RE: anyAttribute via attributeGroup refs problem

> Thank you! Is the "surprising rule" you refer to the one under 
> 
> §3.4.2-"XML Representation Summary: complexType Element 
> Information Item"-"Complex Type Definition with simple 
> content Schema Component"-2.2.1 

Yes.
> 
> (don't know if there is a more concise way of referring to 
> that specific document location)? 

The structure has been improved in XSD 1.1 to make referencing much easier.
> 
> Yes, I would have expected a union in the same way single 
> attribute definitions form a union, since my personal view up 
> to now on "anyAttribute" was that it's just a slightly more 
> generalized "attribute". Obviously, it's not, but treated 
> quite differently. Haven't thought on this long enough, but: 
> are there use cases for namespace intersections? 

I personally can't see any logic behind the current design, but it appears
there are reasonable people who can.

Michael Kay
http://www.saxonica.com/