To: 386users@TWG.COM Subject: 80386 mailing list, vol 4 #2 Date: 8 Feb 89 19:00:18 EST (Wed) From: "Wm E. Davidsen" 80386 User's mailing list vol 4 #2 Jan 9, 1989 In this issue: Ingres running under 386 UNIX(xenix?) Re: 80386 vs. 68030 Re: Xenix/386 2.3.1 on Monolithic motherboard Request for Info on LIM emulators for '386s Merge386 - Mixing DOS and UNIX Effect of cache on system speed Re: vol 3 #27, 386 & 386sx Upgrade 386 PC to UN*X Serial ports and Unix386 DOS emulators Re: Merge386 - Mixing DOS and UNIX Intel System 302 Inboard 386AT cable 386SX for PC/AT Amdek 386-based machines Re: 80386 mailing list, vol 3 #29 RE: cheap 386 box Re: C-Kermit vs SCO Xenix 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: overby@agrigene.UUCP (Scott Overby) Subject: Ingres running under 386 UNIX(xenix?) Date: 1 Dec 88 17:57:20 GMT Anyone have experience with Ingres RDMS running on a Compac 386 under a version of UNI Opinions on: - Performance - Portabilty - Memory requirements - Disk requirements - Bugs Thank you. ------------------------------ From: mike@stolaf.UUCP (Mike Haertel) Subject: Re: 80386 vs. 68030 Date: 5 Dec 88 01:48:18 GMT In article <5375@cbmvax.UUCP> daveh@cbmvax.UUCP (Dave Haynie) writes: >Going all out, I'd expect a 68020 to be slower than an 80386 system, a >68030 to be faster. I disagree. Working for the GNU project this summer I had an opportunity to use various 386 and '020 machines and, taking differences of clock speed, etc. into account I had the impression that the '020 machines were faster. I was using the same compiler (gcc) on both architectures. I believe the main reason the '020 won was simply that it had more registers available for temporary values for the optimizer. Eight registers (as in the 386) is clearly insufficient, and in fact 16 (as in the '020) is probably not enough. Maybe 32 would be right? #if hearsay_evidence_were_admissible_in_court I vaguely recall hearing of some PC type magazine (not Byte, which always says the 386 is faster, but maybe Dr. Dobbs?) running various benchmarks and, however reluctantly, concluding that the '020 was faster. Disclaimer: I personally have read no such article, someone just told me about it. I don't remember who either. #endif -- Mike Haertel mike@wheaties.ai.mit.edu In Hell they run VMS. ------------------------------ From: jbayer@ispi.UUCP (Jonathan Bayer) Subject: Re: Xenix/386 2.3.1 on Monolithic motherboard Date: 29 Nov 88 15:17:23 GMT In article <1988Nov28.184415.1424@ateng.ateng.com>, chip@ateng.ateng.com (Chip Salzenberg) writes: . We recently have had occasion to install Xenix/386 2.3.1 on a 386AT system . based on a Monolithic System Corporation motherboard. . . Or to try, anyway. . . When booting the N1 floppy, everything works fine until the kernel is . loaded. :-) Seriously: at the boot prompt the keyboard works, and during the . kernel load from floppy the Caps Lock and Num Lock lights toggle correctly. . But once the kernel load completes, Caps Lock and Num Lock -- along with the . rest of the keyboard -- do nothing at all. Try slowing down the computer. If it works then I think there is a problem with the keyboard io on the board. Jonathan Bayer Intelligent Software Products, Inc. ------------------------------ From: erc@pai.UUCP (Eric Johnson) Subject: Request for Info on LIM emulators for '386s Date: 1 Dec 88 14:34:01 GMT I have seen a number of discussions lately on the difference between extended and expanded memory on the PC. During those discussions, I saw mentions of LIM emulators, such as QEMM and LIMSIM. Now, the questions I have are... I have a Televideo '386 machine with 2 MB of RAM. I am running DOS (3.X) and I want to use the extra 1 MB of RAM as LIM 4.0 memory for software we are writing. Basically, we are using Intel Inboard 286s for some 286 machines and I would prefer to use the extra RAM I have already in the machine. 