W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > March 2003

Re: Proposal to close reagle-01 reagle-02

From: Dave Beckett <dave.beckett@bristol.ac.uk>
Date: Mon, 31 Mar 2003 12:29:18 +0100
To: "Patrick.Stickler" <Patrick.Stickler@nokia.com>
cc: jjc <jjc@hplb.hpl.hp.com>, w3c-rdfcore-wg <w3c-rdfcore-wg@w3.org>
Message-ID: <7331.1049110158@hoth.ilrt.bris.ac.uk>

>>>"Patrick.Stickler" said:
> Best to have wording that states clearly that whatever lexical
> form exists in the RDF/XML serialization, it is mapped to
> an exclusive XML canonical lexical form by the parser and
> the graph only contains canonical lexical forms for XML literals
> (unlike other literals, for which lexical forms are just passed
> through unmodified).

<snip>

> 
> I think there needs to be a simple, short, but clear comment
> that canonicalization is done by the parser, and that XML
> literals in RDF/XML serializations need not be canonicalized.

We can't speak in terms of "the parser", this is a mapping between
forms.  So if we want the mapping to include XML canonicalization, it
has to go in the mapping section of (in this case) the appropriate
RDF/XML grammar action.

> Having the discussion of canonicalization in the section on
> grammar productions for RDF/XML strongly suggests to me that
> it is relevant to the RDF/XML. ...

It is relevant to implementing the mapping of RDF/XML to the RDF graph.

> ... Perhaps it simply should be
> moved out of that section entirely, and added in an appendix
> or elsewhere, where it is not right in the heart of the 
> RDF/XML grammar specification, since it has nothing to do with
> the RDF/XML grammar itself. 

No.  It belongs where the mapping is defined.  Plus it has been
changed to be normative (MUST), not as present recommended (SHOULD).

> ... In fact, I would almost go so far
> as to suggest that the syntax spec should simply reference
> the concepts spec -- since canonicalization is strictly a matter
> of the graph syntax, not the RDF/XML syntax.

Going
   RDF/XML -> uncanonicalized XML syntax -> canonicalized XML
              in graph syntax               in graph syntax 

seems wrong.  We've agreed that the graph has canonicalized XML,
therefore the documents that map to it must describe how to
get there.

> 
> (but a short note clarifying no canonicalization for RDF/XML
> is required would be enough to satisfy me ;-)

<pause for Jeremy to write it> :)

Dave
Received on Monday, 31 March 2003 06:30:34 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:56:16 EDT