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: Help us defend against VMS!
Message-ID: <12152@brl-adm.ARPA>
Date: 6 Mar 88 04:24:37 GMT
Sender: news@brl-adm.ARPA
Lines: 28

>> Hah.  The only thing the VMS manuals have over the man pages is bulk.

You forgot completeness and consistency. While they may be exceedingly
verbose, they are useable as reference manuals. The Unix man pages, on
the other hand, range from tutorials that are a pain in the ass to
find anything in, to brief scetches that don't do any more than remind
you of the flags.

I've as yet to run into a Unix system where the important part of the
documentation didn't live in the source tree.

Someone mentioned make. VMS has had "run" available for at least five
years. It does most of what make does (in other words, it handles the
important cases), and doesn't reuquire learning yet another language
to use. Run scripts look like (and are usuable as) DCL scripts.

On the other hand, there are lots of OS features - shared memory,
shared libraries, real IPC (as opposed to pipes), remote file systems,
and other interesting things for five years or more. These are things
that either don't exist, or have no standard, in Unix systems. At best
they have defacto standards.

I haven't seen VMS in five years. I assume that it's grown new
features since 3.x (or whatever it was I looked at). They may not be
clean, or neat, or as nice as they'll be when the eventually arrive in
Unix. But they're _there_. If you need them, Unix just won't cut it.

	<mike