MIME-version: 1.0
Content-type: TEXT/PLAIN; CHARSET=US-ASCII
Mail-system-version: <MultiNet-MM(369)+TOPSLIB(158)+PMDF(5.1)@MATH.AMS.ORG>
Message-ID:  <866813982.225310.BNB@MATH.AMS.ORG>
Reply-To: Mailing list for the LaTeX3 project
              <LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE>
In-Reply-To:  <199706201031.LAA19917@knott.elsevier.co.uk>
Date:         Fri, 20 Jun 1997 09:39:42 -0400
From: bbeeton <BNB@MATH.AMS.ORG>
Sender: Mailing list for the LaTeX3 project
              <LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE>
To: Multiple recipients of list LATEX-L
              <LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE>
Subject:      Re: discussing relation of LaTeX to TeX successors
Status: R

regarding tds for system organization, sebastian says
    if all the TeX systems worked in a similar way (I pray to my gods that
    Phil will not start saying VMS is The Only True Way To Organize
    Files), ...

i certainly won't claim it's the "one true way", but at ams we're
in the midst of building a tex production structure on vms that will
replace (in the foreseeable future) the unix system we're using now.
we're following the tds guidelines as best we know how, and don't
see any real problems.

while the exigencies of a convenient *distribution* structure are
obviously different from those of a production structure, at least
to the maintainer, we've taken to heart the many requests for a
more easily installed setup than we now have (with respect to the
ams-tex, ams-latex and amsfonts collections), and the world may
see some changes there as well.  while we're not in a position to
provide the installation scripts for "all" tex setups (or even for
the several different flavors of dos and windows installations),
we do find the tds to be predictable and explainable, and i expect
this to be reflected in our offerings "within my lifetime".  (this
is a personal opinion; circumstances and management may decree
otherwise, but something compatible with tds is what i'm aiming for.)

i encourage other package maintainers to consider how the tds can
best be supported at a distribution level (i've been impressed by
the organization of xypic in this respect), while retaining the
necessary unity required for reliable ctan mirroring and retrieval.
if ulrik vieth or joachim schrod is listening in, perhaps they can
take this as a request to reactivate the tds group to look at the
matter of archiving and distribution, which has some different
requirements than a user system.

so, sebastian, don't give up yet!  and don't knock vms without
looking a little more closely.
                                                -- bb