- From: Jeremy Carroll <jjc@hpl.hp.com>
- Date: Tue, 24 Apr 2007 19:27:09 +0100
- To: GRDDL Working Group <public-grddl-wg@w3.org>
I propose to approve #httpHeaders as an informative test I propose to approve #embedded-rdf7 #embedded-rdf7-alt #loopx #loopx1 #loopx2 #loopx3 Noting that: a) #embedded-rdf7 shows a possible behaviour of an XSLT 1.0 system b) #embedded-rdf7-alt shows the expected behaviour of an XSLT 2.0 system and c) With an XSLT 1.0 system, the instruction <xsl:message terminate="yes"> executes, and so the GRDDL spec does not specify correct processing. There is an addiitonal transform (empty.xsl) which ensures there is a GRDDL result, but we may wish to reject these two tests. I verified this behaviour with Jena before adding the XSLT 2.0 support to the library function. d) #loopx and friends, considers a document that is it's own namespace document, and has GRDDL results that depend on its own GRDDL results. In earlier discussion we were minded that the 'SHOULD' in the specification permitted implementations to not find the logically correct #loopx3 answer. Hence the test shows all possible GRDDL results. Jeremy -- Hewlett-Packard Limited registered Office: Cain Road, Bracknell, Berks RG12 1HN Registered No: 690597 England
Received on Tuesday, 24 April 2007 18:27:50 UTC