Re: following your nose from an RDFa document to the RDFa spec

I just want to bump the thread because of some new work that is going on
 at ping.semanticweb.org [1] which I find highly relevant to the RDFa
work. I was suggesting to Uldis Bojars that we should add RDFa support
to his Semantic Radar Firefox extension [2] as well as to the ping
service. However, we can't just ping the service with *every* XHTML page
so I'm thinking we will need something like a profile or link/@rel
pointer to indicate that there's RDFa in XHTML besides changing
namespace documents at w3.org.

TODO: write PHP RDFa parser.

-Elias


[1] http://pingthesemanticweb.com/
[2] http://rdfs.org/sioc/firefox


Dan Connolly wrote:
> In the new GRDDL WG, Fabien Gandon recently gave a pointer
> to some work on RDFa and GRDDL...
> http://lists.w3.org/Archives/Public/public-grddl-wg/2006Jul/0015.html
> -> http://www-sop.inria.fr/acacia/soft/sweetwiki.html
> and especially
>   http://www-sop.inria.fr/acacia/soft/RDFa2RDFXML.xsl
> 
> In #swig chat[1], Elias reminded me of some RDFa test data
>   http://svn.rdflib.net/trunk/test/rdfa/
> 
> I tried RDFa2RDFXML.xsl on the 000001.htm test and I'm
> getting different results from 000001.ttl .
> 
> So I wonder if the spec says which one is right, and
> I tried to follow my nose to find it.
> 
> There doesn't seem to be anything in the 000001.htm
> test document that says "this is not just any HTML document;
> it's an HTML document with RDF data inside; any RDF triples
> you extract per the RDFa spec are indeed meant by the author."
> 
> Please add some such signal to the test data, so that
> there's a chain of authority from the URI spec
> to the HTTP spec to the MIME specs, then somehow
> to the RDFa spec, a la 
> section 3.1.1. Details of retrieving a representation
> of webarch
>  http://www.w3.org/TR/webarch/#dereference-details
> 
> 
> I prefer a GRDDL-compatible signal; i.e. either
>   (a) change the namespace document
>    at http://www.w3.org/1999/xhtml to refer
>    to have a GRDDL namespaceTransformation link
>    to something like RDFa2RDFXML.xsl, or
> 
>   (b) change the namespace of the 000001.htm document
>    to something like http://www.w3.org/2006/08/xhtml-rdf
>    or perhaps http://www.w3.org/2002/06/xhtml2/ per [xhtml2]
>    and arrange for a namespace document there with
>    a namespaceTransformation link
>    to RDFa2RDFXML.xsl , or
> 
>   (c) add a profile to the 000001.htm document to
>    an RDFa profile, which has a profileTransformation
>    link to RDFa2RDFXML.xsl , or
> 
>   (d) add the GRDDL profile (http://www.w3.org/2003/g/data-view )
>    to the 000001.htm document and add
>    <link rel="transformation"
> href="http://www-sop.inria.fr/acacia/soft/RDFa2RDFXML.xsl" />
> 
> 
> I gather that the draft that was used to prepare the RDF
> test data in http://svn.rdflib.net/trunk/test/rdfa/ is
> http://www.w3.org/2001/sw/BestPractices/HTML/2005-rdfa-syntax
> 
> In there, I see
> "The aim of RDF/A is to allow [RDF graph]s to be carried in XML
> documents of any type."
> 
> p.s. the 000001.htm says that it's XHTML 1.0 transitional,
> but it has role and property attributes. Hmm... the
> validation service says "Sorry, I am unable to validate this document
> because its content type is text/plain, which is not currently supported
> by this service." So actually, the document doesn't say that
> it's XHTML 1.0 transitional; it says that it's plain text.
> So following ones nose a la web architecture won't even get
> as far as the XML or XHTML specs; it'll stop cold at text/plain.
> That's perhaps just a test suite hosting issue, but please let's
> do get some real test data where we can follow our nose all the
> way thru from the URI of the document thru the stack of
> specs to the RDF triples and then thru the URIs in those triples
> and so on.
> 
> 
> [1] http://chatlogs.planetrdf.com/swig/2006-08-03#T18-18-51
> 
> 
> [xhtml2] http://www.w3.org/TR/xhtml2/conformance.html#doccont
> 

Received on Wednesday, 23 August 2006 19:06:12 UTC