake. To be continued... Tom Basham -------------------------------------------------------------------------- "The world was going down the tubes. They needed a scapegoat. They found Wayne." ------------------------------ From: jsilva@cogsci.berkeley.edu.UUCP (John Silva) University of California, Berkeley Subject: Re: Computer Products United 386 16MHz Date: 9 Mar 89 19:28:16 GMT In article <9804@ihlpb.ATT.COM> bash@ihlpb.ATT.COM writes: >Not too long ago I posted on how Computer Products United salesmen >treated me in a fair and friendly manner in my quest for a 16MHz >386 box. > [stuff deleted] > >The thing that bothers me about this is: the processor is a 386 16MHz >chip. The motherboard is copyrighted as a 20MHz board. There is >an oscillator speed jumper that sets the maximum clock speed at >either 16 or 20 MHz. This jumper is missing. Possibly the clock >will default to a 16MHz rate, but I doubt it. I don't have a scope >at home to actually measure the clock. If they have the clock running >too fast the poor 386 would try to keep up as long as possible and >then give up (I've seen this trick in the world of Z80s.). Anyway, to >me, missing this jumper seems to be a blarring error. I'm getting >worried that this may be an indication of poor practices by their >staff. > >Then again. maybe it was a simple mistake. > >To be continued... > >Tom Basham I've found that a considerable number of companies who advertise '20Mhz' 386 systems are shipping them with 16Mhz processors. The reason for this is that they will have a reasonably low failure rate within warranty, while keeping costs down by using a cheaper processor. A 16Mhz 386 can be successfully used at 20Mhz, *as long as they are kept cool*. The reason for this is the chip will generate a certain amount of heat in normal operation (which the specs account for), and running them at a higher clock rate will generate excessive amounts of heat. This reduces the life of the processor, and the operating temperature range. Generally, if you buy a 20Mhz 386 board with a 20Mhz processor you'll be paying a little more for increased lifetime and reliability. Running a 386 out of spec isn't worth the hassle over the long run. -J. --- John P. Silva INTERNET : jsilva@cogsci.berkeley.edu "You don't know what you're UUCP : {backbone}!ucbvax!cogsci!jsilva getting into, friend..." ------------------------------ From: davidsen@crdos1.UUCP (bill davidsen) General Electric CRD, Schenectady, NY Subject: Re: Computer Products United 386 16MHz Date: 10 Mar 89 23:13:11 GMT In article <10866@pasteur.Berkeley.EDU> jsilva@cogsci.berkeley.edu.UUCP (John Silva) writes: | A 16Mhz 386 can be successfully used at 20Mhz, *as long as they are kept cool*. | The reason for this is the chip will generate a certain amount of heat in | normal operation (which the specs account for), and running them at a higher | clock rate will generate excessive amounts of heat. This reduces the life | of the processor, and the operating temperature range. | | Generally, if you buy a 20Mhz 386 board with a 20Mhz processor you'll be paying | a little more for increased lifetime and reliability. Running a 386 out of | spec isn't worth the hassle over the long run. You may have a good conclusion, but I think you got there using incorrect information... Everything I have read or seen indicates that Intel does not have assembly lines producing 386s in 20, and 16 MHz, but that there is one chip foundry which produces 386s using the same parts and masks. Chips labeled as 20 MHz are tested at that speed and are guaranteed for it. Chips labeled for lower speeds may be either (a) chips not tested at the higher speed, or (b) chips which failed the higher speed and passed the labeled speed. Intel says that they test under conditions which can not be duplicated casually, using extremes of rated voltage and temperature. Just trying it at home doesn't duplicate this. Therefore I agree that going with a tested chip for peace of mind is a good idea. The chip does not change by putting a label on top which says 16 MHz. There is no reason to suspect that a chip rated at any speed will run hotter than a chip rated a any other speed. I therefore disagree that running a slow chip at a high speed will "reduce the life of the processor" more than running a fast chip. They should run at the same temperature and have the same mean time to failure. This doesn't mean that the chip won't run hotter at higher speeds, not does it imply that some chips may not run the higher speeds if they get hot. Again you have the right idea, the chips may not be reliable if run fast, but not because they run hotter than "fast" chips, just that they won't run at the higher temperature. Depending on who you talk