Re: Using XPointer with HTML

"Charles McCathieNevile" <charles@w3.org>

> So one approach the RE group could take is to define a document
namespace
> which is in fact defined as the Tidied version of something, where
there is a
> reulst defined for when Tidy just gives up.

Making something dependant on tidy is not an option AFAICT, and isn't
necessary in any case. What we actually need is to stop thinking of the
context as a URI-fragment. We can use a standalone property that we can
define the semantics of without all these fragment references getting in
the way.  XPointer isn't even sufficient for application/xhtml+xml stuff
as they're not robust over changes in the document - so we still need all
the contextual stuff to identify a document.

Jim.

Received on Wednesday, 10 April 2002 12:03:40 UTC