W3C home > Mailing lists > Public > www-tag@w3.org > January 2003

Re: Clarifying what a URL identifies (Four Uses of a URL)

From: Sandro Hawke <sandro@w3.org>
Date: Tue, 21 Jan 2003 17:39:22 -0500
Message-Id: <200301212239.h0LMdM015494@wadimousa.hawke.org>
To: Tim Bray <tbray@textuality.com>
cc: Michael Mealling <michael@neonym.net>, David Booth <dbooth@w3.org>, www-tag@w3.org


[I sure hope the people reading www-tag have "kill-thread" functions
in their mailers.]

> Sandro Hawke wrote:
> 
> > URIs are strings which are used for different things in different
> > situations, in a manner controlled by the semantics of the situation.
> 
> On the other hand, using the same URI to mean different things is a Bad 
> Thing and leads to confusion and misbehavior not only at the Semantic 
> Web level but in terms of general human utility. 

I see people using URIs in at least two completely different ways all
the time.  For example:

<p>If you work for a member-company of the <a
href="http://www.w3.org/">W3C</a> you may find you want to read the <a
href="http://www.w3.org/">W3C Front Page</a> at least once a week.
</p>

In that example the same URI is freely used to identify both the W3C
and the W3C's front page.   

In a more formal situation, I see "http://www.w3.org/2001/XMLSchema"
being used to mean quite different things in different situations.
If there's exactly one resource identified by that URI, can you tell
me some things about that resource?

> When you say that 
> http://example.com/moby represents Moby Dick you need to be clear 
> whether you mean Melville's novel in the abstract, some particular copy 
> on a shelf, the online Gutenberg text, a record in a particular library 
> catalog, or a fictional cetacean.

My claim is that actually using a URI to formally identify some real
or conceptual object requires additionally specifying a URI mapping
function.  The Web architecture doesn't provide one.  To say broadly
that the string "http://example.com/moby" identifies, in all contexts,
a particular book on my bookshelf, ... seems pretty nonsensical when
you imagine it as a working web address.  To talk about what can and
should happen when you use that string as the URI in HTTP operations
-- yes, that's a good and reasonable thing.  To go beyond that to talk
about what one, single thing the URI "identifies everywhere" seems
like a mistake.

>    The
> > notion that there is or should be exactly one conceptual thing
> > corresponding in all situations in some standand way to each URI (that
> > the URI is a logical symbol with a single denotation) is a fallacy
> > which causes the httpRange-14 rat-hole.  
> 
> It's a formalism.  The Web Architecture has a formalism called a 
> "Resource" which is the one thing that corresponds to each URI.  This is 
> fact of life and isn't going away.  Deal with it.  The formalism is 
> plenty complete and consistent enough to support the construction and 
> efficient operation of the Web.  -Tim

It's hard to accept the idea that there is one thing identified by
each URI when no one can tell me anything about that thing, except in
trivial, made-up cases (like DanC's "this is a car" page).  What thing
is, as far as you can tell, identified by
   http://www.w3.org/Protocols/rfc2616/rfc2616-sec10.html
   http://www.uroulette.com/      [ used to pick the others ]
   http://www.avianavenue.com/
   http://ont.net/karate          [ 404, but don't let that stop you ]
   http://www.interactivemarketer.com/
...?

I recognize that in HTTP itself, it's pretty hard to explain "301
Moved Permanently" without saying what moved.  What would it tell you
about XML Schema Datatypes if you got a 301 when you did an HTTP GET
on http://www.w3.org/2001/XMLSchema?  Hmm.  It doesn't actually mean
XML Datatypes moved, but that a source of information about them
moved.  (... and thus I think HTTP "resources" are best seen as
information access locations in mediated shared memory, but that's
getting into the meat of httpRange-14.)  My point, and David Booth's
point at the start of this thread, is that it's reasonable and
appropriate to use HTTP URIs in other contexts (like XML namespace
names, or RDF) to identify other things than they identify in HTTP
GET.)

If the existing formalism could handle defining access rights on a web
page with the same URI as an RDF predicate, I'd never have subscribed
to www-tag [ and we'd all be a lot happier :-) ].   

    -- sandro
Received on Tuesday, 21 January 2003 17:41:48 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:15 GMT