Message-ID:  <m0xJH2C-000OWTC@dcarlisle.demon.co.uk>
Reply-To: Mailing list for the LaTeX3 project
              <LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE>
In-Reply-To:  <97100910281439@multivac.jb.man.ac.uk> (message from Phillip
              Helbig on Thu, 9 Oct 1997 10:28:14 GMT)
Date:         Thu, 9 Oct 1997 12:45:40 +0100
From: David Carlisle <david@DCARLISLE.DEMON.CO.UK>
Sender: Mailing list for the LaTeX3 project
              <LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE>
To: Multiple recipients of list LATEX-L
              <LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE>
Subject:      Re: LaTeX journal and publisher macros
Status: R

Philip writes (as usual I'm only commenting on the bits I don't like,
no one gets any praise around here:-)

> \note
You will need several categories of note. Either by allowing multiple
commands to be declared (as I did in fmatter) or by some extra
argument (which is syntactic dressing for the same thing).

> %THE TWO ARGUMENTS FOR EACH \item IN THE ENVIRONMENTS BELOW WORK SIMILAR
> %TO THE KEY AND THE OPTIONAL ARGUMENT OF \bibitem.

in which case the command shouldn't be called \item (which does not
have any mandatory arguments). The argument structure of latex
commands should not depend on context (which is why for instance
\\ accepts a * form in tabular, even though it doesn't do anything
with it).

\author{long form}{short form}{affiliation-ref}{address-ref}{email-ref}
I don't think that long lists of mandatory arguments with (typically)
short ref strings is very usable in practice, people will get them
out of order, or miss one out. Also it is very inflexible if the
following year you really decide you need one other piece of
informtion then you can not add a new argument to \author or you break
all the existing documents. So it is better I think to have more
commands, or optional arguments rather than mandatory (possibly, or
possibly not) with a key=value syntax

> Should one break the author's name up into initials and surnames, so
Changing order gets complicated, but some classes may want to put
surname only in running head, or to put surnames in small caps.
so some kind of surname markup is probably useful.

Speaking of running heads, you are probably going to need a command
to explicitly set the running head in certain circumstances.
Auto-generating it from the author list is useful and portable when it
works, but you will always find that paper by 10 authors and a robot
where auto generation, even with automatic adding of `et al' doesn't
do the right thing.

> I hadn't been thinking of the abstract as part of the front matter
> (probably because it normally comes after \maketitle) but perhaps I
> should be.

It needs to be part of the frontmatter as often it is set as part of
the title block rather than as part of the main running text.
The standard classes don't do it that way, but many existing classes
do including the AMS classes and many existing journal classes.

Sebastian said

> why not adopt the keyval syntax,

I didn't do it in fmatter because I was trying to keep things looking
more like `normal latex' so I used more comands rather than extensive
use of optional arguments. With a special environment so that
undefined commands were skipped (so any class file automatically can
ignore anything it does not want to deal with).

However... Michael do you want to post your KV suggestion to the wider
list??

David