W3C home > Mailing lists > Public > www-tag@w3.org > July 2002

Re: [Finding] Using Qualified Names (QNames) as Identifiers in Co ntent

From: Joseph Reagle <reagle@w3.org>
Date: Fri, 26 Jul 2002 13:52:35 -0400
To: "Hallam-Baker, Phillip" <pbaker@verisign.com>, "Ian B. Jacobs" <ij@w3.org>
Cc: www-tag@w3.org, www-xkms@w3.org
Message-Id: <200207261352.35640.reagle@w3.org>

Phil,

I don't think we're on the same page.

On Friday 26 July 2002 01:32 pm, Hallam-Baker, Phillip wrote:
> I do not believe that the UDDI schema centric C18N is relevant to the
> QNAME point.

My point is simply, "look at some of the things that become necessary when 
you have QNames in attribute values." in schema-centric C14N they had to 
introduce the concept of EmbeddedLanguages and take care of them: not 
pretty.

> Finaly I read the architecture board's statement quite differently, they
> state that they are not going to consider change on that particular topic
> as the legacy usage is significant. I read that as stating that they have
> not engaged in a comprehensive examination of the issue because it is
> moot.

Yes, there are legacy issues. Again, my concern is simple, just because it's 
done, doesn't mean everyone should do it, particularly without 
understanding the issues you're inviting upon yourself. This strikes me as 
similar to the relative URIs in namespaces issue. It wasn't precluded, it 
created lots of headaches, but (in this case) there wasn't that much legacy 
so it was deprecated [1]. In the QName case, there's too much usage to 
deprecate it, but I still don't think it should be encouraged. But that's 
just IMHO which isn't binding on anyone <smile/>.

[1] http://lists.w3.org/Archives/Public/xml-uri/2000Sep/0083.html
Received on Friday, 26 July 2002 13:53:10 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:10 GMT