RE: Yet more XSD/RDF

on goofiness

>     # dt:QName # omitted due to ambiguity

needs in scope namespaces

>     # dt:NOTATION # omitted; DTD support not merited

needs user defined types (merely an abstract datatype at this level)

>     # NMTOKEN

OK - but really corresponds to an XML DTD concept.
>     # NMTOKENS

ditto

>     # ID
>     # IDREF
Both duplicates of NCName so no need to use them, (except for XML DTD
compatibility issues that I haven't yet come across).

>     # IDREFS
(There isn't an NCNames built-in ...)

>     # ENTITY
>     # ENTITIES
Need surrounding document

I am largely in agreement with Dan. When we recommend the use of XSD, we
probably should recommend a subset. I am a little worried about the equality
issues on the datetime objects .... Mainly XSD jsut defers to an ISO
standard for those.

Jeremy

Received on Tuesday, 26 November 2002 10:41:10 UTC