Message-Id: <9302032103.AA22730@sc.zib-berlin.dbp.de>
Reply-To: Mailing list for the LaTeX3 project <LATEX-L@vm.urz.Uni-Heidelberg.de>
In-Reply-To:  <01GU9KPQD0K2FN50HM@MATH.AMS.ORG>
Date:         Wed, 3 Feb 93 22:04:12 CET
From: Michael Downes <MJD@MATH.AMS.ORG>
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: document classes & numbering systems
Status: R


> -... the number forms used in a document are, I
> -would argue, part of the logical structure of the document, rather
> -than part of the superficial presentation details.
>
> The key question is why is numbering style part of the logical
> structure?

Yes, I agree, that's the key question. Actually, I don't have much
more than a gut feeling about the numbering system being part of the
logical structure, and I was sort of hoping that others would be able
to furnish supporting examples from their own experiences. In average
documents changes in the numbering system are harmless. But here's a
couple of examples to indicate what I had in mind.

(a) Exercises in a textbook numbered like an enumerate environment. At
the beginning of each exercise section the instructions say "Answers
for odd-numbered exercises are found in Appendix~\ref{answ}." If a
different documentstyle is substituted that changes the default
numbering style for "list of exercises" from arabic numerals to
alphabetic, the adjective "odd-numbered" will no longer make sense.

(b) A mathematics article is full of a particular notation that uses
lots of asterisks. The numbering style for footnotes is arabic
numerals. If a different documentstyle changes the numbering style to
\fnsymbol, it might be a disservice to the reader because of confusion
between math asterisks and footnote asterisks.

All the vowel items in the above list may be regarded as hypothetical.

> Any hard-coded references to numbers are asking for
> trouble regardless of style changes

I certainly agree with that; but the mail from Michael Barr and Frank
Poppe indicates that there are practical problems which perhaps should
be taken into consideration for LaTeX3 development, to see if
something can be found to make users' life easier. Enforcing a
standard numbering system for standard elements of a given document
class (and thus requiring that numbering variations be explicitly
specified in the individual documents) might be one way, though my
intent was more to provoke discussion than to claim that that is the
best solution.

> -If so, then
> -documents of the same class are interchangeable only if:
>
> -(1) All documentstyles in a given class conform to a single numbering
> -system for all predefined environments, sectioning commands, and other
> -numbered elements.
>
> Absolutely untenable. You'll have a near-infinite number of book
> sub-classes, to say the least.

[Subclasses weren't what I was arguing for here, (1) was intended only
to be taken together with (2).]

> -(2) Any numbering changes are done in the individual documents.
>
> To be discouraged at best. Numbering is to be handled by the
> document style. The only excuse for changing it is when using a
> generic document style.

Why? You didn't explain the reasons.

> Hard references should be checked in any event because of
> problems with text being edited and numbers changing. See above.

Yes and no. At the point when an electronic document is turned over to
a publisher, the publisher usually prefers to minimize changes, thus
the chance of numbers changing is quite small. (Except for page refs,
but hard-coding of page refs is rarer than hard-coding of item refs,
and more obviously counterproductive.)

Michael Downes                              mjd@math.ams.org (Internet)