path: root/info/ltx3pub/l3d002b.tex

%%% ====================================================================
%%%  @LaTeX3-article{ LaTeX3-LTX3-002b,
%%%  filename        = "l3d002b.tex",
%%%  archived        = "ctan:/tex-archive/info/ltx3pub/",
%%%  related-files   = "part of l3d002.tex",
%%%  author          = "David Rhead",
%%%  doc-group       = "Project core team",
%%%  title           = "Some ideas for improving {\LaTeX}\\ General",
%%%  version         = "1.1",
%%%  date            = "18-Mar-1993",
%%%  time            = "20:19:36 GMT",
%%%  status          = "public, official",
%%%  abstract        = "Ideas and suggestions from David Rhead for
%%%                     improving various areas in LaTeX",
%%%  note            = "prepared for the workshop at Dedham 91",
%%%  keywords        = "",
%%%  project-address = "LaTeX3 Project            \\
%%%                     c/o Dr. Chris Rowley      \\
%%%                     The Open University       \\
%%%                     Parsifal College          \\
%%%                     Finchley Road             \\
%%%                     London NW3 7BG, England, UK",
%%%  project-tel     = "+44 171 794 0575",
%%%  project-FAX     = "+44 171 433 6196",
%%%  project-email   = "LTX3-Mgr@SHSU.edu",
%%%  copyright       = "Copyright (C) 1993 LaTeX3 Project
%%%                     All rights reserved.
%%%
%%%                     Permission is granted to make and distribute
%%%                     verbatim copies of this publication or of
%%%                     coherent parts from this publication provided
%%%                     this copyright notice and this permission
%%%                     notice are preserved on all copies.
%%%
%%%                     Permission is granted to copy and distribute
%%%                     translations of this publication or of
%%%                     individual items from this publication into
%%%                     another language provided that the translation
%%%                     is approved by the original copyright holders.
%%%
%%%                     No other permissions to copy or distribute this
%%%                     publication in any form are granted and in
%%%                     particular no permission to copy parts of it
%%%                     in such a way as to materially change its
%%%                     meaning.",
%%%  generalinfo     = "To subscribe to the LaTeX3 discussion list:
%%%
%%%                      Send mail to listserv@vm.urz.uni-heidelberg.de
%%%                      with the following line as the body of the
%%%                      message (substituting your own name):
%%%
%%%                        subscribe LaTeX-L First-name Surname
%%%
%%%                     To find out about volunteer work:
%%%
%%%                      look at the document vol-task.tex which can
%%%                      be obtained electronically, see below.
%%%
%%%                     To retrieve project publications electronically:
%%%
%%%                      Project publications are available for
%%%                      retrieval by anonymous ftp from ctan hosts:
%%%                          ftp.tex.ac.uk
%%%                          ftp.dante.de
%%%                          ftp.shsu.edu
%%%                      in the directory /tex-archive/info/ltx3pub.
%%%
%%%                      The file ltx3pub.bib in that directory gives
%%%                      full bibliographical information including
%%%                      abstracts in BibTeX format.  A brief history
%%%                      of the project and a description of its aims
%%%                      is contained in l3d001.tex.
%%%
%%%                     If you only have access to email, and not ftp
%%%                      You may use the ftpmail service.
%%%                      Send a message just containg the word
%%%                          help
%%%                      to ftpmail@ftp.shsu.edu
%%%                      for more information about this service.
%%%
%%%                     For offers of financial contributions or
%%%                      contributions of computing equipment or
%%%                      software, contact the project at the above
%%%                      address, or the TeX Users Group.
%%%
%%%                     For offers of technical assistance, contact the
%%%                      project at the above address.
%%%
%%%                     For technical enquiries and suggestions, send
%%%                      e-mail to the latex-l list or contact the
%%%                      project at the above address.",
%%%  checksum        = "20144 934 5970 41620",
%%%  docstring       = "The checksum field above contains a CRC-16
%%%                     checksum as the first value, followed by the
%%%                     equivalent of the standard UNIX wc (word
%%%                     count) utility output of lines, words, and
%%%                     characters.  This is produced by Robert
%%%                     Solovay's checksum utility.",
%%%  }
%%% ====================================================================



\chapter{Some e-mail comments about structure}
\label{structure}

\begin{footnotesize}\begin{verbatim}

From:     MITTELBACH FRANK <PZF5HZ@EARN.DRUEDS2>
Date:     Fri, 5 Oct 90 17:43:46 CET

I would like to foward a mail from Nico which was a reaction
to some discussions at the Cork meeting.

