I agree with Sebastian on this matter Guy,
Here at the University of Montreal we use an Access database to manage the
metadata to our theses. Then, with an export module we convert to the
TEIHeader, DC in HTML, and other interchange formats. We haven't programed
the RDF export template yet since we are waiting to see IF it is worthwhile
to use this format.
And since we are closely working with other universities that uses
proprietary formats for metadata, building a new export template is very
simple and with a simple click you have all your metadata in the new format.
> Guy Teasdale writes:
> > we are setting up a FileMaker Database on the web that will be used
> > by the students to enter themselves their cataloguing
> > information. Then, always
> > with Omnimark, we will convert the FileMaker records to a Teiheader
> > structure that will be pasted on top of the thesis.
> > As I don't want to make all these efforts if TEAHEADER is
> > eventually going to an RDF schema, I asked this question.
> I don't think I see the problem. You want the students to provide
> metadata; great. You give them a front-end to type it into. You
> then convert the database record to some *ML markup. Why does it
> matter whether its TEI header or RDF? You can go from one to the other
> later, surely? You are not tying yourself into either system.
> This assumes that the TEI header can encompass all that is needed by the
> RDF, of course! That, surely, is the issue, as to whether RDF adds
> lots more stuff undreamt of by the TEI frontierspeople.