On Thu, May 30, 2002 at 05:22:34PM +0200, Roland Schwaenzl wrote:
> > This gets back to the basic distinction between "canonical"
> > (or "authoritative" or "definitive") representations of DCMI
> > terms and representations derived from the same, such as an
> > RDF schema.
>
> Not at all. RDF is not a miracle, but a spezified gadget -
> just as XML Schema.
>
> The schema is saying in it's metadata, which sources it uses.
You sound here like you are disagreeing with me, but I do not
see how. You have drawn my attention to two statements (in
http://www.mathematik.uni-osnabrueck.de/projects/dcqual/qual21.3.1/Schema/A/dcterms)
which I had not in fact noticed before:
| <dc:source
| rdf:resource="http://www.dublincore.org/documents/2000/07/11/dcmes-qualifiers/"/>
| <dc:source rdf:resource="http://www.dublincore.org/usage/decisions/"/>
To me, this says that the RDF schema is based on the Qualifier
recommendation and the Decisions documentation -- in the
terms I have been using, it is a "derived representation"
of the "canonical" Qualifier document. And this is exactly
what I would want and expect the RDF schema to say -- no
disagreement there.
So remind me what are we arguing about?
> > Then I have correctly understood that this particular RDF
> > schema is but one possible representation among others, even
> > other RDF schemas.
>
> Sure...the essential point is just, that the dc website should not
> distribute RDF Schema, which make erronous assertions.
> Most importantly it should not define resources in the namespaces,
> which haven't been accepted by Usage.
I absolutely and emphatically agree. All I am adding is
that we had better define a way to get from one base document
that is revised, reviewed, and approved by the Usage Board to
the set of "derived representations", such as your RDF schema,
as automatically as possible. The experience of DCMI to date
makes me pessimistic that the accuracy of a multiplicity of
representations could be maintained by relying on humans to
proofread and hand-edit them all on a regular basis.
> > If the RDF schema is not presented as, and does not declare
> > itself to be, the canonical representation, then I agree that
> > loading versioning information into that schema may indeed not
> > be the point.
>
> There is no such claim.
Then there is no problem. My misunderstanding. I had not
noticed this:
| <dc:source
| rdf:resource="http://www.dublincore.org/documents/2000/07/11/dcmes-qualifiers/"/>
| <dc:source rdf:resource="http://www.dublincore.org/usage/decisions/"/>
> > 3) an RDF schema in which each term declaration says it
> > "isDefinedBy" the very same namespace URI; and
>
> Please: This is an RDF construct. Don't argue by English grammar.
> That's not the one, which applies.
Fair enough, though it would appear from Pete's posting that
exactly what "isDefinedBy" does mean -- in English or any
other conceptual form -- is not entirely clear...
> > To me, such a construct is misleading because it creates a
> > self-referential circle, saying in effect that "DCMI declares
> > this RDF schema to be _the_ definitive representation." Whereas
> > -- as you have yourself confirmed -- the RDF schema is but "a"
> > definitive representation.
>
> NO! Please read the metadata (dc:source).
Got it. Objection withdrawn.
> In case there is something the pointer goes to the relevant
> Usage decision. I don't know to which other documents you want
> to see a pointer. To the terms-latest.html ? That's just in draft
> status - with wrong hyperlinks.
No, the pointers are fine as is. In the near future, though,
I would expect the pointer to
http://www.dublincore.org/documents/2000/07/11/dcmes-qualifiers/
to be replaced with a pointer to
http://dublincore.org/usage/terms/dc/.
In the meantime, terms-latest.html has been superseded by the
draft at http://www.gmd.de/People/Thomas.Baker/usage/terms/dc/.
> It took me a while to figure out some working URL's.
> Also DC1.1 and the DCQ rec are not cited correctly.
They are correct on
http://www.gmd.de/People/Thomas.Baker/usage/decisions/ --
I just checked.
> This document will undergo changes till it
> becomes stable. I find it rather confusing, that it doesn't
> give dates.
Should I ask Beth to archive the versions in dated directories?
> In case terms-latest converges there will be no problem to add pointers.
Excellent. Problem? What problem??
Thanks,
Tom
--
Dr. Thomas Baker [log in to unmask]
Institutszentrum Schloss Birlinghoven mobile +49-171-408-5784
Fraunhofer-Gesellschaft work +49-30-8109-9027
53754 Sankt Augustin, Germany fax +49-2241-14-2619
|