------------------------ forwarded mail -------------
...

> Front matter
> ------------
> Currently, an article starts with \title, \author and \date
> instructions. In the Elsevier styles I have added \address,
> \received, \revised and \accepted commands, and also a keyword
> environment, similar to the abstract environment. By doing so, we can
> automatically convert a LaTeX-coded document to an SGML-coded
> document.
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     Don Hosek <DHOSEK@EDU.CLAREMONT.HMCVAX>
Date:     Mon, 18 Feb 91 16:36:00 PST

A few items which have been brought up previously, but I'm
willing to bring them up again now that I have some direct
experience with the situation in a production situation:

\chapter*, to make sense, should work as follows:
- A TOC entry should be generated
- Headers should be appropriately modified
+ All sections contained in it should be unnumbered. Not sure
  about "numbered-within" items like equations and figures. The
  situation has yet to come up

Other sectioning commands should work similarly

Items like \tableofcontents and \thebibliography which currently
use \something* to generate heads can use the low-level commands
to get at the header alone. Things would make more sense in the
style files this way as well.

My use of \chapter* is to mark the prefatory sections of a
document. The first non-starred chapter command switches to
arabic numerals for pages.

...
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     Sebastian Rahtz <S.P.Q.Rahtz@UK.AC.SOTON.ECS>
Date:     Tue, 19 Feb 91 13:54:39 gmt

 > \chapter*, to make sense, should work as follows:
 > - A TOC entry should be generated
no, not necessarily. that should be clearly up to the style designer.
I dont want Preface and Acknowledgments in my TOC, thanks, in style B

 > + All sections contained in it should be unnumbered. Not sure
 >   about "numbered-within" items like equations and figures. The
 >   situation has yet to come up
again, this is not a universal rule. it might make sense, though, to
copy the (S)GML world, and have a higher level constrcut like
\begin{prefacesection} and \end{prefacesection} in which, by default,
sections are all * type

isn't Don's requirement catered for by the values of tocdepth and
secnumdepth, used judiciously? one can reset them on the fly, yes?
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     Don Hosek <DHOSEK@EDU.CLAREMONT.HMCVAX>
Date:     Wed, 20 Feb 91 19:14:00 PST

> > \chapter*, to make sense, should work as follows:
>4 > - A TOC entry should be generated
>no, not necessarily. that should be clearly up to the style designer.
>I dont want Preface and Acknowledgments in my TOC, thanks, in style B

Well you might not, but picking five books at random from my
bookshelf (it's so nice having a computer at home, finally),
reveals:

Norman Sherry, _The Life Of Graham Greene, Vol. I_: Various *-forms
in TOC.

Jan White, _Graphic Design for the Electronic Age_: ditto.

Floyd L. Moreland & Rita M. Fleisher, _Latin: An Intensive
Course_: ditto

Leslie Lamport, _LaTeX: A Document Preparation System_: ditto.

Bentley Layton, _The Gnostic Scriptures_: ditto.

I haven't tried, but I suspect that there are no examples of the
opposite practice in my library.

> > + All sections contained in it should be unnumbered. Not sure
> >   about "numbered-within" items like equations and figures. The
> >   situation has yet to come up
>again, this is not a universal rule. it might make sense, though, to
>copy the (S)GML world, and have a higher level constrcut like
>\begin{prefacesection} and \end{prefacesection} in which, by default,
>sections are all * type

>isn't Don's requirement catered for by the values of tocdepth and
>secnumdepth, used judiciously? one can reset them on the fly, yes?

Kind of. You get one tocdepth per document, the one in affect
when \tableofcontents is executed. One _could_ exert some effort
in writing \setcounter{tocdepth} commands to the toc, but I'm not
inclined to bother. Setting secnumdepth to a negative number for
the prefatory material, though is a very appealing idea and
solves the problem quite nicely. I think I'm going to change a
few sections of my LaTeX book once I finish this note.

Incidentally, in keeping with established LaTeX usage, I would
suggest that we have single commands to indicate the beginning of
prefatory material and the beginning of the body a la \appendix
and \ps (think about that one... the analagy seems to be the best
way for my students to understand what \ps is about when I teach
the LaTeX course).
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     David Rhead ...
Date:     20 Feb 91 16:20:30

BACKGROUND

Most users of TeX and LaTeX 2.09 are not committed TeXies: they just want
typesetting that looks as if it was produced by traditional means.  If
TeX/LaTeX are working in terms of structures/concepts/practices that are
different from those adopted in mainstream publishing, authors (and support
staff) are going to be forever translating from LaTeX-speak to publisher-speak.

