assigning semantics to XML, Re: [xml-dev] Re: URIs, concrete (was Re: [xml-dev] Un-ask the question)

Walter Perry wrote:
> 
> But my question was whether (and if so, why) we could reasonably expect a
> discernible difference in the semantic outcome from processing a particular
> instance of a single given attribute in no namespace as if it were 'in the
> namespace' of the element which it modifies.
> 
As a followup on my last response:

One may assign a different range of interpretations to a given namespace qualified element with an attribute depending on whether the attribute is:

a) not qualified (no prefix)
b) qualified with the namespace name of the element
c) qualified with any other individual namespace name

This does not mean that one is _required_ to assign _a different_ interpretation to each of the above cases, rather it means that one _is not required_ to assign _the same_ interpretation to each of the above cases. The distinction is important.

Jonathan