W3C home > Mailing lists > Public > www-rdf-interest@w3.org > June 2001

Re: URI etymology

From: Sean B. Palmer <sean@mysterylights.com>
Date: Tue, 12 Jun 2001 16:24:19 +0100
Message-ID: <022601c0f353$da3c6480$9dda93c3@z5n9x1>
To: "Pierre-Antoine CHAMPIN" <champin@bat710.univ-lyon1.fr>
Cc: "Sampo Syreeni" <decoy@iki.fi>, <www-rdf-interest@w3.org>
> >    :x a :Picture;
> >       :returnedFrom <http://example.org/img>;
> >       :date "2001-03-17";
> >       :containsARepOf [ a :Person; :name :y ] .
> Pardon me, Sean, but I don't understand how your example
> respond to Sampo's problem. [...] there is a problem when
> the *same* URI identifies two different things, [...]

Where in my example is a URI being used to represent two different
things? The "problem" was that what you get back when you dereference
a URL is a resource in its own right, and that any interpretations
thereof are also resources. So just give new URIs to those resources;
there isn't a conflict because the URL only identifies one thing - the
resource that you get a representation of on dereferencing.

On the Semantic Web, you can cast a further contextual light upon this
resource. As all uses of URIs are purely contextual anyway, this is
great. For example, if I said that your homepage was actually a
rabbit, would that suddenly break it because I'm using it for two
diffferent things? I could do the same thing with a URN or any other
URI that you care to mention. There are no roots of the ontology tree
in the Semantic Web, and yet people seem to want to cast permanent
restrictions on how a URI can be used. This is contrary to the entire
SW principle of being able to say anything about anything", and is in
my opinion the bigest problem that the Semantic Web has to face.

If you want to somehow build a trust mechanism on top of this free
contextualization, I suggest that you look into the various digital
signatures on top of RDF developments that are going on.

Another point to mention is that many of these so called "problems"
that people keep talking about are simply theoretical ratholes (such
as, "what happens when you have a representation of a resource from a
URL which is also being used in some context in the Semantic Web?"),
and they do not relate to the practical tools available today.

The "FragId definition is MIME related" is the only issue where I
would concede that there is a problem... Oh and the "{abc}xyz {abcx}yz
both concatenate to the same URI" issue. Both of these have been
debated to pieces in the RDF IG and RDF Logic mailing lists, and I
have nothing further to add to them.

Kindest Regards,
Sean B. Palmer
@prefix : <http://webns.net/roughterms/> .
:Sean :hasHomepage <http://purl.org/net/sbp/> .
Received on Tuesday, 12 June 2001 11:23:51 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:07:36 UTC