>>>Roland Schwaenzl said: > Dave wrote: > > I don't understand what your problem is > > RDF/XML parsed by > different parsers is supposed to yield > the same models or not? Yes. In the existing M&S, the language as part of the literal could be ignored, so the two examples can be seen as equivlent. We are making a change - xml:lang is required to be supported and is significant. This is a good thing, right? > Apparently W3C's RDF validator gives > models different from those you suggest. The validator will be updated in due course - when there is a written up RDF Core WG decision, approved test cases, and the test cases language has been updated. Please be patient. Dave