Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!caen!umich!sharkey!wyn386!wybbs!ohenry!doug
From: doug@ohenry.UUCP (Richard H. Douglas)
Newsgroups: comp.unix.sysv386
Subject: Re: ISC vs Adaptec 1542b vs 80387
Message-ID: <101@ohenry.UUCP>
Date: 4 May 91 17:16:39 GMT
References: <451@bartal.BARTAL.COM>
Organization: Douglas Bros., Inc.; Grand Rapids, MI 49503
Lines: 29

phillip@BARTAL.COM (Phillip M. Vogel) writes:
> Due to a change in system requirements at bartal, I decided to
> remove the 80387 from the unix box, and put it in the dos box
> where I run most of my CAD applications.  Well, The unix machine
> won't get past "loading the unix system" wit the 387 removed.
> However, if I also pull the SCSI controller, it will boot just
> fine from the st-4096.  I checked the cmos, and all is correct,
> as far as I can tell.
> 
> Is this for real? do you need to have a math coprocessor to run a
> SCSI disk? This may be the most ridiculous thing I've seen since
> they announced that you need to have a coprocessor to maintain
> system security...
> 
I too ran into this problem. Then I read the release notes. It seems
that whether you have a mathco installed or not has nothing to do
with whether the machine will boot or not. Per my SCO Unix notes,
the kernal must think that a matchco is present. Thus I believe
in /etc is a program called emulator or some such which is loaded
at boot time (If the mathco isn't present). I suspect that your
4096 has this program and you scsi drive does not. BTW, I have
a scsi setup (1542B Adaptec with a Seagate 670Meg drive) and it
works great. I also have the mathco installed not because I use it
so much, just because I like faster toys!!!

Hope that this helps.

Rich Douglas  doug@ohenry.        No sig yet. Lost in the bitbucket.
              Ma BELL (616)459-9157 --  Voice only yet.