%%% ====================================================================
%%%  @LaTeX3-article{ LaTeX3-L3-006,
%%%  filename        = "l3d006.tex",
%%%  archived        = "ctan:/tex-archive/info/ltx3pub/",
%%%  author          = "David Rhead",
%%%  doc-group       = "Project core team",
%%%  title           = "Structures to be supported",
%%%  version         = "1.00",
%%%  date            = "April 1992",
%%%  time            = "",
%%%  status          = "discussion paper",
%%%  abstract        = "",
%%%  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        = "61107 558 3086 23976",
%%%  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.",
%%%  }
%%% ====================================================================


%  These notes are along the same lines as Wessel Kraaij's comments, as
%  copied from comp.text.tex by Rainer.

%  Please feel free to replace ...
   \documentstyle{l3ms001}
   \sloppy
%  ... by a command to map a "2.09 article structure" into whatever
%  design you like which is compatible with that structure.

\newcommand{\BibTeX}{{\rm B\kern-.05em{\sc i\kern-.025em b}\kern-.08em
      T\kern-.1667em\lower.7ex\hbox{E}\kern-.125emX}}

\begin{document}

\title{Structures to be supported}
\author{David Rhead}
\date{April 1992}
\maketitle

\tableofcontents

\section{Motivation}

\subsection{Dual role of 2.09 style files}

In \LaTeX\ 2.09, ``style files'' confuse 2 roles:
\begin{itemize}
\item definition of a structure.
      (I think I've heard this given as a justification for the designs:
      ``the designs don't matter, because the style-files are just there
      to define what structures are supported''.)
\item mapping the structure into a design.
\end{itemize}
For example:
\begin{itemize}
\item {\tt article.sty} and {\tt xarticle.sty} seem to map the same
      structure (the ``\LaTeX\ 2.09 analysis of the structure of an {\tt
      article}'') into different designs
\item the command \verb+\documentstyle{siam}+ seems to map ``{\sc siam}'s
      analysis of the structure of an article'' into ``{\sc siam}'s design for
      an article''.  (If so, then the roles of \verb+\documentstyle[11pt]{siam}+
      and \verb+\documentstyle[12pt]{siam}+ are unclear.  Are they intended
      as ``preprint styles'', for an author to use while a paper is
      being drafted?)
\item the commands \verb+\documentstyle[onecolumn,10pt]{iso}+ and
      \verb+\documentstyle[twocolumn,9pt]{iso}+ seem to map ``the structure
      of an ISO standard'' into ``two designs for ISO standards''.
\end{itemize}
Changing from \verb+\documentstyle{article}+ to
\verb+\documentstyle{xarticle}+ will work (i.e., give a document with the
same structure but a different design), but changing from
\verb+\documentstyle{article}+ to \verb+\documentstyle{siam}+ or to
\verb+\documentstyle{iso}+ won't work.

The analysis that was done for version 2.09 is mostly implicit in {\tt sty}
files, rather than being available explicitly.  Subsequent providers of
{\tt sty} files have generally followed the same practice (although some
provide supplementary documentation).  Thus, a user who just wants his/her
current structure laid out in a different design may end up reading lots of
archived {\tt sty} file code to see whether an alternative {\tt sty} file
supports the structure they are currently using.

\subsection{Practical difficulties at 2.09}

Someone attempting to produce a book/thesis with \LaTeX\ 2.09
has to change various defaults:
\begin{itemize}
\item they will want their preliminary pages numbered in roman, but will
      want to switch back to arabic at the start of their main text.
      They may need a \verb+\setcounter+ to start the roman sequence at the
      right place.
\item they will probably want units such as ``acknowledgements'' and
      ``references'' to appear in their ``table of contents''. They will
      be using \verb+\chapter*+ for such units (to get headings that
      look appropriate), so will have to use \verb+\addcontentsline+
      to get the units mentioned in the ``table of contents''.
\item page-selection can be a problem, since \verb+\count0+
      doesn't distinguish between roman and arabic.
\end{itemize}

People producing ``an issue of a journal'', or a conference-proceedings,
may have additional problems:
\begin{itemize}
\item if they treat the work as {\tt book} \cite[p.\ 23]{lamport}, they
      will be faced with trying to get authors' names, affiliations, etc.,
      typeset consistently at the start of each chapter.
\item if they leave the work as a series of {\tt article}s, they will be faced
      with ensuring that numbering (of pages, etc.) follows on.
\end{itemize}

I think that these problems arise because the analysis of ``document
structure'' for the \LaTeX\ 2.09 ``standard styles'' is inappropriate:
\begin{itemize}
\item the concepts of ``front matter'' and ``back matter'' are well
      known in publications about book design, etc., but aren't supported
      by the 2.09 ``standard styles''
\item an ``issue of a journal'', and a conference-proceedings, have structures
      of their own (which are different from the structure of an ordinary book).
\end{itemize}