I appreciate that TeX and LaTeX 2.09 were produced in limited time, and that
if their authors had had to completely research structures/concepts/practices
before they started coding, they might never have had time to write any code.
But if, with the move to LaTeX 3.0, there are opportunities to phase out
the old ways of doing things and to phase in new ways of doing things,
I think that it is worth taking these opportunities to move LaTeX in the
direction of traditional mainstream publishing.  Then the communications
problems between designers and computer-people will be reduced, and authors
will be able to get what designers intend fairly painlessly.

Thus, I tend to think that, if some aspect of LaTeX 2.09 is out of line
with traditional mainstream publishing practice, then it is LaTeX 2.09
that is in "the wild blue yonder".  Admittedly, it's not always easy to
identify what is fundamental in traditional mainstream publishing, but its
better TO TRY than to risk leaving LaTeX out on a limb.

\CHAPTER*

As far as I can see, there is no concept in traditional mainstream
publishing practice that really corresponds to \chapter*.

The relevant concepts seem to be "front matter" (or "preliminary pages",
or "prelims") and "back matter" (or "end pages" or "back matter").
Within these divisions, there are top-level units (Preface,
Acknowledgements, Glossary, References, Index), but they aren't
really chapters.  See, for example,
"Chicago Manual of Style", Chicago University Press, 1982,
      pages 4-5;
Judith Butcher, "Copy-editing", Cambridge University Press, 1981,
      chapters 7, 8, 9;
Hugh Williamson, "Methods of Book Design", Yale University Press,
      1983, chapter 8;
Ruari McLean, "The Thames and Hudson Manual of Typography",
      Thames and Hudson, 1980, chapter 10;
Jan V. White, "Graphic Design for the Electronic Age", Watson-Guptill,
      1988, pages 156-161;
John Miles, "Design for Desktop Publishing" Gordon Fraser, 1987,
      pages 58-61.

This divergence between LaTeX 2.09 and traditional publishing practice
seems to be what causes: Don's problems; Sebastian to have to
countenance resetting tocdepth and secnumdepth "on the fly" (OK for
Sebastian, but I think that it would just put ordinary users off);
ordinary mortals to mess around with \pagenumbering{roman} and
"\pagenumbering{arabic} right after the first \chapter command".

Don wants to signal "prefatory sections" (i.e., what a publisher would
call "front matter", "preliminary pages" or "prelims").  In his design,
"front matter" has:  TOC entry; modified headings; page numbering
in roman; equations and figures unlikely.  I presume that modifying the
effect of \chapter* to meet Don's requirements would violate the
decision that "LaTeX 3.0 should be able to process LaTeX 2.09 input
files", so we'd be talking in terms of phasing something else in to
meet his requirements.

Sebastian's design happens to treat "front matter" differently from Don's,
but I think he still has a chunk of document that would be recognised
as "front matter", etc.

I agree with Sebastian that "a higher level construct" would be appropriate.
I'd suggest something like \begin{prelimpages} ... \end{prelimpages}
\begin{maintext} ... \end{maintext} \begin{endpages} ... \end{endpages}
[If you want a precedent, I think that MIT-press-book.sty envisages that
the author will go \begin{frontmatter} ... \end{frontmatter}.]
Then the user wouldn't have to bother about: the \chapter/\chapter*
distinction; \pagenumbering; \addcontentsline.  Such details should
be taken care of by the designer, via the style file.  The designer
will be thinking in terms of "preliminary pages" and "end pages",
the user will "tell" LaTeX (via the environments) which pages
are "preliminary pages" and "end pages", the user will have to think
in the terms in which his/her publisher is thinking but won't have
to bother with how to coerce LaTeX to make the "preliminary pages"
and "end pages" conform to the publisher's design.  As a bonus, "preliminary
pages" could have a \countN set appropriately to help page selection, e.g.
\count2 = 1, 2, 3 might mean "preliminary pages", "main text", "end pages".

