draft XML query results format spec

I attach two possible result formats that vary in one design choice -
  1) variable names are element names and hence the XML is schemaless
  2) variable names are content and the XML has a tight schema.

I presently prefer the former result1.xml since it seems more natural
XML to use the application terms - the variable names - as element
names.  The result XPath, XSLT etc. would read better I think.

result2.xml comes with a RelaxNG compact, RelaxNG and W3C XML Schemas
that all validate the document.  It's got inline schemaLocation as an
example, but that is optional if I understand WXS correctly.

Dave