[gmx-developers] Re: markup for macromolecules in biology

[Marc Baaden]

Hi,

just a quick comment,

>>> Michael Banck said:
 >> IMHO XML is not really useful as an input file format, at least not in
 >> general (there might be nice XML editors and emacs-modes, but that would
 >> cut off general easy editing). It's really good for data files though.

I don't totally agree. I can see that it is more tedious to edit an XML input
file (except eg with emacs :))) ), but there is one big advantage: due to the
strict XML structure you have a very good chance that this input file can still
be read by a future program version (eg in 10 years or so). This is often quite
important in the context of a lab where people change frequently and you might
want/need to work with their data a couple of years after they have left.

Furthermore, once you have a "standard" input file, you will only need to
change small bits of it from time to time. You usually don't have to re-create
it from scratch every day. So if Gromacs already came with a simple useful
example, people could restrain to modify that, which even or should I say
in pareticular in XML is an easy task.

Also, XML provides tools to check your document for validity. So you can
eliminate a large number of input file format/syntax problems before even
starting up Gromacs.

And then also XML is - because of its verbosity - very human readable.
As an example: for entries like T-coupling or P-coupling where there
can be several groups, it is not obvious in an mdp file how you need
to format/seperate them (except by reading the doc or having an example).
XML would make that clear from its very structure (DTD) which good editors
(yes they exist !) do parse and can even suggest you which tags you are
allowed at a given position in the file.

Hope that makes my point.

Cheers,
Marc
-- 
 Dr. Marc Baaden  - Institut de Biologie Physico-Chimique, Paris
 mailto:baaden at smplinux.de      -      http://www.marc-baaden.de
 FAX: +49 697912 39550  -  Tel: +33 15841 5176 ou +33 609 843217