...
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     N.POPPELIER@NL.ELSEVIER
Date:     Thu, 21 Feb 91 09:03:38 +0000

I agree with David Rhead that there is a definite need for a division of a
book in front matter, body and back matter. In fact, last summer I suggested
that we start defining the structure of every class of documents for which we
will make a document style in SGML-like terms. Existing SGML dtd's for books
already contain the division David suggested.

...
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     CA_ROWLEY@UK.AC.OPEN.ACS.VAX
Date:     Mon, 8 Apr 91 13:22:53 GMT

This message is partly ... and partly to
say that David Rhead's suggestions seem to be a very reasonable
outline programme for when we get to that stage.

However, in the last sentence, he says:

> I'd suggest a move towards the "publishing-industry standard" structures.
>
The only such standard I know of at present for the logical structure
of books is the AAP's DTD.

I should therefore like to hear people's views as to whether we should
be thinking about doing this and, also, from those more familiar with
it (Nico!), whether it would be feasible/sensible etc.
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     Michael Downes <MJD@COM.AMS.MATH>
Date:     Mon, 8 Apr 91 18:46:30 CET

I am currently working on a documentstyle with a slightly weird design.
It's for a book consisting of chapters by various authors; the chapters
are collected from various sources, and may be previously published, or
given as a talk at some conference but not previously published, or
written specifically for the current book. I am wondering about
questions such as the following.

---Structure of a chapter: Make it exactly like a journal article?

  \title{Chapter title}
  \author{Chapter author}
  \affiliation{Author's university or institution}
  \maketitle

Or use \chapter instead of \title?  Each chapter is numbered with a big
chapter number, which wouldn't normally be the case in, say, a
conference proceedings volume.

Maybe it's only a question of semantics; but if the publisher wants to
call them chapters, my first choice is to use the \chapter command, to
avoid confusing end users and the publisher's staff.

However, this immediately raises another difficulty: \chapter
does not normally have an associated \maketitle command, but the format
of the author names and affiliations is such that typesetting
each one as it comes along, and getting it in the right place,
would be technically difficult. So I am leaning to the format

  \chapter{Chapter title}
  \author{Chapter author}
  \affiliation{Author's university or institution}
  \makechaptertitle

But this is a little out of synch with the LaTeX manual.

---Processing each article separately or processing them together using
\include. The main benefits of processing them together seem to be (a)
less chance of error in the page numbering and (b) cross-references
between chapters---but this is unlikely to be applicable for
independent articles.  The main drawback of processing articles by
different authors together is the possibility of a global change in one
article affecting subsequent articles in undesirable ways. Local
changes can of course be limited to the original article by enclosing
it in a group. In AMS journals we have experimented with processing
all the articles of an issue together, and encountered undesirable
global changes, as well as save stack overflow from the extra
level of grouping, often enough that eventually we went to separate
processing, with the sequence of articles controlled in a VAX/VMS DCL
command procedure, and with the page number being passed from one
article to the next using \write and \read.

---There is also a possibility in any given volume that certain commands
might be required both within a chapter and at the outer level:
\tableofcontents, \bibliography, \appendix. Something of course
can be managed, but does anyone have previous experience and/or
suggestions on how the user interface ought to look? Should I
provide \tableofcontents and \chaptertableofcontents,
\bibliography and \chapterbibliography, etc.? Or how about
putting commands such as \frontmatter, \middlematter, \backmatter
in the driver file that would change the effect of \tableofcontents,
\bibliography, and \appendix commands? I.e.,

       \documentstyle{X}

       \begin{document}
       % front matter beginning here
       \tableofcontents
       \include{preface} % containing \chapter{Preface}
       \include{notation} % containing \chapter{List of Notation}

       \middlematter
       \include{chapter1} % normal chapters, maybe containing
       \include{chapter2} % their own \tableofcontents, \appendix,
       \include{chapter3} % or \bibliography commands.
       ...

       \backmatter
       \appendix % switch to appendix format
       \include{app1} % \chapter{Appendix title}
       \include{app2}

       \include{biblio} % \bibliography
       \include{index} % \begin{theindex}...\end{theindex}

       \end{document}


---On another tangent, the way the \appendix command works has always
seemed slightly odd to me. It would seem more natural to make \appendix
a direct substitute for \section (in an article documentstyle) or
\chapter (in a book documentstyle). Maybe in LaTeX 3.0, with its
attribute handling, we'll be able to say something like

  \section[variant=appendix]{...}

?