to, there is a 50-80% chance that any chip which works at all will work at 20 MHz. If I had a system which I could increase to 20MHz, I'd try it for any non-critical application. I think we've gotten to the same point from different directions. NOTE: I don't have information on the 25 MHz parts for sure, but since they are they same 'D' stepping using the same design rules you may conclude that it's worth trying. Keeping *any* CPU cool is a good idea, regardless of it's rating for speed or temperature. Remember "a cool electron is a happy electron." -- bill davidsen (wedu@ge-crd.arpa) {uunet | philabs}!steinmetz!crdos1!davidsen "Stupidity, like virtue, is its own reward" -me ------------------------------ From: abcscnge@csuna.csun.edu (Scott Neugroschl) CSU Northridge Subject: Re: Inboard 386 PC Date: 9 Mar 89 08:04:13 GMT In article <1803@uwovax.uwo.ca> 16012_3045@uwovax.uwo.ca (Paul Gomme) writes: ]In article <1070@blake.acs.washington.edu>, diehr@blake.acs.washington.edu (George Diehr) writes: ]> Yesterday I received a catalogue from J & R, and found that ]> Inboard 386 PC is selling at $599. Another 1 M daughter board ]> will cost $399 according to the ad. Anyway, I have been ]> interested in upgrading my old Zenith and hope to get ]> your opinions regrading my upgrade. Please help me. ]> Now the questions are: ] ]I've noticed messages from individuals who either _have_ or are ]_thinking_ of upgrading their machines by installing an Inboard 386 or ]similar product. I have questions of my own: Why do it this way in ]the first place? Why not sell the old machine and buy either a fast ]286, or a 386? It$ known a$ a $hortage of ca$h. The Inboard specified costs a total of $1K with 2M of memory. How much does a 2M 386 system cost? Even a bare bones: 2M memory, HDFD controller, 1.2M drive and nothing else (use the old XT stuff -- but then it's hard to sell the XT...) co$t? To many people it may be cheaper and more cost efficient to hold onto the XT and upgrade it as far as it will go. $incerely, ------------------------------ From: barry@dgbt.uucp (Barry Mclarnon) CRC, Ottawa CANADA Subject: Re: Inboard 386 question Date: 8 Mar 89 15:17:14 GMT >From article <6605@ecsvax.UUCP>, by mvolo@ecsvax.UUCP (Michael R. Volow): > > How does the speed of the IBM PC with the Inboard 386/PC compare > with a 80286 0 Wt State 8 Hz AT Clone? The Inboard runs at 16 MHz, but with 2 wait states (it uses 120 ns DRAMS). It still should have a significant edge over the AT clone in benchmarks that don't involve disk i/o though. > > Is the 1 meg of included memory enough for applications other than > [Dlarge spreadsheet/data base files, e.g. enough for applications such > as WordPerfect 5.0? Yes. Of the 384K of "nonconventional" memory on the board, 128K is used for ROM BIOS shadowing (optional) and internal uses, and is not otherwise available. The other 256K is available as extended memory, or it can be converted to 192K of expanded memory. The Inboard comes with custom versions of the Super PC-Kwik cache software (uses extended memory) and 386-to-the-max (for memory mapping). > > If you added memory, how much did you add? > What is the smallest amount of memory you can add with the daughter- > board? You can add either 1, 2, or 4 Megs. The first two options via the daughterboard which takes 256K chips and is available half- or fully- populated, and the second via the daughterboard which uses 1 Meg chips and is only available fully stuffed. Current prices are roughly $350, $700, and $1200 respectively (e.g., $349 for the 1 Meg from PC Connection in NH). > > M Volow, VA Medical Center, Durham, NC 27705 mvolo@ecsvax.UUCP Barry McLarnon -- Barry McLarnon Communications Research Center Ottawa, ON Canada UUCP: ...utzoo!bnr-vpa!bnr-rsc!dgbt!barry INTERNET: barry@dgbt.crc.dnd.ca Compu$erve: 71470,3651 Packet radio: VE3JF @ VE3JF ------------------------------ From: uunet!physics.rice.edu!zielke Subject: Loss of characters above 9600 baud... Date: 11 Mar 89 22:00:00 CST A possible cause is lost interupts. If the cpu does not get back to the com port before the next character comes in it will be lost. I have heard of com boards which buffer the ports but they tend to be somewhat expensive. David Zielke =============================================================================== ARPA==> Zielke@Physics.Rice.Edu * Zielke@128.42.9.23 * After Three Days MaBell==> 713-527-8101 ext. 4018 work * Without Programming 713-666-2982 home * Life Becomes US Snail==> David M. Zielke * Meaningless 7490 Brompton #110 * Houston, Tx 77025 * -The Tao of Programming =============================================================================== ------------------------------ From: "NJITX::HXN8477" Subject: 286 vs. 386sx Date: 12 