Re: Spec update with vocabulary expansion

On Aug 31, 2011, at 22:04 , Gregg Kellogg wrote:

> I just finished implementing @vocab expansion in my processor and came across a couple of things we might consider adding to the spec:
> 
> For my own testing, I used DOAP as the @vocab value, which returns an RDF/XML representation of the vocabulary. Without having an RDF/XML processor also loaded, this results in a load error, but many other things could result in an error. I elected to add a rdfa:UnresovedVocabulary warning to the processor graph in this case.
> 
> To appendix B.2, add:
> 
> [[[
> rdfa:UnresolvedVocabulary a rdfs:Class, owl:Class;
>  dc:description "warning; to be used when a referenced vocabulary fails to load" .
> ]]]

I agree, with a tiny editorial change. We already have an error condition "ProfileReferenceError" which was kept for the management of default profiles, so, for consistency, I'd prefer "rdfa:VocabReferenceError".

I have modified the spec, and have also updated the namespace document.

> 
> Also, rules rdfs5 and rdfs11 don't add interesting information to the default graph; they're only useful as rules to expand of subClass/subProperty chains. I place these in a separate context (t-box). There's no harm in adding them to the default graph, but we should consider adding wording that allows these triples to be removed from the default graph might be useful as a note.
> 
> To 10.1 Details of the RDFa Vocabulary Expansion, change:
> 
> [[[
> The processor SHOULD NOT add the triples appearing in the vocabulary graph to the (expanded) default graph.
> ]]]
> 
> to
> 
> [[[
> The processor SHOULD NOT add the triples appearing in the vocabulary graph, or created as the result of expanding rules rdfs5 and rdfs11 to the (expanded) default graph.
> ]]]

Doh. I intended something like that, but I failed. Actually, I made a little bit more precise:

"The processor SHOULD NOT add the triples appearing in the vocabulary graph only, or created as the result of expanding rules rdfs5 and rdfs11 on the vocabulary graph, to the (expanded) <tref>default graph</tref>."

indeed, it may well be that the expansion with rules rdfs5 and rdfs11 occurs on predicates in the default graph (eg, if the RDFa source described subproperty relationships!) which I believe should be left there. I agree this is a very rare case, but nevertheless...

I have updated the -src.html file!

Thanks

Ivan


> 
> We also need to discuss how we should add tests to the RDFa Test Suite, wether as "optional" tests in the main manifest, or as a completely separate manifest and test suite.
> 
> Gregg
> 
> 


----
Ivan Herman, W3C Semantic Web Activity Lead
Home: http://www.w3.org/People/Ivan/
mobile: +31-641044153
PGP Key: http://www.ivan-herman.net/pgpkey.html
FOAF: http://www.ivan-herman.net/foaf.rdf

Received on Thursday, 1 September 2011 09:19:38 UTC