Path: utzoo!attcan!uunet!samsung!zaphod.mps.ohio-state.edu!wuarchive!mit-eddie!mit-amt!snorkelwacker!spdcc!merk!xylogics!world!madd
From: madd@world.std.com (jim frost)
Newsgroups: comp.unix.i386
Subject: Re: Taking risks on software (ISC)
Keywords: unix dos vpix
Message-ID: <1989Dec3.143323.1037@world.std.com>
Date: 3 Dec 89 14:33:23 GMT
References: <240@hdchq.UUCP> <1989Nov30.135631.12796@virtech.uucp> <[2575a8ad:160.11]comp.unix.i386;1@nstar.UUCP> <1989Dec2.041718.15310@pcrat.uucp>
Organization: Software Tool & Die
Lines: 33

rick@pcrat.uucp (Rick Richardson) writes:
>In article <[2575a8ad:160.11]comp.unix.i386;1@nstar.UUCP> akcs.larry@nstar.UUCP (Larry Snyder) writes:
>>How has the file system managed with complete un-attended power downs?  Try
>>pulling the plug a couple of times and see what happens.  I've lost the
>>power a couple of times to this machine (running SCO) and have NEVER lost a
>>file let alone the ability to boot the machine (which was the case SEVERAL
>>times when running ISC 2.02).

>This is completely different from my experiences.

Mine too.  I used to accidentally kick the plug out all the time and
it'd happily reboot and rebuild the free list.  Never lost a file.
This was with 2.0.2.

>Larry's experiences seem to be wildly different from those of
>most ISC users.

Different than mine, that's for sure.  Consider that the Xenix
networking support configured STREAMS incorrectly, causing the network
to fail sporadically (and hang the system; all kinds of strange things
happen when STREAMS runs out of NBLKs).  ISC worked out-of-the-box.
The Xenix compiler seemed very flaky when optimizing, the ISC compiler
worked well with only one exception (fns.c in GNU emacs).  There is no
such beast as X11R3 for Xenix, the ISC one worked extremely well.

I've used both, and ESIX too, and ISC provided better value than the
others, although I would really like better documentation (and job
control, but that's another story).

A satisfied developer,

jim frost
madd@std.com