 |
 |
On Thu, Sep 30, 2004 at 07:25:56AM +0300, Patrick Stickler wrote: > So, presuming that <http://purl.org/dc/elements/1.1/> does not denote > a schema (and not caring further what it might denote): > > ... <rdfs:isDefinedBy rdf:resource="http://purl.org/dc/elements/1.1/"/> > > should rather be > > ... <rdfs:isDefinedBy rdf:resource="http://dublincore.org/2003/03/24/dces"/> > > Thus, IMO a reasonable set of URIs could be: > > http://dublincore.org/2003/03/24/dces denotes (one version of) the DC RDF schema > http://purl.org/dc denotes the complete DC vocab > http://purl.org/dc/dcmitype denotes a DC subvocabulary > http://purl.org/dc/elements/1.1 denotes a DC subvocabulary > http://purl.org/dc/terms denotes a DC subvocabulary I had not understood why a string commonly used as an XML namespace in encodings of DC metadata today -- "http://purl.org/dc/elements/1.1/", with the slash -- could not also function as DCMI's official identifier for the notional "vocabulary" in an RDF sense (i.e. a "set of terms"). I believe Patrick said that there is no requirement in the current XML namespace spec that the URI denote anything at all, so such a use would not actually violate any rules. However, I understand there are at least three dangers: -- There is a long-term risk that the set of members of the XML namespace could become disjunct with respect to the set of members of the RDF vocabulary. -- There is risk that people who incorrectly assume they should be able to click on this "URL" will get confused. -- In the future, someone might want to build a pig farm at that location (metaphorically). If that is the case, then Patrick's proposal above makes sense. (I wonder whether isDefinedBy would perhaps need to point to a "latest version" identifier like http://dublincore.org/dces, but that is a minor quibble). 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]