All,
I've updated
Guidelines for encoding DC in XML
http://www.ukoln.ac.uk/metadata/dcmi/dc-xml-guidelines/
Apologies for the delay in doing this.
The text below summarises the changes that have been made based on
comments sent to the list. Apologies if I've missed comments!
Given the current discussion about what 'simple DC' means, I guess we need
to extend the comment period for this document?
<changes>
1) We've removed the link to the XML schemas from the references. There
were a number of specific comments related to the schemas we proposed.
We agree that these XML schemas need more work. Enhancing the XML schemas
will continue separately from this document.
2) We've removed the section on use of Xlink. We agree that this proposal
is premature. We will make a separate proposal about the use of Xlink to
this WG.
3) There were some comments about the use of the dcxml namespace prefix
for the 'scheme' attribute. Having investigated the use of scheme in
XHTML, we recognise that this namespace is not strictly necessary.
Therefore we have removed the use of this namespace from the guidelines
document.
</changes>
Finally, there was a question about why the guidelines have moved away from
recommending nesting of attributes to indicate element refinement. This
recommendation has changed a number of times during the life of the
document. We currently propose this
<dcterms:alternative> ... </dcterms:alternative>
rather than this
<dc:title>
<dcterms:alternative>
.....
</dterms:alternative>
</dc:title>
We propose to leave this recommendation unchanged. The reasons for this are
as follows:
- this recommendation is in line with the current RDF/XML encoding
recommendation
- using nesting to indicate 'refinement' would be a DC-specific
convention. This convention is not shared by other XML applications, e.g
IMS and ODRL. We feel that a flat encoding structure is more in line
with other XML applications. (Note, nesting of elements is used in IMS
and ODRL, but not to indicate refinement).
We acknowledge that the downside to this recommendation is that metadata
applications will not be able to dumb-down metadata based solely on the
instance metadata. Some knowledge about DC will have to be embedded into
the metadata application to perform dumb-down. Our view is that most
non-RDF XML applications are likely to have to have quite a lot of knowledge
embedded into them anyway. (This is clearly one of the significant
advantages of the RDF approach).
There were also comments about the use of xml:lang rather than dc:language
to encode the language of the element value. My understanding is that use
of xml:lang is now consistent across these XML guidelines and the RDF/XML
recommendations. Therefore, our recommendation to use xml:lang remains
unchanged.
Regards,
Andy and Pete
--
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/
|