W3C home > Mailing lists > Public > uri@w3.org > August 2001

Re: I-D ACTION:draft-masinter-dated-uri-00.txt

From: Sean B. Palmer <sean@mysterylights.com>
Date: Sun, 26 Aug 2001 16:13:14 +0100
Message-ID: <00b601c12e41$c063c340$91ed93c3@Palmer>
To: "Larry Masinter - LMM@acm.org" <lmnet@attglobal.net>, "Al Gilman" <asgilman@iamdigex.net>
Cc: <uri@w3.org>, <www-rdf-interest@w3.org>
> 'tdb' may be heuristically expanded "that described
> by" and it is not the representation or recovered value
> of the resource that this indicates at all, [...]

Yeah, that's what I said: a "duri" identifies the representation of a
resource, and a "tdb" identifies the resource itself. Of course, a
representation of a resource can be a resource itself, and is now
identified as such by a "duri". But a "tdb" is only useful on concepts
which are not time-independant, and so to use a "tdb" on a "data:" URI, or
a "duri" URN is worthless.

> This makes "tdb(XHTML 1.1 definitive URI)" a perfectly
> beautiful 'predicate' [...]

It goes further than that; it's an rdfs:Resource. It can be a predicate,
subject or object. Examples:-

   as subject:-
     <tdb:(XHTML 1.1 definitive URI)> :title "XHTML 1.1" .

   as predicate:-
     <tdb:(RDFS subPropertyOf URI)> a rdf:Property .

   as object:-
     <duri:(some page)> :format <tdb:(XHTML 1.1 definitive URI)> .

I'm not so sure about "duri"... it can probably only be considered an
instance of classes which are not properties.

   { { :x :uri [ :startsWith "urn:duri:" ] }
   log:implies
   { :x a [ daml:complementOf rdf:Property ] } }
   a log:Truth; log:forAll :x .

Ah, the Semantic Web :-)

> to apply in EARL even if it is a funny sort of URI.  But
> EARL wants that sort of funny URIs, doesn't it?

EARL conventionally expresses this sort of information in the model,
because that way, processors can get to it. I have a sneaking suspiscion
that "tdb" and "duri" violate the opacity of URIs axiom, in that they
contain information about the resource explicitly in the URI scheme itself.
I don't like that at all. On other words, the following DURI:-

   duri:2001:http://infomesh.net/

is just short for:-

   [ :date "2001"; :representationOf <http://infomesh.net/> ] .

The second form is better because you don't have to do weird things with
the URI scheme in order to get the consituent information.

--
Kindest Regards,
Sean B. Palmer
@prefix : <http://webns.net/roughterms/> .
:Sean :hasHomepage <http://purl.org/net/sbp/> .
Received on Sunday, 26 August 2001 11:13:26 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 13 January 2011 12:15:29 GMT