W3C home > Mailing lists > Public > public-rdf-in-xhtml-tf@w3.org > February 2009

Re: Named Graphs in RDFa

From: Shane McCarron <shane@aptest.com>
Date: Sun, 01 Feb 2009 09:40:40 -0600
Message-ID: <4985C278.2010007@aptest.com>
To: Kjetil Kjernsmo <kjetil@kjernsmo.net>
CC: RDFa <public-rdf-in-xhtml-tf@w3.org>

I am sure you are aware of this, but just in case...

XHTML permits @xhtml:id on everything.  In XML, there can ONLY be one 
attribute of type ID on any given element.  Therefore, attempting to use 
@xml:id would by definition conflict with @xhtml:id in at least some 
cases.  Moreover, in other XHTML family specifications @xml:id is used 
interchangeably with @xhtml:id (you can put one or the other on any 
element).  This will become the default case for all elements with the 
advent of XHTML 2.  Consequently, I am reluctant to agree with you that 
@xml:id can be used for the purpose you have identified.

Kjetil Kjernsmo wrote:
> On Sunday 01 February 2009, Michael Hausenblas wrote:
>> You rightly conclude that we very likely need a new attribute but
>> then state that 'this document does not define what that attribute is
>> called'. Hm. I think this is too much relying on a convergent market.
>> I can't see how this should scale.
> Yup, this is also my main concern with the current draft. We've 
> discussed it at some length, then we figured that we have a 
> philosophical disagreement here, and that we'd better take it to the 
> list. :-) Then, I asked Toby to hold off until I had written up my RDFa 
> Templates idea, which is the motivating use case, and that's now here:
> http://www.kjetil.kjernsmo.net/software/rat/
> As you can see in this draft, I use xml:id for the named graph. Toby has 
> several good arguments against that, but I kept it as a starting point 
> for discussion. 
> I think the very reason to write a spec (rather than just API 
> documentation) is to avoid that people need to read documentation that 
> I write, instead they just need to read the spec, and it will just 
> work. So, a private agreement, I feel doesn't achieve that goal, though 
> it may be useful as an extension mechanism.
> I'm attaching a diff I wrote to a Jan 19th draft of the spec. The 
> essence of this was to define @xml:id as the simplest way of creating a 
> graph name, I wrote:
> The simplest way to markup a graph name is to use the xml:id attribute. 
> The value space of this attribute is full URIs and the lexical space is 
> similar to xhtml:id. Therefore, if the base URI of the document is 
> http://example.com/document then the attribute xml:id="foo" represents 
> the URI http://example.com/document#foo. Implementations of this 
> specification must accept graph names set by this attribute by default.
> So, my idea was that users would always be able to identify a graph 
> using xml:id, but could also use a more flexible graph naming scheme by 
> private agreement.
> Cheers,
> Kjetil

Shane P. McCarron                          Phone: +1 763 786-8160 x120
Managing Director                            Fax: +1 763 786-8180
ApTest Minnesota                            Inet: shane@aptest.com
Received on Sunday, 1 February 2009 15:42:09 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:50:30 UTC