The chief difficulty seems to be resetting the counter to 1, for the
first such command, and incrementing it naturally thereafter. Other
changes would be straightforward.
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     bbeeton <BNB@COM.AMS.MATH>
Date:     Mon, 8 Apr 91 19:42:51 CET

mike downes has suggested the term \middlematter .  the term
\bodymatter is generally accepted among the sgml people i hang
around in standards meetings, so i'd like to suggest that as a
substitute.
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     N.POPPELIER@NL.ELSEVIER
Date:     Fri, 12 Apr 91 15:10:43 +0000

Chris more or less invited me to comment on the point of publishing-industry
standards and SGML. By coincidence, I have been looking at these things the
past week; next week I will be in CERN working on the AAP dtd and conversions
SGML <--> LaTeX.

O.K., just a quick reply. Sorry haven't got more time. You can skip
everything of this message if you wish, except the part with the ! in
column 1.

There are no real publishing-industry standards: there is a group of
publishers who support the AAP dtd, but the AAP dtd's for scientific articles
and books are not always suited to the needs of publisher X. Therefore,
publisher X adapts the AAP dtd to his purposes and (hopefully) tries to stay
as close to the original as possible.

[The original, by the way, contains several errors, as was pointed out in
EPSIG News by Derek Coleman and Jan Bleeker of our company.]

! Nevertheless, the AAP dtd's for scientific articles and books serve as good
! starting points and I would suggest that we try to create document styles
! that are as close to these dtd's as possible.

What you will immediately notice when you try this, is that they contain much
more elements and sub-elements, especially in the front matter. The
front-matter elements of LaTeX's standard document styles are totally
insufficient. I can understand Michael Downes' attempts at refining this
structure -- in fact I've been doing the same here.

[Another interesting initiative in the SGML field is the TEI initiative,
which is an initiative from the area of humanities and linguistics. A
different approach, but draft version 1.1 of TEI P1 contains interesting
thoughts.]
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From      David Rhead ...
Date:     12 Apr 91 20:46:08

Here are a few comments about Chris's and Nico's postings.

Martin Bryan gives a DTD for textbooks in Appendix C of his "Author's
Guide to SGML".  I also seem to remember that the British Library had
a project that was defining some DTDs (but I could have remembered wrongly).

Irrespective of whether there are formal standards yet (in terms of SGML
DTDs, etc.), there does seem to be an informal consensus about "the
structure of a book" within the publishing industry.  See, for example,
      Hugh Williamson.  "Methods of book design", chapters 7 and 8.
      Judith Butcher.  "Copy-editing", chapters 7-10.
      Ruari McLean.  "Thames and Hudson manual of typography", chapter 10.
      Jan V. White. "Graphic design for the electronic age", pages 157-161.
      "Chicago Manual of Style", pages 4 and 5.

With a bit of luck, we'll find that the emerging DTDs are consistent
with the previous informal consensus.  For example, if the AAP and Chicago
University Press both manage to produce books in the end, they must both
be talking about the same thing!

As regards the "more elements and sub-elements, especially in the front
matter", I think that this is a price worth paying if it means that
LaTeX and the .sty file can relieve an author of more of the work.
With more accurate knowledge of the structure, LaTeX could take care of
details like:
- when to have page-numbers, and whether they should be arabic or roman
  (rather than authors having to mess about with \chapter*,
  \addcontentsline, \pagenumbering{roman}, \pagenumbering{arabic}
  "right after the first \chapter command")
- when to have table-of-contents entries.
This would help "free authors from formatting concerns to allow them to
concentrate on writing".

\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     CA_ROWLEY@UK.AC.OPEN.ACS.VAX
Date:     Sat, 13 Apr 91 16:01:11 GMT

Nico writes (apart from the very useful earlier parts of his message):

> [Another interesting initiative in the SGML field is the TEI initiative,
> which is an initiative from the area of humanities and linguistics. A
> different approach, but draft version 1.1 of TEI P1 contains interesting
> thoughts.]
[TEI = Text Encoding Initiative]
I have not read this document but knew of its existence:
my reason for not mentioning it in my message is that my understanding
of the TEI project is that it is intended to code a far larger range
of "texts" (in particular, historical material) and a far larger range
of properties of those texts (eg what prining press and which
compositor produced them) than any system aimed at typesetting
documents would need to encompass.