\subsection{Suggestion for 3.0}

I think that there would be advantages in:
\begin{enumerate}
\item
     keeping a clear distinction between ``the structure supported'' and
     ``the design into which the structure is mapped'', so that the
     end-user will known when they can/cannot change design by changing
     just one line of their {\tt tex} file.
\item
     associated with (1), thinking in terms of ``software that maps a
     structure into a design'', rather than ``style file'' which confuses
     ``structure supported'' with ``design into which the structure is
     mapped''
\item
     analyzing in more depth the structure of the types of documents to be
     supported.  Then, for example, the end-user will be able to just say
     ``this is front matter'', and so have details such as roman/arabic
     numbering, heading style, and ``table of contents'' entries taken care
     of automatically in accordance with the relevant design (or
     house-style).
\end{enumerate}

\section{Structures to be supported}

Various gurus \cite{chicago,aap,majour,white,bs-thesis,bs-report,tei} give
analyses of the structures of the document-classes for which people use
\LaTeX\ 2.09.  Although, there may be differences between the analyses
offered by different gurus, I think that it would be better for the project
to consult the gurus, rather than to ignore them (since otherwise the
project will waste time re-doing the work already done by the gurus).

\subsection{First proposition}

It is easier to select the ``good bits'' from off-the-shelf analyses than
to start from nothing.

\subsection{Second proposition}

For overall structure, particularly ``good bits'' are to be found in the
{\it Chicago Manual of Style} \cite[pages 4,5]{chicago} and in the SGML DTDs
published by the Association of American Publishers \cite[appendix B]{aap}.

\subsection{Notes about AAP analysis}

\subsubsection{Three basic structures}

Broadly, the AAP analysis defines three basic structures:
\begin{description}
\item[BK-1] book/monograph/textbook, conference proceedings, technical report,
      thesis/dissertation
\item[ART-1] article, feature
\item[SER-1] serial,\footnote{%
      I think they mean ``an issue of a serial''.  See section
      \ref{not-a-serial}.}
      conference-proceedings.  (Articles are embedded within
      a serial without any modifications.)
\end{description}

Thus, although the \LaTeX\ 2.09 and AAP analyses both define 3 main structures,
the boundaries are drawn differently:
\begin{itemize}
\item
     the AAP regards reports and books as having the same basic BK-1
     structure (unlike \LaTeX\ 2.09, which has two distinct ``standard
     styles'', {\tt report} and {\tt book}, the main difference being that
     {\tt book} isn't allowed to have an abstract)
\item
     whereas the \LaTeX\ 2.09 manual suggests \cite[p.\ 23]{lamport} that
     ``it is easy to include an article as a chapter in a report or book'',
     the AAP defines a special structure, SER-1, for multi-author works
     made up of separate articles.
\end{itemize}

\subsubsection{Borderline cases}

Some types of document may lie on the borderline between two AAP
categories.  For example, a long report might be divided into units called
chapters and be appropriately classified as BK-1, while a short report
might be divided into units called sections and be more akin to ART-1.
(Compare \cite{aap}, which envisages that technical reports will have BK-1
structure, with \cite{bs-report} which envisages that technical reports
will be divided into sections.)

At worst, cases that cross borderlines might need two mappings.  In the
report example, they might be (1) a mapping of BK-1 structure to a report
design, and (2) a mapping of ART-1 structure to a report design.  (This is no
worse than the situation with the \LaTeX\ 2.09 ``standard styles'', which
also envisage that {\tt report} is made up of chapters.) Thus, adoption of
AAP classification would not cause insuperable difficulties.

\subsubsection{Advantages of AAP analysis}

Paying serious attention to the AAP analysis would have various advantages:
\begin{itemize}
\item the analysis is well-known
\item it covers the types of documents that current \LaTeX-users generally
      want to produce
\item it generally (but not always) gives analysis to the depth that
      \LaTeX\ 3.0 might need
\item other gurus cite it (if only to disagree with it)
\item it embodies knowledge about publishing practice that the
      average \LaTeX-er doesn't have
\item it is finite.  The project would not be attempting to analyse
      all possible documents, but would be concentrating on the structures
      that are most commonly required.
\item support for AAP-like structures might attract ``real publishers''
      to \LaTeX
\item {\it it exists now} (unlike, for example, the European
      journal-publishers' work \cite{majour}, which is still in progress).
\end{itemize}

