ISSUE-147 Our DatatypeUnionOf is not XSD's (semantically)?

Issue 147 proposes union datatypes (i.e., integer OR string). XML  
schema already has a sort of union datatype:
	http://www.w3.org/TR/xmlschema-2/#dt-union

We're not going to use the XSD union (at least its semantics).

(Everything I say is a bit provisional the moment as I've not  
spelunked in any formal semantics.)

Intuitively, XML Schema union vs what we want is a bit like union in  
SPARQL vs. unionOf in OWL. I hope. I was looking for backup on this  
and am not immediately finding it. It depends on the problem one is  
considering (i.e., recognition vs. schema consistency checking).

So I was bit glib in my assertion that one *couldn't* reuse union,  
but maybe it's just easier not to use it?

Cheers,
Bijan.