Message-Id: <9302081920.AA02289@sc.zib-berlin.dbp.de>
Reply-To: Mailing list for the LaTeX3 project <LATEX-L@vm.urz.Uni-Heidelberg.de>
Date:         Mon, 8 Feb 93 19:19:06 +0000
From: Paul Taylor <pt@DOC.IC.AC.UK>
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: task list, diagrams in particular
Status: R


George Greenwade:
> If I have unintentionally offended an adequate number of coordinators, ...
Really, George, I didn't mean that, as I indicated by  the :-) in my message.
It's just that you are an "Economics and Business Analysis" person and
talk one language, whilst I'm an academic researcher who dabbles in public
domain software and talk a different one.

Let me pose my question in a different way.

LaTeX (2.09 or 3) is a "document preparation system" and as such its main
business is accents, tabulation, baseline alignment and all the other things
which occupy this list most of the time. Clearly these need to be closely
co-ordinated, and depend on a "core" of typographical utilities.

On the other hand, I have provided a package for a specialised application.
There are other such applications (tables, chemistry, syntax diagrams, BNF)
and whilst using commutative diagrams as an example, I shall try to make
my comments in a generic way.

In my package, which I began to develop long before I'd heard of LaTeX 3
and is extensively used in the category theory and theoretical computer
science communities, it has been my policy not to rely on LaTeX or any
other environment. I have used plain TeX (as well as initex) commands,
but have allowed for the modifications made by LaTeX, AMS-TeX, and even
texinfo.  Of course I shall also make any adjustments appropriate to
LaTeX 3, whether or not I am considered a part of it.

Kris Rose also has a package (which is similar, and indeed he credits me
for some of the ideas) with its own user community. As I understand it,
he is also sympathetic to LaTeX 3 but not reliant on it. Different comments
probably apply to Mike Spivak.

So, none of us is waiting for direction from the LaTeX 3 management before
proceeding with development. Design and development, in different directions,
have been done, the packages are in use, and these will continue in the future.

Maybe convergence is possible, at least to some extent in design, though
not in code: my methods are completely different from what cursory
examination of their code suggests Rose's and Spivak's to be.
I am completely opposed to simply joining them together, just as I'm sure
no-one would seriously suggest converging C and Pascal by writing a
common compiler.  Adding to one features corresponding to the other is
possible and likely, whether done in co-operation or competition.
Emulation and translation may also be possible, and desirable in the event
that the maintainer of one of the packages decides to retire.

So, there is a general question: What should be the interface (both in terms of
organisation and in terms of code) between LaTeX 3 with its main tasks of
document preparation and specialist applications macros such as ours,
particularly in the case where two or more such packages already exist in an
advanced, well maintained and widely used form?

Paul Taylor