Nevertheless, I am sure it contains some matter of interest.  Is it
any easier to study than the AAPs DTD?
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     David_Rhead@UK.AC.NOTTINGHAM.CCC.VME
Date:     Sun, 14 Apr 91 17:36:30 BST

Barbara suggested \bodymatter (as apparently used by SGML people)
as an alternative to the \middlematter used in one of Michael's examples.

The publishing gurus I mentioned in my last contribution use the terms:
"The text" or "main text" (Williamson, p. 180)
"The text" or "main book" (Butcher, pages 119 and 130)
"Main text" (McLean, p. 157)
"Text" (White, p. 159)
"The text" or "text proper" (Chicago, pages 4 and 22)

Would \maintext be worth considering as a command-name?  (If we use existing
publishing-industry jargon rather than defining new jargon, the end-user
gets spared the job of interpreting new jargon in terms of old jargon.
Unless, of course, the publishing industry's jargon is in the process of
changing, in which case we'd be better off using what they are changing to.)
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     David_Rhead@UK.AC.NOTTINGHAM.CCC.VME
Date:     Mon, 15 Apr 91 09:54:21 BST

Michael raised the question of a book that consists of contributions by
various authors.

Obviously, at the moment, he has to do whatever is necessary to get this
particular book published.  However, it would be nice if, in the long term,
there was a recognised way of producing these things.  So I'll comment on
the questions he raised from a long-term point of view, even if the
comments are not relevant to his short-term difficulty.  It seems part
of the question of a "conference-proceedings style" that I think Frank
mentioned informally at Cork (unless I've remembered wrongly).

Generally, I think that having jargon/structures designed for one situation
and then mis-using them in another situation will tend to give confusion
and trouble.  If computer-people and publishing-people use different jargon,
everyone is going to waste time translating from one to the other.  If
they use the same jargon (in this case "chapter") but mean different things,
that's even worse!

In this situation, "being out of synch with the LaTeX manual" is only
to be expected, since the LaTeX 2.09 standard styles weren't really
designed to handle multi-author works.  Using \author when its not
clear whether the author is the "overall editor" or the "author of this unit",
or \chapter when the unit concerned isn't what everyone thinks of as a chapter,
seems like asking for trouble.  The fundamental problem seems to be that
the 2.09 standard styles don't include one that is designed for dealing
with multi-author works.

So, what's required?  Judging by the Chicago Manual of Style and Butcher's
"Copy-editing", we are talking about "multiauthor works", that consist
of "contributions" which are written by "contributors", where the whole
thing is made into a book-like object by a "volume editor".  So an
appropriate set of jargon/structure might go like:
      \documentstyle[...]{multiauthor}
      \volumeeditor{...}
      \begin{frontmatter}
      ...
      \end{frontmatter}
      \begin{maintext}
      \begin{contribution}
          \contributor{...}
          ...
      \end{contribution}
      \begin{contribution}
          \contributor{...}
          ...
      \end{contribution}
      \begin{backmatter}
      ...
      \end{backmatter}
"Contribution" sounds like a word that is already in use in this context, so
that people in the same situation as Michael and his publishers could use it
to avoid confusion about "the publishers calling the units chapters, but
\chapter not being appropriate".  Its sufficiently neutral to allow both
designs that have "chapter numbers" and those that have no "chapter numbers".
[Within each contribution, it does seem more like a "journal article" than
like a "\chapter" so, if \volumeeditor was adopted to mean "volume editor",
one could probably use \author rather than \contributor without (too much)
confusion.]  Hence, if necessary, \volumetableofcontents,
\contribtableofcontents, etc. perhaps?

If an approach such as that outlined above was adopted, only a small
proportion of the people involved (the volume-editor, the publisher's staff,
the style-file hacker) would need to be aware of any of the above.
The individual contributor to a multi-author work can regard themselves
as producing a paper (e.g. in 2.09's \documentstyle{article});  indeed
the volume-editor might send them some "instructions for contributors" that
tell them to do just that.  So specialised commands/environments (such
as "contribution") needn't necessarily be mentioned in the documentation
aimed at ordinary end-users/contributors.  Contributors could pick up
\section etc. from the body of the LaTeX 3.0 manual; details (of specialised
environments like "contribution") needed by volume-editors could be given
in an appendix of the LaTeX 3.0 manual or just as comments in the style files.
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     David_Rhead@UK.AC.NOTTINGHAM.CCC.VME
Date:     Mon, 15 Apr 91 10:28:13 BST