1) Are there any packages that would allow me to do this? Can you give me names and perhaps an address or phone number for a company contact? (If your company sells such a product here is a chance to justify your net access.) Are there any hidden gotchas in using such software? Hidden incompatibilities, etc.? 2) Will a LIM 4.0 emulator slow down the machine? Or, is the LIM software fast enough that performance is near that of a straight LIM board? I'm really unfamiliar with this whole topic of LIM et al, so any information will be helpful. 3) Will the Intel Inboard 286 work with a Televideo '386 machine? The 286 in the product name seems to imply no. I note that Intel also makes an Inboard _386_ which is a product that perhaps I will have to get. I would prefer to use the RAM already on board, but whatever works works... I can compile a summary for the net of any information that is emailed to me. Thank you for your help, -Eric -- Eric F. Johnson | Phone +1 612-894-0313 | Are we Prime Automation,Inc | UUCP: bungia!pai!erc | having 12201 Wood Lake Drive | UUCP: sun!tundra!pai!erc | fun Burnsville, MN 55337 USA | DOMAIN: erc@pai.mn.org | yet? ------------------------------ From: mrh@camcon.co.uk (Mark Hughes) Subject: Merge386 - Mixing DOS and UNIX Date: 1 Dec 88 09:40:17 GMT Do you have any experience with a product called Merge386 from Locus? It includes UNIX and MS-DOS for $795 and allows DOS to run as a UNIX process. I shan't bore everyone with all the details, as it is reviewed in the current issue of Byte, but it looks very nice, allows mixing of MS-DOS and UNIX commands and shared file space. I'd be interested to hear of any first hand experience, and in particular comments on its use of memory. The Byte article says that you need at least 2.5M, but does not say how useful it is at this level. The article goes on to describe benchmarks with up to four DOS processes (running concurrently under UNIX), but does not say how much RAM was needed to achieve this. -- ------------------- or <...!mcvax!ukc!idec!camcon!mrh> | Mark Hughes | Telex: 265871 (MONREF G) quoting: MAG70076 |(Compware . CCL) | BT Gold: 72:MAG70076 ------------------- Teleph: Cambridge (UK) (0)223-358855 ------------------------------ From: leech@threonine.cs.unc.edu Subject: Effect of cache on system speed Date: 3 Dec 88 01:33:13 GMT Any opinions or insights into which would be faster as a single-user Unix machine? The ubiquitous Norton SI seems useless for this type of comparison. - 20 MHz 80386, 100 ns memory, AMI write-through cache. - 25 MHz 80386, 70 ns memory, no cache (not sure if it's interleaved or page mode, though). Thanks! -- Jon Leech (leech@cs.unc.edu) __@/ ``My goal is simple. It is complete understanding of the universe, why it is as it is and why it exists at all.'' - Stephen Hawking ------------------------------ Date: Sat, 3 Dec 88 08:55:05 CST From: James Van Artsdalen Subject: Re: vol 3 #27, 386 & 386sx > Date: Tue, 8 Nov 1988 12:37 EST > From: The Time Traveler > Subject: 25Mhz? > I heard a rumor that the 25 Mhz '386 chips in such computers as the > PS/2 Model 70 are not really 25 Mhz. [...] I suppose it means that > they're certified to run at 20 Mhz only, but the manufacturers just > run them at 25 Mhz regardless. Not true. 25MHz parts are readily available, perhaps too much so if you believe Intel's financial reports. All of the other support chips also exist at the desired speed, so there's no problem with doing a design that's fully within the chip specifications. --- > From: rob@conexch.UUCP (Robert Collins) > Subject: Re: 386SX - what is it? > Date: 7 Nov 88 08:43:00 GMT > But please note, the 386SX is actually slower than the 286 running > at the same speed. This is a matter of some debate. I do not think it true at all when running unix or any other software taking advantage: unix on an SX will be faster and better than on a 286. Same with OS/2-386 whenever uSoft decides to do it. As for MS-DOS, I am not so sure. Extended memory access will be a good deal quicker with a 386SX, and that may make up any difference for people who use RAM disks or caches. Most humorous is the price difference between the 386SX and 386. The parts