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

Re: "absolute URI reference" considered awkward (and in one case, overly constraining)

From: Roy T. Fielding <fielding@apache.org>
Date: Sat, 7 Sep 2002 12:07:25 +0200
Cc: Tim Berners-Lee <timbl@w3.org>, www-tag@w3.org
To: Tim Bray <tbray@textuality.com>
Message-Id: <9C04C65E-C249-11D6-98A4-000393753936@apache.org>

> Trouble is the RFC process (as I understand it) is pretty 
> nondeterministic, and it's gonna REALLY SUCK if the W3C and the IETF mean 
> different things when they say "URI".  It would be so much nicer if we 
> could have some comfort that there will be consistency.

The RFC process is a lot more deterministic than the W3C process.
Those who show up with an opinion AND an implementation determine
what goes in the standard.  People who do not show up, or who just
have an opinion, have to depend on others.

The issue is well-known.  The problem has been that some folks don't
consider the fragment to be part of the URI, and for good reason:
outside of the href element, it has no relationship with the URI.
They aren't processed together, they aren't dependent on one another,
and they don't communicate the same concepts.  The only way to
justify them being joined together is to first define what a
resource is and then define a fragment as an identifier of a
resource, which is not a technical argument.  So, the majority of
opinions will carry the issue and people who do not express their
opinions on uri@w3.org will not be heard.

....Roy
Received on Saturday, 7 September 2002 06:10:45 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:11 GMT