Michael commented that:
    " ... the way the \appendix command works has always seemed slightly odd
    to me. It would seem more natural to make \appendix a direct substitute
    for \section (in an article documentstyle) or \chapter (in a book
    documentstyle). Maybe in LaTeX 3.0, with its attribute handling, we'll be
    able to say something like \section[variant=appendix]{...}"

I presume that doing the above literally would conflict with the requirement
that LaTeX 3.0 should be able to process LaTeX 2.09 input files (which might
use \appendix as defined in the 2.09 manual).  So \appendix would have
to be left with its 2.09 meaning, but something better could be phased in.
Perhaps one could have
..
\begin{backmatter}
   \begin{appendices}
   ...
   \end{appendices}
   ...
\end{backmatter}

I think that, if a document has just one appendix, then it is silly to call
it "Appendix A".  It can just be called "Appendix" (since there is no
"Appendix B").  If LaTeX can't work out for itself that there is only
one appendix (e.g. from the .aux file), it might be appropriate to allow
a variation of "appendices" that tells LaTeX that there is only one appendix,
e.g. \begin{anappendix} ... \end{anappendix}.

There is the problem of how to introduce the individual appendices:  by using
\chapter (in book, report, etc.) or \section (in article etc.), or by using
something else?  I don't have any specific suggestions here (unless anyone
likes \toplevel), but I think that the question needs considering in
conjunction with questions about how to introduce other "top level" units
that appear in the front-matter and back-matter (such as Preface,
Acknowledgements, Glossary).
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     Michael Downes <MJD@com.ams.math>
Date:     Mon 15 Apr 91 12:55:35-EST
To:       David_Rhead ...

> I think that, if a document has just one appendix, then it is silly to call
> it "Appendix A".  It can just be called "Appendix" (since there is no
> "Appendix B").

I was thinking about this very thing over the weekend. I surmised
that in LaTeX 2.09 the user should be instructed to do something like

       \appendix
       \renewcommand{thechapter}{}
       \chapter{Text of the title}

(in a book). But I haven't tried this yet to see what might happen with
table of contents entries and running heads.

...
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     N.POPPELIER@NL.ELSEVIER
Date:     Mon, 22 Apr 91 08:34:57 +0000

The TEI draft report is certainly easier to read than the AAP documentation.
Furthermore, the AAP dtd is more or less a dtd for the exact sciences, you
know: with lots of math and tables. TEI aims at linguistics, humanities, etc.
It is useful to have a look at both, since two of the flaws of LaTeX's
document styles are, in my opinion,

1. not enough structure
2. aimed at exact sciences (with exception of chemistry),
   hardly usable for humanities
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     N.POPPELIER@NL.ELSEVIER
Date:     Mon, 22 Apr 91 08:48:54 +0000

I am strongly in favor of the nomenclature of the Electronic Manuscript
Standard (ANSI Z39.59-1988), otherwise known as the AAP dtd.

book|article|journal == (fm, bdy, bm)

fm: front matter
bdy: body, which in turn is defined as 1 or more chapters|sections|articles
bm: back matter
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     CHAA006@UK.AC.RHBNC.VAX
Date:     Mon, 22 Apr 91 13:57:27 BST

Bearing in mind Franks' frequent injunctions to make a contribution
rather than allow silence to imply tactit acceptance, I would like
to add my vote to support (front matter, body matter, back matter).
Strangely I would prefer end matter to back matter for reasons which
are not clear even to me, but realise that the term is ambiguous and
could encompass front matter as well as back matter.

     ** Phil.
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     MITTELBACH FRANK <PZF5HZ@EARN.RUIPC1E>
Date:     Tue, 23 Apr 91 15:34:29 CET

I just finished an issue of some internal newletter printed at the EDS
which has the problem of being a typical multi-author work. There is
the overall stuff like table of contents as well as individual articles
from different authors. There is also some `backmatter' material
like a glossary which is compiled by the editor.
The individual articles are of (at the moment) two different types
namely `contributions' and `notices' in our internal jargon.

Guided by production experiences the style for this newsletter
was set up in a way that individual articles are prepared in
the standard article style syntax.

Reasons:

   1) The individual author is able to typeset his/her article
      with any style (house or otherwise) that implements `article
      makeup'. He is not forced to know about special names and
      functions but can view his article in the way the document is
      natural to him, namely as an individual article which is so far
      not necessarily part of some larger structure.

