Phillip has raised the point of non-front matter issues of the standard journal
class. To this I would like to comment as follows:

1. BibTeX: yes, there should be an officially approved bst file for each
   journal. BibTeX does not have to be obligatory, but it should
   certainly be available. With my makebst (aka custom-bib, merlin,
   genbst) generating such files is greatly simplified.

   Marcel's comment that making the bibliography by hand is trivial
   compared to the rest of the article makes me laugh and cry together.
   The bibliography (by hand) is far from trivial, if you want it right,
   and it is an enormous  effort for really arbitrary, fiddly, piddly
   stuff. I would rather make up a bib file for a one-shot paper than
   manually make up a thebibliography  listing.

2. Most journal style files (few issue classes) do include extra features
   for handling tables, figure captions, references, sublabelling
   equations. Much of this can be handled by existing packages, such as
   my natbib. It would therefore be better to recommend the standard
   packages that one should  be able to use, such as the tools collection
   and amsmath. That is,  a paper using the model journal class could
   include all these packages  without having to submit them separately
   to the journal; he can assume they will be at any installation that
   contains journal.cls.

3. Abbreviations and spelling should not be accommodated in any such

   packages. The author can provide this himself. I did this in my Guide
   to LaTeX (and I  even explain how I did it that book) and the
   publisher really did demand US spelling after saying British spelling
   would be okay at first. (Actually in  the book I describe how one
   could do it; what I really did was
      \newif\USspell \USspelltrue
      \newcommand{\USUK}[2]{\ifUSspell #1 \else #2\fi}
   Now you can refer to \USUK{color}{colour} as you please, and just
   change the state of the \USspell flag to switch. You can even define
   \newcommand{\colour}{\USUK{color}{colour}} to simplify typing. (Note
   that \color is probably already taken.)

4. The journals I have written classes for want a manuscript first with
   the figure captions and tables all at the end, not included within the
   text. This is anarchistic today. However, I have a package figcaps
   that  allows this. Either the figures appear as normal in the text, or
   the captions (and optionally the figures themselves) and tables are
   written to  separate files to be reread at the end. I have not made
   this package  public, but parts of it are available inside my class
   files.

   I use this as an example of some very complex things that some
   journals require. My suggestion for journal.cls would be to FORGET IT.
   A draft or  manuscript option should be available (for single column,
   double spacing) but such juggling of figures and tables should be left
   out. Let the  figures and tables appear in the manuscript text too. It
   is only out of historical reasons that these are to be listed at the
   end, from the bad old typewriter days.


Those are my comments for now.

Patrick
-----------------------------------------------------------------------------
   Dr. Patrick W. Daly                    Tel. [+49] 5556-979-279
   Max-Planck-Institut fuer Aeronomie     Fax. [+49] 5556-979-240
   Max-Planck-Str. 2
   D-37191 Katlenburg-Lindau              Internet: [log in to unmask]
   Germany
-----------------------------------------------------------------------------