[Bcc: to AGU infotech list] Report on the AGU InfoTech Committee meeting October 2002, Washington DC
Of the many items that this meeting had to address, the ones that
held the most interest for me were the handling of abstracts and
manuscripts for the meetings and various publications of the AGU. The
organization is moving rapidly into all-electronic publishing and
handling of abstracts and manuscripts, making the electronically
archived version the "version of reference" -- a process not without
its frictions and problems.
As for the abstract submission, this is currently being done by an
HTML form. A back-end first roughly checks syntax and then processes
using LaTeX, presenting the user with a graphical preview. It appears
that earlier problems with diacritics have been solved, a convenience
thing for the many non-Anglo-Saxon abstract submitters. Also the
problem of inserting a percentage sign, commenting out the remaining
LaTeX line, is now caught by the preprocessor. It should be possible
for the innocent contributor to to use a "%" without disaster
striking! Including formulas into the abstract will always require
writing LaTeX manually, as must be expected. Some $$ balancing checks
do take place in the back end.
The AGU publications are rapidly going / have gone towards electronic
archiving in the form of "rich-tagged XML", from which printable (PDF)
and Web (HTML + pixel map graphics) versions are generated on request.
Articles currently may be submitted in many formats, including MS
Word, WordPerfect [Note: AGU HQ continues to use WP and reasons for
not switching to the "industry standard" were graphically expressed
:-)] and the well known LaTeX document classes.
Processing of received manuscripts currently consists of automatic as
well as manual (mark-up) steps. This holds for all formats. LaTeX is
fairly easy to process automatically, but user-produced rich-tagged
XML would definitely cut down on the manual work. *Any* form of XML
would.
Carter Glass showed the XML format currently used by AGU; it differs
substantially from, e.g., DocBook. He promised to send me the DTD with
a view to making LyX produce output in this format.
Having had a look at the way LyX produces XML output disclosed that
only two specific formats are supported in a hard-wired way: DocBook
and LinuxDoc. To support other formats, like the AGU format or also
OpenOffice, would require
(1) merging back DocBook and LinuxDoc support into a single XML
support methods tree, parametrized by the document class layout
files [somewhat the LyX equivalent of DTDs];
(2) writing these layout files separately for DocBook, LinuxDoc and
AGU-XML.
This will be more difficult than the similar mechanism already in
place for the many different -- but among themselves much more similar
-- LaTeX classes; but will also have a greater pay-off. LyX will
probably never be a great rich-tagged XML editor, but it can
definitely be better than it is now.
BTW about rich-tagged: AGU-XML allows/requires you to tag, e.g.,
authors' first and family names separately. LyX/LaTeX was never meant
to do this. This can probably be faked by introducing "user-defined
character styles", a good idea anyway.
In the coming months I intend to write LyX layout files supporting the
various AGU journals' LaTeX formats, a realistic proposition now that
the layout-defined floats mechanism is in place. This will help those
contributors with a preference for LaTeX to easily use LyX for this
purpose.
...freely from memory :-)
--
Martin Vermeer [EMAIL PROTECTED]
Helsinki University of Technology
Department of Surveying
P.O. Box 1200, FIN-02015 HUT, Finland
:wq
msg47701/pgp00000.pgp
Description: PGP signature