   2) The only thing we request is to separate the article from
      the driver file to keep unnecessary editorial work to a minimum.
      That is, we suggest to use

       \documentstyle{article}

       \begin{document}
         \input{myarticle}
       \end{document}

      The file myarticle then contains the real article with
      definitions and all, and can therefore be pasted directly into
      other structures.

   3) The style itself contains environments like `notice' and
      `contribution' that redefines \author \maketitle and the like
      to produce a desired layout. Some of these commands are ignored
      depending on the kind of article. This means that the editor is
      free to turn any article into a contribution or a notice or ...
      without changing the source.
      This concept was considered important as it speeds up processing
      time. Changes to articles are done only for the sake of
      editorial corrections (typos, grammar, contents) but are not
      necessary when simply fitting the article into the larger
      context of the newsletter.

   4) On the newletter surface, commands like \tableofcontents
      are available but produce information assembled from \maketitle
      etc of individual contributions and notices.

   5) We think of adding a section `letters to the editor' which will
      certainly be implemented as some letter environment that accepts
      standard letter GML but turns this into suitable output for the
      newletter.


Problems:

  As mentioned by several people on the list, the front matter
  declarations in LaTeX are clearly not sufficiant. In fact they
  are nearly non existant. This means that we currently have to
  help ourselfs by abusing fields like \date which then in turn
  produce a problem in such a concept.
  But this only means that we have to define a proper front matter
  concept that will allow to specify necessary information in a
  natural manner.

  I don't consider the question of compatibility here a difficult one
  since the current LaTeX provides nearly nothing which can be easily
  provided in a `compatibility style' or changed in an older source.

  The use of identical names(e.g. \tableofcontents) for slightly
  different things might be considered confusing but this is something
  I think is in practice perhaps even more natural then inventing new
  names (like \volumetoc \issuetoc ...) that are actually all commands
  for the same concept only that the special outcome is determined by
  the position in the document.
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     MITTELBACH FRANK <PZF5HZ@EARN.RUIPC1E>
Date:     Tue, 23 Apr 91 15:35:34 CET

David said some time ago:
>
> I presume that doing the above literally would conflict with the requirement
> that LaTeX 3.0 should be able to process LaTeX 2.09 input files (which might
> use \appendix as defined in the 2.09 manual).  So \appendix would have
> to be left with its 2.09 meaning, but something better could be phased in.

Let me say a few words to the compatibility problem.

While compatibility is certainly an important issue, we
should not make it the main issue. Seeking better
solutions to some problems might result in some
incompatibility in the end but once we found a solution
we will probably also find a way to successfully parse
older documents.

I think that it is wrong to start with the idea ``LaTeX
2.09 has an \appendix therefore LaTeX 3 needs to have
the same concept.'' To start with the goal that we
``necessarily can process LaTeX 2.09 *within* 3.0
documents without any changes to the old sources''
seems to be too restrictive (to me).  We should try to
achieve this if it is feasible but we should not defend
bad concepts for the sake of full compatibility.

In other words: when we found some concept feasible we
can then ask how to achieve compatibility but not the
other way around.
\end{verbatim}\end{footnotesize}
\begin{center} --- \end{center}
\begin{footnotesize}\begin{verbatim}
From:     Don Hosek <DHOSEK@EDU.CLAREMONT.HMCVAX>
Date:     Tue, 30 Apr 91 23:34:00 PDT

Some days ago, it had been suggested that the current syntax of

\appendix
\chapter {...}
..
\chapter {...}
..

be replaced with something more like

\appendix{...}
..
\appendix{...}
..

I would like to suggest that this is a mistake. When I teach
LaTeX classes, my presentation of the \appendix command is that
it switches LaTeX into an "appendix mode" where formatting rules
may change. In the standard report style, the numbering of
chapters is changed from arabic numerals to uppercase latin
letters. In other styles more dramatic changes may be called for:
at BRL their article style, in addition to changing section
numbering also changes equation numbering from 1,2,3 through the
document to A1,A2,A3...; B1,B2,B3... in each appendix. The
working design for my class notes switches to nine point type and
two columns for the appendices.

I find that presenting the material in this fashion (and making
analogies with it for the \ps command of the letter style),
allows users to see pass the seeming illogicality of the scheme
and have it make sense again. What's more, it seems to me that
this scheme is somewhat _more_ logical in terms of document
structures and also (slightly) easier to implement.
\end{verbatim}\end{footnotesize}