On Jan 29, 2006, at 6:20 PM, pt wrote:
One thing that particularly concerns me is Word to OpenDocument
interchange (both directions). My experience in our practical
courseware publishing project is that we have to work at the level of
standardised style names to interchange documents (this is the
microformat approach). If you go ahead with an elaborate bibliographic
data model built into OpenDocument then interchange will be very hard,
won't it? Maybe using the forthcoming MS office with native XML it
would be doable, but there are still lots of copies of Word 2000 and
earlier running out there, with no compelling reason for people to
upgrade.
OK, the goods news is that it should be possible to have full fidelity
-- and easy -- interchange between future versions of ODF and MS XML.
Some of this will depend on corporate decisions I have no control over,
but I'm optimistic.
One thing that would very much help future compatibility for everyone
out there with any kind of institutional or corporate connections to
press on Microsoft to include compatible citation support as part of
their ECMA submission of the Office 12 file formats. I have already
mentioned this to Brian Jones at Microsoft, and to people in one or two
other places. In general corporations care nothing about higher ed,
much less citations, so the more they hear this, the better.
But even without that, it would still be possible to use our new
citation support in Word, with its support for custom schemas. It just
wouldn't be standardized.
The metadata issue is actually really easy to solve. Both ODF and the
next versions of Office will have equivalent file structures. In both
cases, then, we could just store the metadata in the file wrapper.
I think compatibility is likely to be very hard to do with older
versions.
I think we could probably go a long way using existing mechanisms like
plain-old hyperlinks to allow people to insert citations which would
point back to an outboard application using your new data model, that
in turn could post-process documents to build a bibliography.
Anyway, I think the new ontological approach looks promising, and I
will try to find the time to look at the foundation Python code you
would need to process it; I'm thinking the first thing to get is an
RDF library so we can read and write the RDF format you're proposing
in and out of a data structure.
In my current coding, I'm just using native data structures to create
the objects. An import driver for the metadata ought to be fairly
simple, and could be an RDF library (like rdflib) or even just an XML
one. But I see that as abstracted from the core code. It should be
possible for someone to write a BibTeX input driver if they want.
Bruce
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
