- From: Elliotte Harold <elharo@metalab.unc.edu>
- Date: Thu, 16 Dec 2004 10:50:38 -0500
- To: Michael Kay <mike@saxonica.com>
- CC: 'Rich Salz' <rsalz@datapower.com>, xml-dev@lists.xml.org, public-qt-comments@w3.org
Michael Kay wrote: > The XQuery specification, incidentally, defines a mapping of error QNames to > error URIs. It's different from this one, of course. The error URIs in the XQuery spec look like this: http://www.w3.org/2004/10/xqt-errors#XPST0017 Something similar has been tried before in XML digital signatures and RDF. For reasons elucidated in the draft proposal, this format is not suitable for all QNames and does not work as a generic mechanism for encoding Qnames as URIs. The XQuery format also seem to suffer from the disease that everything must be an http URI even when, as in this case, the URI is not actually resolvable. (http://www.w3.org/2004/10/xqt-errors is a real page but the fragment ID doesn't point to anything.) I agree with most of your concerns about the details of the proposed qname URN scheme (IRIs vs. URIs, percent encoding, etc.) but at a fundamental level I think the proposed scheme is better than what XQuery is proposing. I would suggest that XQuery adopt the new qname URN scheme rather than using http URIs to identify error codes. In fact, I'll cc this to the XQuery working group to register it as a formal comment. -- Elliotte Rusty Harold elharo@metalab.unc.edu XML in a Nutshell 3rd Edition Just Published! http://www.cafeconleche.org/books/xian3/ http://www.amazon.com/exec/obidos/ISBN=0596007647/cafeaulaitA/ref=nosim
Received on Thursday, 16 December 2004 15:50:42 UTC