On Wed, 25 Oct 2000, Dave Beckett wrote:
> >>>Andy Powell said:
> [...]
> > Therefore, our simple description should take the form
> >
> > <dc:creator>
> > <dct:Agent>
> > <rdf:value>Dave Beckett</rdf:value>
> > </dct:Agent>
> > </dc:creator>
>
> The document is for unqualified DC so this is out of scope; you may
> argue that it is too simple, but I would plead that we should chop up
> the DC syntax format problems into small bits, and encoding the
> 15-element DC in XML is a discrete part.
As a bit of XML it is fine. As a set of RDF statements I am still not
convinced that it says what we want it to say. What we want to say is
that there is a resource (which has some properties) and there is a
related agent (which also has some properties) and the two things are
related. This is true even in DC-Simple. The RDF arc-node diagram that
you get from your current RDF/XML encoding does *not* say this.
If we don't make the separation of resource and agent explicit in the
simple RDF/XML encoding (ie, if we don't make it explicit that two things
are being described) then we force people to embed the understanding of
what is really meant by the encoding into their applications. I think
that this runs counter to the 'semantic Web' spirit of RDF ?
I accept that the use of dct:Agent in the above RDF snippet is out of
scope - because we don't have, or understand, dct:Agent in simple DC.
However,
<dc:creator>
<rdf:Description>
<rdf:value>Dave Beckett</rdf:value>
</rdf:Description>
</dc:creator>
would be perfectly allowable.
Andy.
> I'm updating the draft after DC8 and one of the changes is to allow a
> URI reference (RDF/URI people - don't start!) for every element. I
> added it to every one since I'm sure people could give examples of
> why it was needed there, rather than just restricting it to the
> elements that specifically said that they could take a URI.
>
> So you could use this to point to an 'agent resource':
>
> <dc:creator rdf:resource="http://resource-describing-creator-agent/"/>
>
> and still be simple-ish. In this case, the creator is an identifier
> of the agent resource as Simon Cox prefered.
>
> The form you mention above including the dct: namespace will be in
> the next XML document, something like 'An XML Encoding for Qualified DC'
> (again using a DTD)
>
> Dave
>
Andy
--
Distributed Systems and Services
UKOLN, University of Bath, Bath, BA2 7AY, UK [log in to unmask]
www.ukoln.ac.uk/ukoln/staff/a.powell Voice: +44 1225 323933
Resource Discovery Network - www.rdn.ac.uk Fax: +44 1225 826838
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|