[foaf-dev] Parsing RDF from namespace documents - anyone reading RDF from inside XHTML? (foaf ns)
danny.ayers at gmail.com
Sat Jun 6 02:07:55 CEST 2009
2009/6/5 Dan Brickley <danbri at danbri.org>:
> Hi all,
> In the FOAF spec we have for a while had various ways of finding an RDF
> description of the vocabulary. The server supports content negotiation,
> or you can go to http://xmlns.com/foaf/spec/index.rdf ... or the HTML
> doc has in it's header,
> <link href="http://xmlns.com/foaf/spec/index.rdf" rel="alternate"
> type="application/rdf+xml" />
Good so far...
> In addition, the XHTML spec has usually had RDF/XML embedded directly
> inside it. I think - especially with the rise of RDFa, the time has come
> to switch this off. It makes the document ill-formed, and isn't a widely
> used deployment style.
> Would any object if future versions of the FOAF spec didn't embed
> RDF/XML in the XHTML? The RDF (ie. RDFS/OWL) will still be accessible
> via content negotiation and a link to index.rdf as above. In addition we
> can add some RDFa, covering some (and perhaps eventually all) of the RDF
> statements from the schema. Libby's begun working on the latter piece.
I would very much like to see the continuation of there being a HTML
version and an RDF/XML version of the spec available through conneg.
Standards and all that. What troubles me a little is that if the HTML
version uses RDFa, it's going to be hard to keep versions in sync, all
Libby, what are the chances of you being leveraged into doing the
whole lot in RDFa? - so something like Jeni Tennison's XSLT can be
used (occasionally) to derive the RDF/XML doc. (or even use GRDDL)
On the other hand, maybe the completeness thing isn't such an issue -
if the RDFa version has a pointer (seeAlso?) to the RDF/XML, any
half-sane agent will get the full vocab. But then that begs the
question of why use RDFa at all, rather than just a rel="alternate"
Best bet is likely sameAs the effort folks are willing to provide :-)
More information about the foaf-dev