Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site lanl.ARPA
Path: utzoo!watmath!clyde!burl!ulysses!allegra!bellcore!decvax!genrad!panda!talcott!harvard!seismo!cmcl2!lanl!jlg
From: jlg@lanl.ARPA
Newsgroups: net.arch
Subject: Re: Magic Cookies and File Systems
Message-ID: <23050@lanl.ARPA>
Date: Fri, 8-Mar-85 16:19:36 EST
Article-I.D.: lanl.23050
Posted: Fri Mar  8 16:19:36 1985
Date-Received: Mon, 11-Mar-85 05:37:23 EST
References: <917@sjuvax.UUCP> <538@rlgvax.UUCP> <190@u1100s.UUCP>
Sender: newsreader@lanl.ARPA
Organization: Los Alamos National Laboratory
Lines: 25

> [ does this really belong in net.arch? ]

Yes, it pertains to the archetecture of operating systems.  There is no
other network forum for this discussion and I don't know how to make one.

> It really burns me the way some message passing, locking,
> and semaphore systems use a globally administered pool of names (or
> worse yet, numbers!!) for the names of entities.  The file system directory
> structure is the ONLY logical place for global names of ANY kind to reside.

This assumes that the shared resources are global, that processes that
don't share resources are willing to put up with the overhead of having
this capability in the file interface, and that the designer agrees with
your religious conviction that the file system should be the only global
entity.  SOME information about shared resources needs to in the system
just as SOME I/O interface must be in the system.  But the place for rarely
used or very sophisticated features is at some level higher than the system
kernal.

> IF SOMETHING NEEDS A NAME, PUT IT IN THE FILE SYSTEM!!!!!!!!!!!!!!!

And therby slow the file system down for ALL users whether they use the
additional global naming capibility or not.

J. Giles