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

Re: fragment identifiers

From: Tim Bray <tbray@textuality.com>
Date: Mon, 29 Jul 2002 11:32:48 -0700
Message-ID: <3D458A50.2070109@textuality.com>
To: Dan Connolly <connolly@w3.org>
Cc: Joshua Allen <joshuaa@microsoft.com>, "Roy T. Fielding" <fielding@apache.org>, "Sean B. Palmer" <sean@mysterylights.com>, www-tag@w3.org

Dan Connolly wrote:

> Whether http URIs can identify cars and such or not isn't observable
> from the HTTP protocol itself.

Indeed.  And given that the server can decide to send back whatever it 
$#^!%#&!@ well wants to (and I've written a lot of server code that uses 
some *very* non-obvious inputs in deciding what to emit in response to a 
URI)...

And given that outside of RDF, the Web architecture has no built-in way 
at all to talk about what a resource "is"...

the only world-view that makes sense to me is what my section 1.1 draft 
says, a resource is just whatever it happens to be, you can get 
representations of it by asking (sometimes), and you can make assertions 
about it regardless of whatever it "is", whatever "is" means, whatever 
"means" means, and neatly dodge all the metaphysical bushwah.

And the more I think about it, the less I can believe in any special 
distinguished status or limitations for HTTP URIs, aside from noting the 
obvious fact that they have the useful advantage of a well-defined 
builtin protocol for requesting/receiving representations.  -Tim
Received on Monday, 29 July 2002 14:33:46 GMT

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