FYI: Attaching Meta-Information ROLE To XHTML Elements

On an HTML WG call today, Mark Birbeck mentioned this document:
  http://lists.w3.org/Archives/Member/w3c-html-wg/2003AprJun/att-0426/role
of using a role attribute to annotate metadata on HTML elements, such that, 
in the long term, even a RDF parser could parse a HTML document and pull 
the triples out. The discussion touched on three items it'd be worthwhile 
capturing more explicitly.

1. That sounds fantastic, but more aggressive than I was trying to achieve 
with respect to valid instances, and more related to the TAG issue:
  http://www.w3.org/2001/tag/ilist#mixedNamespaceMeaning-13
2. It would cool that if instead of an RDF application only accepting a 
particular syntax, it had a MANDATORY mechanism (i.e. XSLT) that enabled it 
to transform any XML to a suitable serialization prior to consumption. (I 
actually don't know if there's any definition of conformance for an RDF 
application? Does RDF 1.0 only specify a format?)
3. Instead of decorating XML instances with annotations or "wearing the data 
model on the instance" folks have been tooling around with Schema 
annotations such that an XML instance can yield RDF when instance is 
validated against the schema.
  http://www.w3.org/2003/02/schema-annotation.html

Received on Wednesday, 9 July 2003 11:35:27 UTC