Mar 89 09:31:00 EST In article <1215PICHER@MAINE> PICHER@MAINE.BITNET (Michael W. Picher) writes: >Could somebody possibly sum up the advantages of using a 386sx over >say a 20 MHz 286??? Can they 387 be used at high speeds with the >386sx?? This question could not have come at a better time. There is an article in this month's issue (March 89) of Byte magazine. There, you can find very illuminating points which clear up the misconceptions about 386. First, when 386 machines were first introduced, people thought that it ran programs much faster than 286 machines. However, that speed was not so much because of 386 power as it was because of the clock that was used. This was particularly evident when Harris and ADM introduced their high speed 286 chips. As a matter of fact, 286 machines outperformed 386 machines when they both ran at the same speed in some benchmark tests. It seems Intel did not itself introduce the fast 286 chips to make way for its new line of 386 chips. Second, part of the extra speed that 386 machines offer when they run DOS applications are due not so much to the CPU but to the caching techniques used in the system. Third, the vast majority of todays DOS applications are optimized for 8088 machines. There are very few applications that are written only for 286 machines, and much fewer for 386 machines. As a matter of fact most of the 286 machines today are running as virtual 8086 machines 90% of the time. Byte lists only about a dozen programs that are written only for 386 machines (i.e. the powerfull 386 instruction set is actually utilized). Fourth, when you buy a 386 machine, the data bus used is only 16 bits. So you have a CPU that is capable of handling 32 bits at a time but you are not able to move data to and fro memory 32 bits at a time. Fifth, a 386 machine costs on average 35-40% more than a 286 machine. Sixth, the 486 is due on the market sometime soon. As a matter of fact that is what has discouraged many software developers from writting software for the 386. It is also the reason why many potential buyers backed off waiting for the new chip which will deal the 386 a fatal blow. Seventh, OS/2 runs on 286 just as well as does on 386. Unix also runs on 286 machines, albeit with poorer performance than on 386 machines. But with uncertainty about future operating systems, one cannot really make any good plans at this points as to which machine will serve tommorows needs better. 1989 is supposed to be a very crucial year in defining the standards for the ninties. Byte predicts that OS/2 will not be as popular as DOS is today until 1991. I predict that OS/2 will never be popular and it is a big failure that IBM will regret for ever. But this is another topic anyway. >From the seven points above, it is obvious that if you are going to run *DOS* applications, you are better off with a 286 than with a 386 (that's unless you have some extra $$$ you want to get rid of :-) I would not pay for hardware that is not fully utilized by the software I am going to run. +---------------------------+------------------------------------------------+ |Hamed Nassar |Internet : hxn8477%njitx.decnet@njitc.njit.edu | |EE Department |UUCP : bellcore!argus!mars!nancy | |NJ Institute of Technology |CompuServe: 74000,130 | |Newark, NJ 07102 |Fidonet : 1:107/701 | +---------------------------+------------------------------------------------+ ------------------------------ From: davidsen@crdos1.UUCP (Wm E. Davidsen) Subject: Re: 286 vs. 386sx Date: 13 Mar 89 10:24:53 EST (Mon) This is a carefully constructed posting, with many good points. Unfortunately I have to disagree with some of them... > > [ question about 286 vs. 386sx ] > This question could not have come at a better time. There is an article > in this month's issue (March 89) of Byte magazine. There, you can find > very illuminating points which clear up the misconceptions about 386. > > First, when 386 machines were first introduced, people thought that it > ran programs much faster than 286 machines. However, that speed was not > so much because of 386 power as it was because of the clock that was > used. This was particularly evident when Harris and ADM introduced > their high speed 286 chips. As a matter of fact, 286 machines > outperformed 386 machines when they both ran at the same speed in > some benchmark tests. This is true, but by percentages so small that it's hardly fair to claim "faster." I generally disregard any comparison showing only a few percent difference. > Second, part of the extra speed that 386 machines offer when they run > DOS applications are due not so much to the CPU but to the caching > techniques used in the system. > > Third, the vast majority of todays DOS applications are optimized