X-VM-v5-Data: ([nil nil nil nil nil nil nil t nil]
	[nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil nil])
Date:         Fri, 25 May 90 17:11:00 GMT
Reply-To: LaTeX-L Mailing list <LATEX-L%DHDURZ1@unido.bitnet>
From: "Chris Rowley - Open University UK (R01/Maths)" <CA_ROWLEY@VAX.ACS.OPEN.AC.UK>
Subject:      RE: Attributes
To: Rainer Schoepf <BK4@DHDURZ1.uucp>
Status: R
X-Status: 
X-Keywords:                  
X-UID: 117

Rainer writes:
>  The scheme Leslie proposed looks very complicated
> to me, and I suspect that Mr. Average U. Ser  will not grasp the
> concept very well.  On the other hand, Frank's proposal is very
> easy to understand, even if it's a deviation from the current
> syntax.
I think that they both could be easy to use, or too difficult,
depending on what attributes (or parameter assignmemts are allowed)
and on how robustly they are implemented.  I do not think that the
syntax is such a major issue: and I can see advantages in both Frank's
and Leslie's.

> I want to add a few remarks concerning the question of a command
> versus an environment: In my eyes footnotes are a classical example
> of an entity that should be implemented as an environment.
Strong agreement, they should be, along with many other things such as the
box-generating commands, marginals, etc.
But can we make such radical changes at this stage?

> For things
> like sections my point a view is a bit different.  Frank's proposal
> was something like (if I understand it correctly):
>
>    \begin{section} <section title> \end{section}
>
> I think it is much better to use it like
>
>    \begin{section}{<section title>}
>      <Contents of the section>
>    \end{section}
>
> After all, \begin...\end enclose a logical entity.
Strong disagreement: the point is that a section is very much like an
item in a list: it can only be ended by the start of one of a
well-defined (by the document style) collection of other entities,
namely those whose level in the sectioning hierarchy is at least as
high as the level of section.  Thus, unless we wish to go to the
extreme of insisting that everything has a \begin and \end, then
\section is a natural candidate for "ommitted end-tag" status.

Thus I do not like   \begin{section} <section title> \end{section}
either (did Frank really suggest this?).

At the logical level, <section title> is merely a logical attribute of
the section thus, using (without prejudice) Leslie's notation, the
syntax should be something like this:

\section[\renewcommand{\sectitle}{<section title>}...]

NOTE: I am NOT advoctaing this, just pointing out that

\section[...]{<section title>}

is logically speaking a short form for the above.

Which reminds me:  whilst I agree with Leslie that a usable system
must blur the line between logical entities and formatting, and provide
convenient short forms and defaults for commonly used forms, the
implementors of the system need to understand and dcocument
exactly when and why this is done.


chris