(unknown charset) Re: TAP classes are strings, not classes

Hi Rob,

(glad to see things happening on this list again :)

* Rob McCool <robm@robm.com> [2004-06-11 11:49-0700]
> 
> > The data at http://tap.stanford.edu/tap.rdf is all like this:
> > 
> >  <rdf:Description rdf:about="EarDrops">
> >     <rdf:type>ProductType</rdf:type>
> >     <rdfs:subClassOf>EarCareProduct</rdfs:subClassOf>
> >  </rdf:Description>
> > 
> > when I assume you mean something like
> > 
> >  <rdf:Description rdf:about="EarDrops">
> >     <rdf:type rdf:resource="&tap;ProductType" />
> >     <rdfs:subClassOf rdf:resource="&tap;EarCareProduct" />
> >  </rdf:Description>
> > 
> > Let me know if it gets fixed.
> 
> Yes, this was fixed a while ago but the downloads weren't updated. That
> entry now looks like this:
> 
>   <tap:ProductType rdf:ID="http://tap.stanford.edu/data/EarDrops">

I believe this should be rdf:about not rdf:ID. The latter gets expanded 
in context of base URI, but shouldn't itself contain characters beyond 
XML NCName. So you could either write rdf:ID="EarDrops" assuming a 
base URI context of http://tap.stanford.edu/data/ or else write 
rdf:about="http://tap.stanford.edu/data/EarDrops".

Spec nerd detail:
http://www.w3.org/TR/rdf-syntax-grammar/#section-Syntax-ID-xml-base
http://www.w3.org/TR/rdf-syntax-grammar/#rdf-id
"An attribute ·string-value· matching any legal [XML-NS] token NCName".
-> http://www.w3.org/TR/REC-xml-names/#NT-NCName
   NCName ::= (Letter  '_') (NCNameChar)*
   NCNameChar ::=Letter |Digit  '.'  '-' |'_' |CombiningChar  Extender


>     <rdfs:subClassOf rdf:resource="http://tap.stanford.edu/data/EarCareProduct"/>
>   </tap:ProductType>

> 
> If you want to download an updated tap.rdf, go here:
> 
> http://panic.stanford.edu/tap-all.rdf.gz

Cool. I'd like to recommend TAP as one of the 'bulk' namespaces that
could be used inside FOAF files, eg. for describing weblog post topics, 
personal interests, stuff being sold etc. Is 
http://tap.stanford.edu/data/ the offial-est namespace URI you folks 
are deploying? Should it be OK to encourage folk to write RDF using this 
namespace and expect it to be dereference-able over the coming years...?

cheers,

Dan

Received on Friday, 11 June 2004 15:15:00 UTC