Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!zaphod.mps.ohio-state.edu!uakari.primate.wisc.edu!uflorida!mephisto!mcnc!decvax.dec.com!maxx!tyager From: tyager@maxx.UUCP (Tom Yager) Newsgroups: comp.unix.i386 Subject: Re: ISC 2.2 Installation Troubles Summary: Watch your step... Keywords: ISC, installation Message-ID: <47@maxx.UUCP> Date: 4 Jul 90 03:30:32 GMT References: <1990Jun22.052443.184@pilikia.pegasus.com> <26825@infotel.UUCP> <1990Jul1.125113.970@pilikia.pegasus.com> Organization: MAXX: Tom Yager's UNIX lab, Amherst, NH. Lines: 64 In article <1990Jul1.125113.970@pilikia.pegasus.com>, art@pilikia.pegasus.com (Art Neilson) writes: > Well, at the suggestion of several people I pulled my hard drive and > controller, went over to the local computer store and hooked up my > drive to a 386 Micronics 20Mhz motherboard. My ISC 2.2 installed > *just fine*. Problem fixed I thought, after all I had gotten net mail > from a guy in Australia who had done this same thing to bypass the > 2.2 install hang problem. I took my drive home with the new ISC > release installed on it, put it back in my system and fired the > system up. Lo and behold, the kernel panic'ed. Below you can > see the gory details firsthand: [... detail deleted to spare the squeamish ...] > This pretty much proves that the install disks aren't defective, what > we have here is a major incompatibility with the new release 2.2 and > my TransComputer 386 33Mhz motherboard. > -- > Arthur W. Neilson III | ARPA: art@pilikia.pegasus.com When ISC and SCO publish their compatibility lists, they ain't kidding! I just spent a full month trying to get a system configured to run UNIX. I had a 25MHz 386, 16MB of memory, a brand-X ESDI controller and a 300MB disk. The first system that came in passed every test in BYTE's arsenal with flying colors, but panicked the moment I tried to run UNIX. SCO wouldn't even install. ISC would, but died with the same symptoms described above. The long-story-short version is that I went through three replacement motherboards (each a different design and manufacturer), two memory expansion boards, and three complete sets of memory chips. I can't fault the vendor, because in the end I got a 33MHz system for the price of a 25--they felt horrible about all the problems. The third time must have been the charm, for the system's been running around the clock ever since. The moral: Never buy a system mail-order to run UNIX unless a) it's on the UNIX vendor's recommend list, or b) you got it on 30-day terms or have a return privilege. I think the best way to buy these days may be to get the whole sha-bang from someone like Dell, Everex or Mobius. You get the system loaded, it's warranteed for use with UNIX, and you don't get blank stares from tech support people who make you spell U-N-I-X. That's not an endorsement, but I know I'm far from the only person who's been bitten by the "incompatible compatible system" snake. I guess the manufacturers must be at fault, because there are more than enough systems out there that just work. I got fooled into thinking that all ISA-based 386 boxes had enough in common to run UNIX without problems. Baloney. By the way, the techies at this company said they burned the first two systems in using both Xenix and OS/2. I believe them, but neither OS found the problem. On the third one, I told them not to bother. (Hey, maybe it was running OS/2 that fried the other two boards! I must look into that.) Don't be too hard on ISC (and others) for failing to account for every flako timing or similar problem on every motherboard ever made. New Taiwanese and Korean companies are entering the market every day, it seems. No one could ever keep all those workarounds straight. As the man said on Hill Street: "Let's be careful out there." (ty) -- +--Tom Yager, Technical Editor, BYTE----Reviewer, UNIX World---------------+ | NET: decvax!maxx!tyager -or- uunet!bytepb!maxx!tyager | | I speak only for myself "UNIX: It's not a job, | +-------------------------------------it's a Jihad!" -co-worker------------+