- From: John Aldridge <john.aldridge@informatix.co.uk>
- Date: Thu, 07 Sep 2000 17:59:27 +0100
- To: "Henrik Frystyk Nielsen" <frystyk@microsoft.com>, <michaelm@netsol.com>
- Cc: "Simon St.Laurent" <simonstl@simonstl.com>, <XML-uri@w3.org>
At 09:05 07/09/00 -0700, Henrik Frystyk Nielsen wrote: > > Do you go further and say that XSLT (which matches elements > > and attributes > > based on their namespaces being the same) must therefore use this > > definition of equivalence? I.e. two QNames match if their > > namespace URIs > > are byte-for-byte identical and their local parts are the same? > > > > Again, I agree, but Henrik Frystyk Nielsen (at least) has > > expressed the > > view that this is not what the relevant RFCs/RECs say (or > > what they should > > say). > >That is over-simplifying what I stated - in [1], I said that > > 1) The generator of a name has the responsiblity to > know the semantics of the URI space that she is using > > 2) It is sufficient if a basic consumer only uses > octet-by-octet comparison taking into account relative > URIs. However, it is also fully acceptable for the > consumer to know about special normalization rules of a > URI space and apply those if so desired. > >which I further elaborated on in [2] mentioning the responsibilities of >the party generating the document and the party consuming the document. > >Henrik > >[1] http://lists.w3.org/Archives/Public/xml-uri/2000Jun/0721.html >[2] http://lists.w3.org/Archives/Public/xml-uri/2000Jun/0739.html I was referring to the following exchange Me: >You are saying, then, that the XPath REC should permit, _but not require_, >XPath implementations (including XSLT) to normalise NSURIs in a scheme >dependent manner when comparing names? Henrik: >You can not require them to do that just as well as you can not require >them to know about a particular namespace. Me: >This can clearly result in a given XML source/XSLT stylesheet combination >producing different results depending on the normalisations performed by >the particular XPath implementation. Henrik: >If you don't want that possibility then don't use a URI space that has >this behavior - it is up to you, see [1]. -- Cheers, John
Received on Thursday, 7 September 2000 12:59:37 UTC