On Wed, Mar 12, 2003 at 10:14:45AM -0500, Harry Wagner wrote:
> > In this case, does the literal "2003-03-12" fit into "en-US", or
> > is that an example of a squeeze...?
> If we don't assign an xml:lang attribute to all literals, then how do you
> propose applicaitons, such as the registry, differentiate between
> natural-language style literals and non-natural-language style literals?
> When asked to display information about a term in a specific language how
> would the application know to display literals that happen to be dates?
That seems like an application question...? The modeling issue
is whether it makes sense to qualify the literal "2003-03-12"
with "en-US", and Jon seems to confirm what I had suspected --
that it doesn't really "make sense"...
> We should be consistent with our use of xml:lang. If the property value is
> a literal we should use it.
This seems wrong to me ("if the property value is a literal
we should say it is in English"), but harmless enough...?
> > If dc:type is more specific than rdf:type, it is not clear to
> > me whether that extra specificity is really helpful in this
> > case, where it is being used to point to the type of a term
> > -- which happens to be how the term rdf:type is (more often)
> > used in RDF schemas.
> If we would not use dc:type for this then I have a difficult time imagining
> what we would use it for. It seems a perfect fit to me.
Yes, we _could_ use dc:type, but rdf:type seems like a
perfect fit too, and is (I would guess) perhaps the term that
other RDF schemas are more likely to use?
Alternatively, we could be even more specific by coining
a dcu:termtype. I'm generally interested in hearing any
opinions people on this list may have about when to coin and
when to re-use.
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