\subsubsection{Disadvantages of AAP analysis}
\label{not-a-serial}

The AAP analysis is not suitable in all respects as a model of what
\LaTeX\ 3.0 (and associated software) should do:
\begin{itemize}
\item The analysis may go deeper in some areas than is required for \LaTeX\ 3.0.
      (If something with the ``look and feel'' of SGML is required, one might
      as well use SGML.)
\item On the other hand, the analysis doesn't go deep enough in other areas.
      For example, the AAP analysis is no better than the \LaTeX\ 2.09 analysis
      for:
      \begin{itemize}
      \item citations and reference-lists \cite{iso-690};
      \item captions, legends and credit-lines \cite[ch.\ 11]{chicago};
      \item notes to tables \cite[ch.\ 12]{chicago}.
      \end{itemize}
      Other gurus' analyses would have to be used in such areas.
\item At a recent SGML meeting \cite{exeter8}, the AAP standard was
      described as too Anglo-centric.  Hence the European journal-publishers'
      work \cite{majour} was criticised as too AAP-influenced.
\item ``Serial'' is usually used (e.g., by librarians) to describe a
      publication that could potentially continue indefinitely (e.g., all
      the issues of a journal, including those not yet published).
      Therefore SER-1 may be a poor choice of name for a structure that
      would represent ``a single issue of a series'' or a ``one-off
      conference proceeedings''.  MULTI-1 might have been a better name
      for such multi-author (or multi-article) works.
\end{itemize}

\subsubsection{Attitude to AAP analysis}

An analysis based on that given by the AAP (minus a few details, plus a few
other details which could be taken from \cite{chicago} and other places)
might give a reasonable compromise between:
\begin{itemize}
\item ``re-inventing the wheel''
\item blindly obeying something (AAP structure) that hasn't been thought through
 in all
      respects.
\end{itemize}

\subsection{Example}

For example, if some compromise was made between the Chicago analysis and
the AAP analysis, one might envisage a user preparing a file of the form:
\begin{verbatim}
.
\begin{frontmatter}
   \frontelement{Foreword}
   ...
   \frontelement{Preface}
   ...
   \frontelement{Acknowledgements}
   ...
   \frontelement{Dedication}
   ...
   \frontelement{Abstract}
   ...
\end{frontmatter}
\begin{bodymatter}
   \chapter{...}
   ...
\end{bodymatter}
\begin{appendices}
   \appendix{...}
   ...
\end{appendices}
\begin{backmatter}  %  short for "other back matter"
   \backelement{Glossary}
   ...
   \backelement{Notes}
   ...
\end{backmatter}
\end{verbatim}
%  Perhaps \oneappendix{...} could be substituted for
%     \begin{appendices}
%     \appendix{...}
%     ...
%     \end{appendices}
%  if there is only one appendix.

\subsubsection*{Note}

The above example assumes that, for \LaTeX\ purposes:
\begin{itemize}
\item it will generally be sufficient to think in terms of ``front matter
      elements'' and ``back matter elements''
\item it will not be appropriate to go for complete analogy with the AAP's
      SGML DTD.  Thus, one would not expect typical \LaTeX\ 3.0 software to
      define environments like \verb+\acknowledgements+: if \LaTeX\ 3,0 gets
      used as a back-end for an SGML system, the AAP's \verb+<ack>+ would
      get converted to \verb+\frontelement{Acknowledgements}+.  (However, if
      any elements required special treatment, particular environments could
      be defined for them, as is done in \LaTeX\ 2.09 with {\tt abstract}.)
\end{itemize}

\subsection{Bonuses from using structures influenced by Chicago/AAP}

\subsubsection{Conference proceedings}

A document with structure based on SER-1 will naturally support ``a
reference-list at the end of each chapter'' plus the option of a composite
bibliography at the end of the document (for which one sometimes sees
in requests in electronic digests from editors of conference proceedings).

Because the structure is appropriate, there should be less conflict than if
one is (for example) mis-using the 2.09 {\tt book} structure.  One
shouldn't end up trying to have {\tt thebibliography} at both 2.09
``section'' level and at ``chapter'' level, because the ``references at end
of article'' and ``bibliography at end of complete work'' units would have
different definitions.  For \BibTeX, one could envisage a scheme involving
perhaps {\tt article1.bbl}, \dots\ , {\tt article}$N${\tt .bbl}, {\tt
backmatter.bbl}.

