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

Re: My action item on Moby Dec, issue 14, etc

From: Jonathan Borden <jonathan@openhealth.org>
Date: Fri, 20 Sep 2002 17:42:30 -0400
Message-ID: <001d01c260ee$a08a8230$7c674544@ne.mediaone.net>
To: Bill de hÓra <dehora@eircom.net>
Cc: <www-tag@w3.org>

Bill de hÓra wrote:
> Jonathan Borden
> This _entire_ controversy exists simply because the meaning
> of the term
> "resource" is hopelessly overloaded, perhaps "URI" is as well.

No. A while back a few people on this list asserted that mappings from
URIs to resources are unambiguous. Indeed, this unambiguous mapping is
an axiom of the Web. Tim Bray says different and is prepared to take a
hard line on the matter. The RDF Model Theory says different too, see

I don't see the disagreement having anything whatsoever to do with
definitions of resource and URIref, but having everything to do with how
these sets are related. You indicate as much: "Each URIref uniquely and
unambiguously labels the node.". That is where the disagreement lies.


I am saying this based on the RDF MT as I've quoted in

> A URIref is a _label_ for a _node_. Each URIref uniquely and
> unambiguously
> labels the node. Not all nodes are required to have labels.
> Nodes may have
> properties and these properties may relate one node to another.

You could call that model a "Node Description Framework" or NDF for

As it happens, we have another model called the "Resource Description
Framework" or RDF for short. I know more about the RDF model than the
NDF one, but at first glance it seems there is at least one difference
between the two. In the RDF model, each URIref must be disambiguated.
Thus it is not an axiom of the RDF model that each URIref uniquely and
unambiguously labels a resource, we expect a process to do that work for

I have no idea why you conclude that "in the RDF model., each URIref must be
disambiguated"? What does this mean and why do you say this? Note that what
I am saying above refers particularly to URIref nodes (i.e. named resources)
as described in the RDF model theory. Unnamed nodes may indeed be
identified/disambiguated by unique properties, for example a person
identified by an SSN, as per OWL, but that is a different matter. Named
nodes (in RDF -- my use of the term "node" is identical to RDF) are
labelled/named by a single URIref.

Received on Friday, 20 September 2002 18:00:09 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:32:34 UTC