Re: xmlse:foo slightly substantive comment on CURIEs (both in stand alone document, and in RDFa PR) (PR#8053)

Jeremy,

We weren't ignoring you, I promise!  However, it took a while to get to 
your comment in the stack.  I have carefully reviewed your comment.  My 
position is that we cannot accept your changes "as is".  I do agree that 
there is a potential issue here, and would like to add a note to the 
conformance section of CURIEs that describes the issue.  My rationale is 
that, since we are relying upon "xmlns" to define prefixes in XML-based 
languages, and since we have said that a prefix name is restricted to 
NCName, we have to accept the restrictions as stated in the XML 
Namespaces specification. Obviously with other languages, or if there 
are other prefix definition mechanisms, it would be possible to define a 
prefix name that is inconsistent with the requirements of XML 
Namespaces.  Consequently, I agree that it behooves us to point out this 
potential conflict to language designers.  Here is what I propose:

Note: The XML Namespaces specification states that prefix names are not 
permitted to begin with the characters 'xml' (reference).  While this 
specification does not impose such a restriction, when CURIEs are used 
in an XML language this restriction is effectively inherited from XML 
Namespaces.  If such a language defines an additional mechanism for 
defining prefixes, that mechanism SHOULD impose a similar restriction so 
there is no possibility of conflict between prefixes defined using the 
two mechanisms.

The working group has not yet reviewed this proposal.  I expect them to 
do so at their f2f meeting this coming week.  Please let us know if you 
agree with this suggested wording, or if there is some alternate wording 
that would satisfy you WITHOUT removing the requirement that we use 
'xmlns' syntax - we have many other constituents who view that as an 
absolute requirement for CURIEs.

jeremy@topquadrant.com wrote:
> Please remove www-html-editor from the address list if doing a reply-to-all on a WG list.
>
> Apologies that this is a late comment on both specifications.
> It arises from work we were doing yesterday.
>
> This is a comment on:
> http://www.w3.org/TR/2008/PR-rdfa-syntax-20080904/
> and
> http://www.w3.org/TR/2008/WD-curie-20080506/
> made on behalf of TopQuadrant.
>
> Contents:
>   Comment
>   Related Draft Comments
>
> Comment:
> ***
> The string xmlse:foo is (arguably) a CURIE according to the RDFa PR
> and is not a CURIE according to the CURIE Last Call WD.
>
> See the text:
> [[
> prefix values MUST be able to be defined using the 'xmlns:' syntax specified in [XMLNAMES].
> ]]
> in the latter, which is not in the former.
>
> Note both documents have the rule:
> [[
> prefix      :=   NCName
> ]]
> and NCName is linked to
> http://www.w3.org/TR/1999/REC-xml-names-19990114/#NT-NCName
>
> which says 
> [[
> Prefixes beginning with the three-letter sequence x, m, l, in any case combination, are reserved for use by XML and XML-related specifications.
> ]]
>
> but since xml:lang is clearly intended to match the QName construct
> http://www.w3.org/TR/1999/REC-xml-names-19990114/#NT-QName
> it is, in my view, a misreading of Namespaces in XML to have NCName not matching the string "xmlse"
>
> My preferred fix would be:
> Delete the offending text from CURIE.
> Add the following clarification to both documents:
> [[
> Note: A CURIE prefix value may case-insensitively begin in the string "xml", but then it is not permitted to be used as a prefix in an xmlns declaration in an XML document. 
> ]] 
> ***
>
> Related Draft Comments:
>
> The background to this observation is further explained in:
> http://lists.w3.org/Archives/Public/public-swd-wg/2008Sep/0028
> which is in turn a follow up to
> http://lists.w3.org/Archives/Public/public-swd-wg/2008Sep/0018
> Note msg 0028 took a more positive spin on this observation, but on further reflection this morning led me to this more substantive issue. 
> Since this comment arose during my review for the SWD WG, there may be further comment from that WG (e.g. either an endorsement or an explicit non-endorsement).
>
> Jeremy
>
>
>
>
>   

-- 
Shane P. McCarron                          Phone: +1 763 786-8160 x120
Managing Director                            Fax: +1 763 786-8180
ApTest Minnesota                            Inet: shane@aptest.com

Received on Saturday, 18 October 2008 15:28:13 UTC