Path: utzoo!mnetor!uunet!husc6!cmcl2!brl-adm!adm!mwm@violet.Berkeley.EDU
From: mwm@violet.Berkeley.EDU (My watch has windows)
Newsgroups: comp.unix.wizards
Subject: Re: Jerry Pournelle on UNIX (from BYTE)
Message-ID: <11129@brl-adm.ARPA>
Date: 5 Jan 88 22:43:11 GMT
Sender: news@brl-adm.ARPA
Lines: 24

> OS/2 does things UNIX doesn't.

Any OS designed in the last five years that doesn't do things Unix
doesn't has problems. Unless it was designed to be a Unix-or-similar
clone, of course. But then it's probably got other problems.

>> But is unproven.  Unix has the benefit of over 15 years of learning
>> and research as well as debugging and enhancing.  It's been the
>> subject of intense scrutiny at the source code level.

That's 15 years of debugging and enhancing by a diverse set of people,
of varying abilities and styles. While the semantics of most of the
enhancements is clean, the same cannot be said for the code. And some
of the enhancements look like something slapped on the side of the
system.

A clean OS designed & written with the lessons learned since Unix was
first written has been needed for a while. But a Unix rewrite is a
non-trivial effort, and not liable to happen. So I look for some new
OS to displace Unix. The replacement will have to include replacements
for most/all of the Unix utilities, have the same basic semantics for
many operations, and be at least as portable. That isn't OS/2.

	<mike