cannot be noticeably different in cost, yet there is a fair difference in price. The argument might even be made that the 386 is cheaper, since the R&D and NRE is long ago paid for... > Furthermore, Chips&Technologies, and Faraday both have 386-type chip > sets that are comparable to the 386SX. Remember that it is simple to do an SX daughterboard. It is therefore easy for a board designer to drop an SX onto a motherboard - design-ins are always easier than daughterboards. I don't know if anyone will do a 386SX-specific chipset. > I can't help but wonder if this is because Intel has (or is going to) > introduce a small circuit board w/ the 386SX on it that will slip > right in a 286 socket? Keep in mind that they can internally sell themselves 386SXs cheaper than they sell to outsiders and maintain a cost advantage. --- James R. Van Artsdalen james@bigtex.cactus.org "Live Free or Die" Home: 512-346-2444 Work: 338-8789 9505 Arboretum Blvd Austin TX 78759 Dell Computer Corporation ------------------------------ From: weyrich@csun1.UUCP (Orville Weyrich) Subject: Upgrade 386 PC to UN*X Date: 4 Dec 88 07:22:13 GMT I have a 16-MHz 80386-based AT-clone (Goldstar Technology GST-386) which is built around a Micronics motherboard and which has an Adaptec 2372 RLL controller, Seagate ST-4096 hard drive, 1.2 Meg floppy, 8087 coprocessor, two standard RS-232 ports, two standard parallel ports, a CGA video board, 2-Meg of main memory, and AWARD BIOS version C3.03. I am currently running MS-DOS 3.30 and am using TURBO-Pascal type applications. I only want a one- or two-user system. QUESTION # 1: Has anyone brought up some flavor of UN*X on any or all of the hardware described above? Any recommendations? [I realize that I may need more memory -- but how much?] QUESTION # 2: How heavily does UN*X utilize the PC BIOS? I had trouble getting my hard disk to function reliably using Phoenix BIOS, and am concerned about how UN*X will handle my disk/controller configuration. QUESTION # 3: I still need to use this box for MS-DOS applications. Eventually I would like to migrate my applications to UN*X, and so would like to be able to run TURBO Pascal and applications implemented in TURBO Pascal under UN*X with the ability to have the applications under MS-DOS communicate with processes running under UN*X. Some of my TP applications use LIM EMS. Any recommendations? QUESTION # 4: Alternatively, how hard is it to have both UN*X and MS-DOS to reside on the hard disk and choose one or the other when booting? QUESTION # 5: Finally, I am considering upgrading my CGA card to a Paradise VGA-Professional. Will this mess up UN*X? Any help is appreciated. Thanks, ORWJr. -- Orville R. Weyrich, Jr. | UUCP : ...gatech!ugacs!csun1!weyrich Department of Computer Science | University of Georgia | Athens, GA 30602 USA | MA BELL : (404) 542-1082 ------------------------------ From: ALEWIS%UTCVM.BITNET@cunyvm.cuny.edu (Adam Lewis) Subject: Serial ports and Unix386 DOS emulators Date: 3 Dec 88 17:18:02 GMT recently wrote about problems he was having using the serial port from inside Merge/386 under Microport SysV/386. Merge/386 is not alone in this behavior. The VP/ix based SunView DOS Windows system has similar problems on Sun's 386i boxes. This makes sense if you consider that DOS emulators depend upon Unix for doing things like serial communications. It would seem to me that the DOS way of doing things does not map well onto Unix. I'd be interested in hearing of anybody else using VP/ix has problems with data commuications. ----------------------------------------------------------------------- Adam Lewis Center for Computer Applications University of Tennessee, Chattanooga Chattanooga, TN 37403 ----------------------------------------------------------------------- ------------------------------ From: bill@cosi.UUCP (Bill Michaelson) Subject: Re: Merge386 - Mixing DOS and UNIX Date: 6 Dec 88 00:46:32 GMT In article <2171@titan.camcon.co.uk>, mrh@camcon.co.uk (Mark Hughes) writes: > > Do you have any experience with a