comment on xslt_literal_result test

Concerning
http://www.w3.org/2001/sw/grddl-wg/td/litres.xml
with mimetype application/xml

I think the GRDDL WG should duck this one. It is not a GRDDL document, 
and so the WG does not need to decide on it. Any particular behaviour 
for GRDDL clients, fixed by WG decision, concerning an ill-conceived 
document, puts a burden on implementers. I would find it acceptable to 
simply label this as an error test case, i.e. a test:NegativeParserTest.
For instance, my implementation is likely to try and read it as RDF/XML 
and will both report errors, and give some triples. However, it doesn't 
seem unreasonable to apply XSLT and then read it as RDF/XML, but that 
isn't GRDDL - GRDDL isn't triggered by xsl:version, it is triggered by 
dataview:transformation ...

(The loop example though is a good test!)

Jeremy