Path: utzoo!utgpu!water!watmath!clyde!rutgers!husc6!mailrus!umix!b-tech!zeeff
From: zeeff@b-tech.UUCP (Jon Zeeff)
Newsgroups: comp.unix.microport
Subject: Microport '386 Unix
Message-ID: <4280@b-tech.UUCP>
Date: 19 Feb 88 16:38:27 GMT
Organization: Branch Technology, Ann Arbor, MI
Lines: 70

> 
> > [ ... description of Club '386 machine ]
> > Note: this machine will not run '286 Unix (anyone know why?).  
> Two thoughts: is this the Everex 20MHz motherboard or the second 16MHz rev.
> (I don't think they produce the first rev. any more).

I'm not sure which rev it is.  It's the 3000A and runs at 16 mhz.

> I take it you have more faith in Bell's 386 compiler than uPorts?  I haven't
> tried out uPort's that much: did you have many troubles with it?

Actually, getting Bell's system was just cheaper.  As some point I believe
that uport will offer a faster compiler and it may then be worth the extra
cost.

> > The fuser command doesn't seem to work.
> 
> Ug.  This is a real problem.  I *need* it to prevent uucico from going wild
> and continuing PC Pursuit transfers into the morning hours...  I'll have to
> write a program or script to directly read the lock files I guess.

I did something similar - since the pid is stored in the lock file, you
can just use "kill `cat $lckfile`".

> > The lp driver seems unable to 
> > print very long strings without a newline now and then (although this 
> > might be caused by Merge).
> 
> This is a stupid idea, but could the extra newlines be related to the line
> width stuff in the lp driver?  I don't have my manuals in front of me, but did
> you set the line width to zero (or some special value) to disable newline
> generation?
> 

I was using /dev/lp directly.  Is there an ioctl that will fix this?  My
manuals don't seem to have anything on this.

> > HDB uucp insists on resetting the tty 
> > permissions and seems to use a bit more cpu time than it should.  
> 
> It's probably also a nightmarish security sieve, as is virtually every other
> uucp I've used.  I use a modified old HDB.

I finally "solved" this by making additional nodes (same name, but a
different directory).  It works fine now.

> Do they even include the drive # now?  At one time (286 2.2) they didn't
> include the drive number at all, and when I got a bad block, there was no
> way to tell which drive faulted.

I don't think the drive numbers are in the messages.  There do seem to be a
large number of non repeatable disk errors.  Does anyone else get frequent
disk errors? 

> I tried.  I now have a WD1005 on order for delivery Monday.  I also am in
> the process of borrowing the WD1007 from PCs Ltd to try it out, and maybe

I'd be very interested in any information on the WD1007, WD1007/WA2, and
the performance of the WD1006 (ie. is it worth it).

The OMTI-8620 sounds like the way to go if it would only work.

------------------------------------------------------------------
- Jon Zeeff			uunet!umix!b-tech!zeeff
  Branch Technology		zeeff%b-tech.uucp@umix.cc.umich.edu
  Ann Arbor, MI		
	
-- 
Jon Zeeff           		Branch Technology,
uunet!umix!b-tech!zeeff  	zeeff%b-tech.uucp@umix.cc.umich.edu