On Thu, 8 Jan 2004, Lorna M. Campbell wrote:
> Andy - didn't you mentioned that you had done some work on a document
> that illustrates how to encode different identifier schemes as URIs? If
> so, would you be willing to share this with the group?
The document is at
http://www.ukoln.ac.uk/metadata/dcmi/dc-identifiers/2002-07-24/
As you can see from the date, it is now quite old. Note in particular
that the DC in HTML meta tag syntax that I was using then is now slightly
out of date (e.g. I used DC.Identifier whereas current guidelines say to
use DC.identifier).
Some of the identifier-specific stuff may now also be out of date - for
example, it looks as though the 'issn' URN namespace never made it out of
the trenches?
I wrote the document in order to try and convince the DCMI community that
they should always/only use URIs as identifiers - otherwise there was, and
indeed still is, a real danger that DCMI ends up effectively becoming a
registration agency for identifier 'schemes'. I don't see any point in
this when we already have a registration agency in the form of the URI
machinery. The same argument holds true for the IMS community IMHO.
However, I didn't manage to take this document forward within DCMI - there
was some technical concern about some of the schemes (I can't remember the
details but I suspect that people complained that the doi: URI scheme
isn't registered for example) and, I half suspect, there may be some
politically-led views inside DCMI that DCMI *should* become a registration
agency for these things. But in part it was just that moving these kinds
of documents forward takes quite a lot of will-power and determination
and I probably simply gave up :-(
Andy
--
Distributed Systems, UKOLN, University of Bath, Bath, BA2 7AY, UK
http://www.ukoln.ac.uk/ukoln/staff/a.powell +44 1225 383933
Resource Discovery Network http://www.rdn.ac.uk/
|