Message-Id: <199401281408.AA12561@mail.cs.tu-berlin.de>
Reply-To: Mailing list for the LaTeX3 project <LATEX-L@DHDURZ1.BITNET>
Date:         Fri, 28 Jan 1994 13:32:35 LCL
From: Mike Piff <M.Piff@sheffield.ac.uk>
Sender: Mailing list for the LaTeX3 project <LATEX-L@DHDURZ1.BITNET>
To: Multiple recipients of list LATEX-L <LATEX-L@DHDURZ1.BITNET>
Subject:      Re: form and content
Status: R

From:  David Carlisle <carlisle@computer-science.manchester.ac.uk>

%>For LaTeX the situation is slightly different. I think we would all
%>agree that font changes, explicit spacing and the like should always
%>be hidden behind logical markup, however Mike's statement of *only*
%>user-defined commands is too strong I think.
%>
%>LaTeX *does* provide a set of commands for logically marking up a
%>document: \section \begin{enumerate} etc.
%>By using these standard tags to denote this logical markup, your
%>document is much more easily portable.

The problem is, to change the appearance of a list, you are forced then to
hack a way at the meanings of standard LaTeX commands such as \enumerate or
\item, or even worse, non-standard ones like \@noparitem. I agree with Phil
here, the reasoning still applies to LaTeX. What LaTeX provides is a richer
array of constructs which can be mapped to ones own markup.

One problem with LaTeX is that it is not consistent in the way commands and
environments are used. Thus one says \begin{enumerate}, but not \begin{item},
and not \begin{section}. I see the environment as the main markup tool of
LaTeX, and feel that most declarative commands should disappear. If you say
\section, for instance, it is much more difficult to do something at the end
of the section, and indeed it is sometimes impossible from the markup
to tell when the logical section has ended. Whereas \end{section} tells you
*exactly* when it is ended and what to do there, eg, output horizontal line,
change headers and footers, output a blank page or even output a set of
solutions which have been written to another file during that section.
It would be interesting to know Lamport's design decisions in implementing
LaTeX this way, apart from saving extra layers of grouping.
[One often sees the practice too, of \begin{enumerate} being replaced by \be
and \end{enumerate} by \ee, and then \enumerate has been bound into an
environment which starts a grouping, and is then re-replaced by a single, and
less meaningful, command!]

%>                                        For instance latex2html
%>converters can usually recognise this markup and do the right thing,
%>but if you use \MyLevelTwoSectionStart, where this is defined in a
%>local style in terms of arbitrary TeX commands, then there is no
%>chance of anything parsing your document structure, except TeX.
%>
\be and \ee would destroy that capability anyway. But I thought that this was
the point about markup, that the whole complicated layout of a document could
be described by a few commands in the preamble, and changed by changing those
commands. This is one reason why WYSYWYG TeX systems are so difficult to come
by. Change line 2 then  move to page 312, and watch the delay as the whole
document is reformatted.

%>Where one draws the line between those commands that should and
%>should not be used in the document body is difficult.
%>
Agreed! Who would wish to have to type \begin{em}...\end{em} everywhere?

%>Most people on this list might agree that \bf should not occur in
%>document bodies. If we had modified latex2e so this was impossible, I
%>think that there would have been `a few' complaints raised from the vast
%>majority of LaTeX users.
%>

I am not one of the normative group who would suggest directives like that.


Mike Piff
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%% Dr M J Piff, School of Mathematics and Statistics, University of %%
%% Sheffield, UK.                    e-mail: M.Piff@sheffield.ac.uk %%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%