for > 8088 machines. There are very few applications that are written only > for 286 machines, and much fewer for 386 machines. As a matter of fact > most of the 286 machines today are running as virtual 8086 machines > 90% of the time. Byte lists only about a dozen programs that are written > only for 386 machines (i.e. the powerfull 386 instruction set is actually > utilized). However, among the applications which use the 386 are Desqview (my personal favorite) and Windows/386. These both allow a DOS user to have access to the power of the 386 in an immediately useful way. > Fourth, when you buy a 386 machine, the data bus used is only 16 bits. > So you have a CPU that is capable of handling 32 bits at a time but > you are not able to move data to and fro memory 32 bits at a time. This is generally not true. I know of no machines which don't offer 32 bit memory. Some may limit it to 4MB or so, but everyone I have seen offers 32 bit memory. The 386SX, obviously, has onlky a 16 bit bus. > Fifth, a 386 machine costs on average 35-40% more than a 286 machine. > > Sixth, the 486 is due on the market sometime soon. As a matter of > fact that is what has discouraged many software developers from > writting software for the 386. It is also the reason why many > potential buyers backed off waiting for the new chip which will deal > the 386 a fatal blow. Since the 486 will run 386 programs, and seems to be just a faster 386 in many respects, I don't think that has held back many software authors. At the projected prices of the 486, I don't think that it will kill off the 386 in the forseeable future. The 486 seems to need more support hardware, so that will help keep the 386 alive. Since the 8088 has refused to die five years after cheap 286's, I don't think this will happen. Note that people still make new Z80 products to run CP/M. > Seventh, OS/2 runs on 286 just as well as does on 386. Unix also runs > on 286 machines, albeit with poorer performance than on 386 > machines. That's rather incomplete. UNIX does run on a 286, but the constant need to program around the segmented architecture, and the poor performance on 32 bit integer data make it a hassle (this list runs on a 286, I know). > But with uncertainty about future operating systems, one > cannot really make any good plans at this points as to which machine > will serve tommorows needs better. 1989 is supposed to be a very > crucial year in defining the standards for the ninties. Byte predicts > that OS/2 will not be as popular as DOS is today until 1991. I predict > that OS/2 will never be popular and it is a big failure that IBM will > regret for ever. But this is another topic anyway. It is, and not one for this group unless someone has info on OS/3. > >From the seven points above, it is obvious that if you are going to > run *DOS* applications, you are better off with a 286 than with > a 386 (that's unless you have some extra $$$ you want to get rid of :-) If you don't want to run a multitasking program such as Desqview, and don't intend to buy software which uses the 386 instructions (PharLap or UNIX), he's completely right. > I would not pay for hardware that is not fully utilized by the software > I am going to run. I guess I question the "fully." If I can be more productive by using even a small part of the power, I like it. -- bill davidsen (wedu@ge-crd.ARPA -or- davidsen@crdos1.uucp) {uunet | philabs}!steinmetz!crdos1!davidsen "The world is filled with fools. They blindly follow their so-called 'reason' in the face of the church and common sense. Any fool can see that the world is flat!" - anon ------------------------------ From: uunet!sir-alan!mikes Subject: XENIX 2.2.4 and VGA Date: Sun Mar 12 15:11:25 1989 I finally went back to an EGA monitor and things are working fine. I actually like the EGA monitor better for text as the characters are more sharply defined. I discovered that the Tandy VTA adaptor uses the Paradise chipset so concluded that any of the VGA clone boards would probably have the same problem. When 2.3.1 (or is it XENIX GT? That's what the SCO info number called the SCSI release of XENIX, said to be scheduled for either late March or late September depending on which rumor you believe) I'll try again. Michael L. Squires uucp: {necntc,cwjcc,hoptoad}!ncoast!peng!sir-alan!mikes Department of Political Science ..!{convex,daisy,pitt,uunet}!sir-alan!mikes Allegheny College BITNET: mikes%sir-alan@pitt.UUCP (VAX) Meadville, PA 16335 MIKES AT SIR-ALAN!PITT.UUCP (IBM) Office: 814 332 3347 Internet: sir-alan!mikes@vax.cs.pittsburgh.edu sir-alan!mikes@uunet.uu.net Home: 814 337 5528 Data: 814 {337-3159,337-0348} login of "ubbs" for BBS ------------------------------ End of 80386 M/L ****************