product called Merge386 from Locus? > I'm running the Merge386 product under Microport UNIX. I think it's a good product generally, but I am experiencing problems with serial communications which you should be warned about. Simply put, don't expect to be able to run the usual terminal emulators or file transfer programs that you may be used to using in the MS-DOS environment. I don't think they'll work properly. Memory-wise, I don't know what to tell you. I've got 4MB, and I can run multiple MS-DOS sessions concurrently with no problems. I suspect this is a practical minimum for acceptable performance, but I have not done any exhaustive analyses, so don't take my word as gospel. -- Bill Michaelson, COS, Inc. \ "Better to keep your mouth shut, and let Voice 609-771-6705 / people think you're a fool, than to open rutgers!princeton!mccc!cosi!bill \ it..." CompuServe 72416,1026 / -Mark Twain ------------------------------ From: kmk@icc.UUCP (Krishna Komanduri) Subject: Intel System 302 Date: 6 Dec 88 21:07:05 GMT I am considering purchasing of Intel System 302 25Mhz '386 machine. Does anyone out there who have used and have any comments about its features, performance etc. with respect to IBM PS/2, Compaq 386's, etc. and also any features which are absent from others. Thanks in advance. Krishna Komanduri. ------------------------------ From: Vaprak@cup.portal.com (John C Foy) Subject: Inboard 386AT cable Date: 7 Dec 88 02:27:38 GMT I recently aquired an Intel Inboard 386AT card for my 286. It had the software part of the installation kit and all I am missing is the cable to get it installed. The retail for the whole kit is $200, but all I need is the cable. Does anyone know of where I could find just the cable for a reasonable price? Vaprak@cup.portal.com Sun!Cup.portal.com!Vaprak ------------------------------ From: gary@logico.UUCP (Gary Vrooman) Subject: 386SX for PC/AT Date: 7 Dec 88 17:25:46 GMT Cumulus Corp has announced a 386SX card for the PC/AT. It snaps into the motherboard rather than using up a slot. It costs around $600 but only provides roughly a 15% speedup over an AT. The card that we all really want for $600 would include an 82385 cache controller and 32K of static RAM. This should provide nearly double the speed of a PC/AT. Cumulus can be reached at 216-464-2211. ------------------------------ Date: Fri, 09 Dec 88 11:39:12 CST From: joel kahn Subject: Amdek 386-based machines Does anybody out there have hands-on experience with the Amdek 386? How good are they for general-purpose work? ------------------------------ Subject: Re: 80386 mailing list, vol 3 #29 Date: 11 Dec 88 15:54:55 EST (Sun) From: steinmetz!uunet!ontenv!soley (Norman S. Soley) > ------------------------------ > From: leech@tlab1.cs.unc.edu > Subject: How much memory is needed for Unix/Xenix? > Date: 10 Nov 88 23:00:23 GMT > > > I'm planning a 386 system purchase in the next few months. I would > like to run with 2 Meg, since it's my own $$$ involved. It seems that > Sys V/386 will not run in 2 Meg, but Xenix 386 will; could anyone with > actual experience comment on this? Also, how much impact will this > have on single-user operation - will a 2 Meg system be unusably slow? > Darn those DRAM prices anyway. Depends on what you want to do, I'm running both interactive 386ix and XENIX 386 in machines with only 2 megs, works fine for 2-3 users as long as you aren't trying to run an ethernet at the same time. -- Norman Soley - Data Communications Analyst - Ontario Ministry of the Environment UUCP: uunet!attcan!lsuc!ncrcan!ontenv!soley VOICE: +1 416 323 2623 OR: soley@ontenv.UUCP " Stay smart, go cool, be happy, it's the only way to get what you want" ------------------------------ Date: 13 Dec 88 21:34:00 EST Subject: RE: cheap 386 box Reply-To: "NJITX::HXN8477" >You might be interested to know a local firm here (MC Systems in >San Jose, 408-293-9228) is selling the Everex 386 (the house >brand model - 20MHz, with 64K cache, 1Mb memory, keyboard - >Landmarks at 24MHz) for $1695. >Dion L. Johnson >uucp...