\subsubsection{Page selection}

Such an analysis would lead naturally to schemes for using the \verb+\count+s
sensibly, so as to support:
\begin{itemize}
\item distinction between roman and arabic numbered pages (i.e., front
      matter and main text)
\item selection of ``all the front matter'', a whole chapter, a whole appendix,
      or ``all the (non-appendix) back matter''.
\end{itemize}
For example, one might have:
\begin{center}
\begin{footnotesize}
\begin{tabular}{lllll}
\hline\hline
Major division & Minor divisions&\verb+\count0+&\verb+\count1+&\verb+\count2+\\
\hline\hline
Front matter   &                & page-number  & {\tt -1}    &  {\tt 0}  \\
\hline
Main text      &   chapters     & page-number  & chapter-number  & {\tt 0} \\
               &                &              & {\tt 1, 2, ... }&         \\
\hline
Back matter:            & appendices    &page-number&appendix number& {\tt 1}\\
appendices              &               &           & {\tt 1, 2, ... }& \\
\hline
Back matter:   & glossary          & page-number&{\tt -2}   &  {\tt 0}  \\
other units    & bibliography      &            &            \\
               & index, etc.       &            &            \\
\hline\hline
\end{tabular}\end{footnotesize}\end{center}

\subsubsection{Generally}

Generally, if a correct analysis of structure is made, practical details
will tend to fall into place nicely, rather than needing messy {\it ad hoc}
circumventions.

\section{Modularity}

\subsection{Analysis in general}

Although I've suggested that a Chicago/AAP analysis might provide a suitable
basis for ``structures to be supported by \LaTeX\ 3.0'':
\begin{itemize}
\item any ``\LaTeX\ 3.0 project'' selection of ``the good bits''
      is unlikely to be perfect
\item some better analysis may come along, and some successor to the
      ``\LaTeX\ 3.0 project'' may want to support that analysis rather than
      one derived from the suggestions given here
\item people may have to produce ``structure to design'' mappings for
      structures other than the 3 ``modified AAP'' ones (e.g., legal
      articles, ISO standards, SGML DTDs other than the 3 AAP ones)
\item people may produce ``structure to design'' mappings for
      enhancements of the 3 ``modified AAP'' ones (e.g., a book
      that contains plates  and maps, if they are to be numbered
      separately from other illustrations).
\end{itemize}
It seems desirable that such possibilities should be borne in mind when any
software is being written.

\subsection{Structure supported by mapping software}

I think that mapping software (i.e., whatever we call the successors to
2.09 ``style files'') should make clear the structure that is supported,
for example:
\begin{itemize}
\item by software checks
\item because the mapping software starts with a comment that defines the
      structure it supports.
\end{itemize}
Then end-users will know (or find out in a friendly way) whether or not
they can simply ``change design by changing one line of the {\tt tex} file''.

For example:
\begin{itemize}
\item a thesis and a Wiley book might have the same structure, so
      a one-line change to the {\tt tex} file should lead to a change from one
      design to another (with no change to the structure)
\item papers in different physics journals will probably have the same
 structure, so
      a one-line change to the {\tt tex} file should lead to a change from one
      journal's design to another's
\item ISO standards have their own structure.  Someone who tries to apply
      typesetting software that is intended for a different structure (e.g., an
      AAP-based structure) should get a sensible error message.
\end{itemize}



\begin{thebibliography}{00}
\addcontentsline{toc}{section}{References}
\bibitem{lamport}
     {\sc Leslie Lamport.} {\it LaTeX:  a document preparation system.}
     Addison-Wesley, 1986.
\bibitem{chicago}
     {\it Chicago manual of style.}
     13th edition.
     Chicago: University of Chicago Press, 1982.
\bibitem{aap}
     {\it Electronic manuscript preparation and markup: ANSI/NISO Z39.59-1988.}
     New Brunswick: Transaction publishers, 1991.
     ISBN 0-88738-945-7.
\bibitem{majour}
     {\it DTD for article headers.}
     Amsterdam: European workgroup on SGML, 1991.
\bibitem{white}
     {\sc Jan V. White.}
     {\it Graphic design for the electronic age.}
     Watson-Guptill, 1988.
\bibitem{bs-thesis}
     {\it Presentation of theses and dissertations.}
     BS 4821. British Standards Institution, 1990.
\bibitem{bs-report}
     {\it Presentation of research and development reports.}
     BS 4811.  British Standards Institution, 1972.
\bibitem{tei}
     {\sc C. M. Sperberg-McQueen and Lou Burnard.}
     {\it Guidelines for the encoding and interchange of machine-readable
     texts.}
     Draft version 1.1.
     Oxford, Chicago: Text Encoding Initiative, 1990.
\bibitem{exeter8}
     {\sc Michael Popham.} {\it Report on inaugural meeting of
     UK chapter of SGML Users' Group.} Report number 8, SGML project, Exeter
     University, 1992.
\bibitem{iso-690}
     {\it Documentation --- bibliographic references --- content, form and
     structure.}
     ISO 690.  International Organization for Standardization, 1987.
\end{thebibliography}

\end{document}