Re: Updated Editor's Draft

Shane:
On Oct 20, 2011, at 9:51 AM, Shane McCarron wrote:

Folks,

I have updated our source document and am preparing to push an Editor's Draft into date space.  However, in completing my action about namespaced attributes, I was forced to make a decision about the prose that was not explicitly discussed by the working group.  If you look at [1] you will see:

4.3 XML+RDFa Document Conformance

This specification does not define a stand-alone document type. The attributes herein are intended to be integrated into other host languages (e.g., HTML+RDFa or XHTML+RDFa). However, this specification does define processing rules for generic XML documents - that is, those documents delivered as media types text/xml orapplication/xml. Such documents must meet all of the following criteria:

 1.  The document must be well-formed as defined in [XML10-4e<http://www.w3.org/2010/02/rdfa/sources/rdfa-core/Overview-src.html#bib-XML10-4e>].
 2.  The document must use the attributes defined in this specification through references to the XHTML namespace (http://www.w3.org/1999/xhtml).

When an RDFa Processor processes an XML+RDFa document, it does so via the following initial context<http://www.w3.org/2010/02/rdfa/sources/rdfa-core/Overview-src.html#T-initial-context>:

 1.  There is no default collection of terms.
 2.  There are no default IRI mappings.
 3.  There is no default vocabulary IRI.
 4.  The base<http://www.w3.org/2010/02/rdfa/sources/rdfa-core/Overview-src.html#T-base> can be set using the @xml:base attribute as defined in [XML10-4e<http://www.w3.org/2010/02/rdfa/sources/rdfa-core/Overview-src.html#bib-XML10-4e>].
 5.  The current language<http://www.w3.org/2010/02/rdfa/sources/rdfa-core/Overview-src.html#T-current-language> can be set using @xml:lang attribute

Previously as I recall, RDF Core 1.1 did have a default profile applied to all host languages, including XML [2]. This was, in fact, where all of the prefixes were defined; XHTML+RDFa defined mostly link relation terms. We did decide to keep the default profile, now renamed to "initial context". However, I don't see that we decided that XML+RDFa would not have such an initial context. Did I miss something? (Actually, there's not even an ISSUE recorded for removing @profile, just a meeting note [3].

Gregg

Note that this now says that in a generic document, RDFa attributes MUST be referenced in a qualified manner.  Since this is a generic XML document, we cannot assume that unqualified attributes (ones in 'no namespace') are actually relevant to RDFa.  A generic XML document can have ANY elements and attributes (consider private XML structures) and adding RDFa semantics to them has to be qualified so there is no possibility of a collision.  For example, my Real Estate Annotation Language (REAL) might have a property attribute (property="residential"), but clearly that is not the same as @xh:property.

I trust this restriction is consistent with what everyone was thinking in the call.


[1] http://www.w3.org/2010/02/rdfa/sources/rdfa-core/Overview-src.html#xmlrdfaconformance
[2] http://www.w3.org/profile/rdfa-1.1
[3] http://www.w3.org/2010/02/rdfa/meetings/2011-07-28#Removing___40_profile

--
Shane McCarron
Managing Director, Applied Testing and Technology, Inc.
+1 763 786 8160 x120