W3C home > Mailing lists > Public > public-rdf-in-xhtml-tf@w3.org > May 2004

Re: Other link relations

From: Danny Ayers <danny666@virgilio.it>
Date: Mon, 24 May 2004 11:08:50 +0200
Message-ID: <40B1BBA2.1020204@virgilio.it>
To: Martin Duerst <duerst@w3.org>
Cc: Norman Walsh <ndw@nwalsh.com>, atom-syntax@imc.org, public-rdf-in-xhtml-tf@w3.org

Martin Duerst wrote:

>
> At 13:54 04/05/21 -0400, Norman Walsh wrote:
>
>> / Bjoern Hoehrmann <derhoermi@gmx.net> was heard to say:
>
>
>> | Well, QNames, obviously. The next XHTML 2.0 Working Draft will likely
>> | incorporate <http://www.w3.org/MarkUp/2004/02/xhtml-rdf.html> which
>> | proposes to use QNames in the rel/rev attributes of the XHTML 2.0 link
>> | element. Whatever syntax it will be, it seems undesirable to have more
>> | than one across these formats.
>>
>> Indeed. If HTML is going to put QNames in there then we should too.
>
>
> Or better yet, given the discussion here just a few days ago,
> tell them (XHTML) that they should try to avoid QNames, and
> that better solutions are available. QNames in attributes are
> a bad idea, and using them in XHTML doesn't make them better.

Quite. I'm also not sure why the XHTML 2.0 group might incorporate 
Birbeck's RDF in XHTML, as I believe it works on top of the current 
XHTML 2.0 spec so there's no dependency the other way. I suspect it 
would be a little premature to adopt this approach in any case - the 
proposal has gone down very well with a lot of RDF people, as this has 
been something we've  needed for years. But I think it's likely to hit 
resistance when presented to regular XML folks who've known problems 
with QNames in attributes before (to the extent of calling them evil! 
[1]) , and coincidentally have had issues with RDF syntax in the past. I 
think he's probably on the right lines, but I think it would be reckless 
to jump on the first version that vaguely works as the final answer. As 
far as I can see there are other issues - something that would work 
fully with XHTML 1.0 would be nice for a start, and the relationship 
between the text content and the metadata is limited. But I'm sure with 
a few short cycles the RDF in XHTML proposal could be made palatable for 
a considerably larger number of people.

Cheers,
Danny.

[1] http://www.mnot.net/blog/2003/12/06/qnames_are_evil

-- 

Raw
http://dannyayers.com
Received on Monday, 24 May 2004 05:14:32 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:14:59 GMT