new test case for broken transforms (note I would abstain on this)

See
http://www.w3.org/2001/sw/grddl-wg/td/pendinglist

specifically the file
http://www.w3.org/2001/sw/grddl-wg/td/withErrors
that includes one good and lots of bad transforms.

Personally, I think even this test case is over-specifying error 
handling for GRDDL. Since the Jena GRDDL Reader passes, (to my surprise) 
I don't have sufficient grounds for voting against this test, but I 
would abstain. (It means somebody else also needs to propose this test, 
since I don't believe I can abstain on something I propose)

Procedurally, if it makes the commentator happy, and the implementations 
pass, it makes sense to approve it.

Jeremy


-- 
Hewlett-Packard Limited
registered Office: Cain Road, Bracknell, Berks RG12 1HN
Registered No: 690597 England