Re: Reviews of Self-Describing Web

I've been talking to some other TAG members about updating the Self 
Describing Web draft to reflect the guidance given at the May 2008 TAG 
face to face meeting (Bristol).  In a private reply to me, our scribe Norm 
Walsh confirmed that I had correctly understood that:

| - The paragraph starting "Even though this document is of media type 
| application/xhtml+xml "  needs to be replaced with following your nose 
| through: application/xhtml+xml -> RFC 3236 -> HTML M12N -> 
| http://www.w3.org/1999/xhtml -> RDFa specification

So, I set off to work through the pertinent specifications to understand 
the connection from RFC 3236 ("The 'application/xhtml+xml' Media Type") 
[3] and HTML Modularization (HTML M12N).  I didn't get very far.  In 
particular, the text I see in RFC 3236 that mentions HTML M12N is:

"Work continues in the Modularization of XHTML Recommendation [XHTMLM12N], 
the decomposition of XHTML 1.0 into modules that can be used to compose 
new XHTML based languages, plus a framework for supporting this 
composition."

- and -

"With respect to XHTML Modularization [XHTMLMOD] and the existence of 
XHTML based languages (referred to as XHTML family members) that are not 
XHTML 1.0 conformant languages, it is possible that 
'application/xhtml+xml' may be used to describe some of these documents. 
However, it should suffice for now for the purposes of interoperability 
that user agents accepting 'application/xhtml+xml' content use the user 
agent conformance rules in [XHTML1]."

- and -

"XHTML Modularization provides a naming convention by which a public 
identifier for an external subset in the document type declaration of a 
conforming document will contain the string "//DTD XHTML". And  while some 
XHTML based languages require the doctype declaration to  occur within 
documents of that type, such as XHTML 1.0, or XHTML  Basic 
(http://www.w3.org/TR/xhtml-basic), it is not the case that all  XHTML 
based languages will include it. "

The first of those quotes clearly has no normative force.  The second says 
"it is possible that [the media type] may be used to describe [documents 
that are XHTML family members but not XHMTL 1.0 conformant]"  Does get 
give us the normative follow your nose connection that we need in order to 
claim that RDFa documents are self describing and can be safely 
interpreted by user agents?  I'm not convinced.  I don't think the third 
quote helps much either.

Clearly, the RFC allows for use of the media type with XHTML M12N content, 
insofar as it says that it's "possible" that the media type "may be used 
to describe >some< of these documents".  What's not clear to me is whether 
the appearance of the media type is thus sufficient to trigger formal 
appeal to the M12N specification for interpreting the content.

If any of you who understand these specifications better than I do could 
offer some guidance, I'd be very grateful.  In the meantime, I don't feel 
that it's appropriate for me to update the Self-describing Web draft 
(which is sort of frustrating, since this appears to be the one 
significant technical sticking point in getting it published.)  Thank you.

Noah

[1] http://www.w3.org/2001/tag/doc/selfDescribingDocuments-2008-05-12.html
[2] http://www.w3.org/2001/tag/2008/05/20-minutes#item07
[3] http://www.rfc-editor.org/rfc/rfc3236.txt

--------------------------------------
Noah Mendelsohn 
IBM Corporation
One Rogers Street
Cambridge, MA 02142
1-617-693-4036
--------------------------------------

Received on Friday, 22 August 2008 23:32:07 UTC