Norman Walsh wrote:
| So I guess my question is - how is scoping xml:base to not apply | within atom:content where type is xhtml not profiling XML Base?
That would seem really strange to me. The xml:base attribute on an ancestor of atom:content changes the [base uri] property in the infoset. Down inside atom:content, suppose you find an html:div; what is its base URI? To answer that, you look for the applicable [base uri] property. I don't know any other way to do it.
Sure. My point (which still stands afaict) is that this is a structural matter, to do with bleed from the container in the content and that we've solved it with a different approach altogether for namespaces than what's being suggested for xml:base. I guess I'm trying to get the WG to think about at this an 'document architecture' level rather than patching issues on a per case basis as we find them. But I guess people can judge for themselves whether the fact that Atom might corrupt carried content is an issue.
Do we say that elements put in atom:content SHOULD use absolute URIs. We probably, uhm, should :-)
I agree, but you don't sound sure :-)
cheers Bill
