W3C home > Mailing lists > Public > www-html-editor@w3.org > April to June 1999

Re: Liaison statement on fragment identifiers from Linking WG

From: Bill Smith <bill.smith@sun.com>
Date: Mon, 17 May 1999 06:06:30 -0700
Message-Id: <>
To: Tim Bray <tbray@textuality.com>, Steven Pemberton <Steven.Pemberton@cwi.nl>
Cc: w3c-xml-cg@w3.org, w3c-html-wg@w3.org, www-html-editor@w3.org, w3c-xml-linking-wg@w3.org
The discussion on this topic is encouraging.

 At 03:56 PM 5/13/99 -0700, Tim Bray wrote:
>Yes - when I first read the XHTML response, I was irritated first
>at the lack of dialogue, second at the fact that apparently they'd
>discarded the linking WG input.  Having read it more carefully, it 
>seems to me that they've accepted the *important* part of the input, 
>which was to adopt the semantic that in XHTML, foo#bar means whatever
>element has id="bar" - and to insert an advisory notice that
>this won't work in old browsers, so if you care about them,
>put name="bar" on there as well.  So at least you get (nominally)
>the same behavior with XML, HTML, and RDF, assuming that people
>follow the recommendation.
>I can also see arguments why it might not be worth the pain
>to make name= into an ID attribute when HTML already has a 
>perfectly good ID attribute that is in fact being used in DHTML.
>I agree with Bill that a lot of people are going to ignore the
>recommendation and just go on using "name", and will be surprised
>and upset that this doesn't work when you serve the doc as
>text/xml.  Speaking for myself, I honestly can't really predict
>whether this will be a problem - there is a good chance that anyone
>who cares enough to issue a text/xml media type will take the 
>trouble to get the IDs in order.  
>Steven, is my understanding correct?  Bill, assuming that it is,
>am I missing something? -Tim

You've summarzied what I see as the problem and my understanding of the
HTML WG's decision aligns with yours.

I'm (a bit) more willing to predict that market forces are unrelenting and
that if vendors suspect that users will ignore the recommendation and
admonition on NAME and ID, they will build their products to accept either.
I think this will have unfortunate consequences for XML since the
application semantic of HTML (an application of SGML) will transfer to the
generic XML (a meta language). 

My concern could be adressed by XHTML avoiding the mime-type XML and
instead registering a new (application-specific) mime-type XHTML. This
would cleanly divide the application- pecific from the application generic
and that's at the core of my conccern.
Received on Monday, 17 May 1999 09:43:49 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:08:21 UTC