Grant said:
I am in the process of designing an XML standard that I hope to submit to
W3C . Yes, by myself so far.
What is the general consensus concerning the use of XMP as part of a
larger standard? Would my standard then be held hostage by Adobe? Would
W3C even accept a standard which included the use of XMP?
I don't know if Adobe have any special limitations, but XMP is a set of
RDF vocabularies, and RDF was designed (at the W3C) so that
vocabularies can be created, extended and specialised independently
(and in general without standardization by the W3C). So it would be
rather perverse if Adobe did hold your standard hostage. Whether the
W3C would be interested in standardising a specific vocabulary is
another matter (offhand, of the thousands out there I can only think of
two they've taken on, SKOS and Content Labels). I'd also suggest
checking Swoogle (
http://swoogle.umbc.edu/) to make sure the domain of
interest hasn't already been covered elsewhere.
Extension/incorporation of other vocabs is pretty straightforward. For
example, say I wanted to define a term "projectOwner" which could be
used alongside the DOAP (Description of a Project) vocabulary. DOAP has
the "Project" class and various properties like "maintainer" involving
people, and they use the "Person" class from the FOAF vocabulary. The
"Rights" schema in XMP has xmpRights:Owner. Putting all this together,
my RDF Schema might look like this:
<rdf:RDF xmlns:rdf="
http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:rdfs="
http://www.w3.org/2000/01/rdf-schema#"<rdf
roperty rdf:about="
http://example.org/myschema/projectOwner">
<rdfs:subPropertyOf
rdf:resource="
http://ns.adobe.com/xap/1.0/rights/Owner" />
<rdfs:domain rdf:resource="
http://usefulinc.com/ns/doap#Project" />
<rdfs:range rdf:resource="
http://xmlns.com/foaf/0.1/Person" />
</rdf
roperty>
</rdf:RDF>
Whenever I then used the term, e.g.
<rdf
escription rdf:about="
http://example.org/something">
<myschema
rojectOwner rdf:resource="
http://example.org/something"
/>
</rdf
escription>
- it would be possible to infer that
http://example.org/something was a
doap
roject and
http://example.org/something was the foaf
erson with
the xmpRights:Owner rights.
(That's a contrived example, xmpRights:Owner expects a literal as the
object rather than a resource, rather a sucky bit of XMP, IMHO)
Cheers,
Danny.