{decvax!microsof,ucbvax!ucscc,sun,uunet}!sco!dionj One thing to take into consideration is users' technical support. A year ago I bought Everex's EGA card, because I, too, liked its price. I had problems installing the card in my Leading Edge model D machine. As usual in such cases, I sought help from the dealer who pleaded ignorance and asked me to contact the manufacturer -namely, Everex. I picked up the phone and made a long-distance call to Everex. The operator put me on hold for a period of time so long that I had to hang up before anybody came to my help. Trying to save on my phone bill, I prefered to take a more backward approach -namely write a letter. I wrote 3 consecutive letters ( that's right, THREE) to Everex in Massachusetts. Everex never responded, let alone helped. When my patience ran out (I waited for a response for over 10 weeks), I decided to give them another LONG DISTANCE call, to salvage the $218 I paid for the card (At the time, this price was very attractive). Although I identified my self to the operator as a long distance caller, she put me on hold again for over 45 minutes :-). At the moment when I was about to get a heart attack, a technician (he must have been the only one they had :-)) at the other end came to my rescue and answered my questions to the effect that the card was useless for me. Since it was too late to return the card to the dealer, I had to throw it away. By contrast, whenever I needed help from Leading Edge, I received it almost immediately. Twice, the people there not only responded to my letters with each and every detail, but also included tens of photocopied pages to help me do the application I needed to do. That is why I am waiting for their 386 computer (model D3). They introduced it a couple of weeks ago but it has not reached the dealers yet. The introductory price is $3999 for 2 Mb of ram and a VGA card and monitor. However, I think it will come down to the $2500 level fairly rapidly, speciallay with more companies stepping into the 386 land. The company that has also satisfied its customers considerably, as far as technical support is concerned, is Hundai. I am surprised, though, that they have not introduced a 386 yet. My friends who have Huyndai computers tell me that they received help whenever they needed it, even long after they bought the computers. .;;;. .. /\ //^|^\\ .'..; / \ /\ | _/~~\_.'.' / \/ \ /\ | ,(_ ) "Blanked until further notice" / \ \/ \ | ' \~~~|' / \ \ \ | | | +---------------------------+------------------------------------------------+ |Hamed Nassar |Internet : hxn8477%njitx.decnet@njitc.njit.edu | |EE Department |UUCP : bellcore!argus!mars!nancy | |NJ Institute of Technology |CompuServe: 74000,130 | +---------------------------+------------------------------------------------+ ------------------------------ Date: Wed, 14 Dec 88 06:36 EST From: MAJ David McGuffey Subject: Re: C-Kermit vs SCO Xenix Date: 13 December 1988 15:06 est From: Frank da Cruz Subject: C-Kermit vs SCO Xenix We're catching up on backlogged mail around here, and saw your note about the problems you were having with C-Kermit. Is there any reason why you can't simple change the permission on /dev/tty1A to allow yourself read and write access? I finally tracked the problem down to using the tty1A device as dial-out and dial-in at the same time. Although the SCO manual indicates that this can be done, I've yet to find a complete solution. I did get a response that stated that SCO had a new getty out to fix the problem, another response indicated that I need to ensure that the major and minor device numbers are set correctly for dial-out/dial-in on the same line. I'll pursue both paths and post the results here and on 386users. In the mean time, I'm using cron to enable dial-in during those periods of time when I won't be using the dial-out capability. Once the port is disabled, getty not running, and no LCK..tty1a file present, both kermit and cu run a-ok. ------------------------------ End of 80386 M/L ****************