W3C home > Mailing lists > Public > uri@w3.org > April 2003

Re: Resources and URIs

From: John Cowan <cowan@mercury.ccil.org>
Date: Tue, 29 Apr 2003 08:03:00 -0400
To: Patrick.Stickler@nokia.com
Cc: phayes@ai.uwf.edu, gk@ninebynine.org, uri@w3.org
Message-ID: <20030429120300.GR12348@ccil.org>

Patrick.Stickler@nokia.com scripsit:

> If Tim says that http://www.w3.org/Consortium/ denotes the W3C,
> then that's (probably) what it denotes.

Fair enough.

> If you do an HTTP GET on that URI, and you get a document titled
> "About the W3C"

Well, actually, a GET never returns a "document" as I use that term.
What it returns is a MIME entity-body: that is, a sequence of octets,
and a media type, and possibly some other metadata.

Documents contain words, phrases, titles, and are independent of
MIME entity-bodies and can be instantiated in ways that don't involve
them (most obviously on flattened dead trees, but clay tablets work too).

> then that is a "representation" of the W3C

By fiat, yes.  But what then is the URI of the document, of which the
relevant MIME entity-body is also a representation?
This document *describes* the W3C rather than *being* the W3C.

A MIME entity-body of type image/gif may be a representation of Shakespeare,
but if so, only by being a representation of a *picture* of Shakespeare,
(or possibly a representation in GIF form of a textual *description* of S.)
that is, a representation of a representation.

> is some other, distinct resource. Ideally, the HTTP server should 
> tell you the URI of that other resource in its response header,
> but not all servers are so well behaved.

To say the least.

> It may be that you are not told what URI denotes what you have
> gotten from the HTTP server, but that doesn't mean that either
> (a) it has no known URI to denote it, or (b) that the URI given
> to the server is overloaded to denote both the W3C and the document
> (which would be very bad indeed).

It doesn't absolutely *compel* either of those things, no.  But they
seem quite likely to happen nonetheless unless measures are taken to
stop them.  There is, Heaven knows, no shortage of available URIs.
What there is, is a lack of will to assign them appropriately.

It was impossible to inveigle           John Cowan <jcowan@reutershealth.com>
Georg Wilhelm Friedrich Hegel           http://www.ccil.org/~cowan
Into offering the slightest apology     http://www.reutershealth.com
For his Phenomenology.                      --W. H. Auden, from "People" (1953)
Received on Tuesday, 29 April 2003 08:03:07 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:25:05 UTC