Message-Id: <9311081817.AA08354@sc.ZIB-Berlin.DE>
Reply-To: Mailing list for the LaTeX3 project <LATEX-L@vm.urz.Uni-Heidelberg.de>
In-Reply-To:  <93Nov8.075256mst.95582@relay.Phys.UAlberta.Ca>; from "Jiri
              Zlatuska" at Nov 8, 93 7:37 am
Date:         Mon, 8 Nov 93 09:24:07 -0700
From: Michal Jaegermann <michal@GORTEL.PHYS.UALBERTA.CA>
Sender: Mailing list for the LaTeX3 project <LATEX-L@vm.urz.Uni-Heidelberg.de>
To: Multiple Recipients of <LATEX-L@vm.urz.Uni-Heidelberg.de>
Subject:      Re: the name NFSS
Status: R


I can sympathize with Michael Gossens (sp?) in his fight with VM/CMS
system and its ancient flat file structure.  It is not that entirely
flat, since it can be subdived to "disks".  Still better than MTS -
for example.  But if we will start to prefix anything with anything we
will soon have eight character prefixes (and we have accommodate
case-insensitive file names).  I think that such approach is a blind
alley and leads only to troubles described eloquently by Paul Taylor.

Ok, now NFSS2 is a kind of a "specialty solution", but LaTeX2e is supposed
to be THE LaTeX.  What a used does when trying to re-run an old document
saying 'helv' in options when 'nfhelv' is now expected?  S/he grabs an
ancient 'helv.sty' from wherever is deemed to be a good place and complaints
loudly that something is broken.  What a sysadmin does faced with such
prospect?  Clings for her/his dear life to an ancient system which
"always worked" and disregards "all these crazy TeX hackers".

As for VM/CMS problems.  Michael has at least this advantage over MS-DOS
that he has eight, not three, characters in a filename suffix to his
disposal and three part file names.  Maybe it is possible to use that
in an absence of slightly saner file system?

Rumours about TeX on Un*x suffering long startup times from a "deep" file
structure are obsolete for quite some time.  I know that first hand from
practice on various machines (even with a Un*x-like systems, like
Uglix and HP-UX).  But this is a side issue.

   Michal