Message-Id: <9210080034.AA10038@sc.zib-berlin.dbp.de>
Reply-To: Mailing list for the LaTeX3 project <LATEX-L@vm.urz.Uni-Heidelberg.de>
Date:         Wed, 7 Oct 92 20:29:32 -0400
From: "Charles F. Wells" <cfw2@PO.CWRU.EDU>
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:      Syntax
Status: R

Michael Barr says:

> I can hardly believe that there are people out there who say that
> latex should not have letter.sty, for whatever reason given (and I
> heartily agree with Leslie Lamport about having to learn Word????
> just to do letters, when I use TeX for everything else---including
> my very own baseball scorecard).

I write intemperate letters to the editor using TeX, not to
mention intemperate memos to deans. (I made up my own memo.sty.)

The argument against letter.sty is that different countries have different
expectations.  However, that is NOT an argument against having a syntax for
letters, implemented in different ways for different uses.

This suggests that part of the LaTeX3 project might be create standard
syntax, perhaps guided by SGML, for describing the constituent parts of
books, of articles, of letters, of memos, and so on.

There have been discussions of syntax in particular cases (for example,
what to call level five and six subsections of articles) and vol-task
mentions syntax several times but not as a separate issue.  Perhaps it is
worth while to maintain a clear distinction between what we want for
syntax and how it should be implemented.  Something like this:

1. Make a list of types of documents that we want to provide syntax for
and provide it.

2. Make a sublist of the above that will be provided as a standard part of
LaTeX3, and another sublist of those styles that will vary from country to
country or discipline to discipline.

Of course, it's not that simple.   Syntax design decisions will change
with implementation experience.  You might want to provide a generic
American style and a generic European style for many things and let
country committees produce further variations.  And so on.  But I like the
idea of separating specification-syntax from design.

I'll bet those who are in the thick of this project are already thinking
this way.  But it has not been a clear thread in this discussion.

Charles Wells                         cfw2@po.cwru.edu
Department of Mathematics
Case Western Reserve University       Telephone 216-368-2893
University Circle                     Home phone 216-774-1926
Cleveland, OH 44106-7058, USA