Format of the RDFa profile in RDFa 1.1 Core (Re: Editor's Drafts Updated)

The current document says, section 10 (referring to vocabulary documents):

"These documents must be defined in an RDFa Host Language. They MAY be defined in other RDF serializations as well (e.g., RDF/XML [RDF-SYNTAX-GRAMMAR] or Turtle [TURTLE])."

and it also says, right before Appendix C: "RDFa and RDF/XML versions of the vocabulary SHOULD be provided."

Couple of comments on that

- first of all, the note and the text are not equivalent (should vs. may)

- I am not sure we should keep it as general as RDFa Host Language. Say, in some years, somebody comes up with a XYZML+RDFa spec, and this spec makes additions to the processing steps, a bit like RDFa 1.1 XTHML does. According to this text, from that moment on, it is o.k. to define the vocabulary in XYZML+RDFa. However, RDFa processors may not be able to use that, because they have not yet updated their implementation to the extended processing rules.

I would propose to be more restrictive here, and simply say HTML5+RDFa and XHTML+RDFa.

- That being said, the original proposal (I think it was Toby's) was to say that "Conformant processors MUST understand (X)HTML+RDFa, and MAY understand other RDF serializations formats". There is a difference here. Say that the market evolves in such a way that all RDFa processors understand Turtle. That means that a deployment that relies exclusively on Turtle would be fine, and the authors would not feel like breaking the rules. But the current text would not give this freedom...

Ivan

----
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