>
> On Thu, Nov 13, 2003 at 11:33:17AM -0500, Harry Wagner wrote:
> > > Alternatively, we might generate an RDF schema and pass it
> > > back to the institution (or individual) from whence it came
> > > for posting on their own Web server and harvesting by the
> > > DCMI registry.  I'm assuming that this would not be difficult
> > > to do from a technical standpoint (is that true Harry?).
> >
> > It can be done, but there are some issues that would have to be addressed.
> > For example, the distributed schemas would all be making assertions about
> > the same resources.  These could easily get out of sync.  It seems to me the
> > only assertions we (we being the registry) could trust are the
> > specific-language literals (those that have an xml:lang attribute of fr-FR,
> > in the case of the French schema).  That would solve the matter of trust
> > from the registry perspective, but what about other people & applications
> > that may unknowingly treat those schemas as authoritative?
>
> Ah, yes, it's all coming back...  I guess I had gotten used
> to the idea that these "annotations" would limit themselves
> to asserting the language-specific bits with respect to the
> resource identified by the official term URI, possibly with
> a pointer to the exact historical version of which it is
> a translation.  In other words:
>
>     rdf:about="http://purl.org/dc/elements/1.1/title"
>     Label: [the label for "Title" in Chinese]
>     Definition: [the definition for "Title" in Chinese]
>     Translates: "http://dublincore.org/usage/terms/history/#Title-003"
>
> With respect to the RDF schema of the DCMI-maintained version
> in English, such an annotation for Chinese would _not_
> redundantly carry the following information which users in
> China might want to see displayed when viewing "Dublin Core
> in Chinese":
>
>       <rdfs:isDefinedBy rdf:resource="http://purl.org/dc/dcmitype/"/>
>       <rdf:type rdf:resource="http://www.w3.org/2000/01/rdf-schema#Class"/>
>       <dcterms:issued>2003-11-11</dcterms:issued>
>       <dc:type rdf:resource="http://dublincore.org/usage/documents/principles/#vocabulary-term"/>
>       <rdfs:subClassOf rdf:resource="http://purl.org/dc/dcmitype/Image">
>
> In order to display "DC-Chinese" in its entirety, that
> information would either need to be included in the external
> annotation, which as you point out is redundant and risks going
> out of synch with the original over time.  Or one would rely
> on the DCMI Registry to merge the Chinese annotation with
> the central DCMI schema and display the integrated result.
> Or one would rely on some other server-side application to
> merge and display the schemas in that manner.
>
> I lean towards keeping the external annotations minimal and
> relying on applications to merge them for display.

Is that very practical?

rs



>
> Tom
>
> --
> Dr. Thomas Baker                                [log in to unmask]
> Institutszentrum Schloss Birlinghoven         mobile +49-160-9664-2129
> Fraunhofer-Gesellschaft                          work +49-30-8109-9027
> 53754 Sankt Augustin, Germany                    fax +49-2241-144-2352
> Personal email: [log in to unmask]
>
>