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

Re: Namespaces wihtout "#" Was: Few CWM Bugs

From: Lars Marius Garshol <larsga@garshol.priv.no>
Date: 28 Nov 2001 15:53:17 +0100
To: <www-rdf-interest@w3.org>
Message-ID: <m33d2zeyle.fsf@lambda.garshol.priv.no>

* Graham Moore
| 
| My perception of the issue was not necessarily to do with topic
| occurrences but is related to one of the issues that TopicMaps
| solves very well. This is the distinction between identities that
| serve as identities for 'resolvable resources', using what ever
| resolution function you choose, and identities that are 'just'
| identities for some concept or thing.

I agree completely. I've just been waiting for the right moment to
chip in and say just this.
 
| As Tim pointed out, a person is not a web page but a web page is a
| resolvable resource with identiity and thus it would be useful if
| this was somehow indicated.

Not only that, but with the solutions currently discussed here there
is no way to speak of the person indicated by the web page *and* the
web page qua web page at the same time. So if I make a statement about
the web page (that it has a particular author, for example), and
Graham uses it to make a statement aobut the person indicated there we
get a case of conflicting identities.

In topic maps this is not a problem because you always rigidly
distinguish between talking about the resource qua resource and about
the resource as an indicator for something that is not network-
retrievable. (Yes, I know that RDF calls both a resource; I'm using
topic map terminology where only network-retrievable resources are
called resources.)

The cleanest solution I've seen proposed for RDF is to use anonymous
nodes to indicate resources that are non-retrievable and then to
assign the resources indicating what the identity of the node is with
a property. If that property is a daml:uniqueProperty the effect
should (I am told) be much the same as with subject indicators in
topic maps.

One solution may of course be that we define a set of RDF properties
for topic maps, among them one for subject indicators, and that you
simply take over and use the subject indicator property.

(I'm currently writing the second version of a paper on RDF and topic
maps where I'm delving into just this issue, and I plan to make an RDF
schema for topic maps. How successful that will be I don't know yet.
I'm currently progressing at a pace of one or two paragraphs a day.)

--Lars M. 
Received on Wednesday, 28 November 2001 09:53:08 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:52 GMT