W3C home > Mailing lists > Public > www-rdf-comments@w3.org > April to June 2002

Re: namespace question

From: Brian McBride <bwm@hplb.hpl.hp.com>
Date: Tue, 02 Apr 2002 20:29:57 +0100
Message-Id: <5.1.0.14.0.20020402202131.031e9e10@0-mail-1.hpl.hp.com>
To: "Tim Sebel" <tsebel@enleague.com>(by way of "Ralph R. Swick" <swick@w3.org>), www-rdf-comments@w3.org
Hi Tim,

At 12:52 02/04/2002 -0500, Tim Sebel wrote:
>[freed from spam trap -rrs]
>
>Date: Tue, 2 Apr 2002 12:25:56 -0500 (EST)
>Message-ID: <FLEFILNFDLBKOIAHHDHEMEALCAAA.tsebel@enleague.com>
>From: "Tim Sebel" <tsebel@enleague.com>
>To: <www-rdf-comments@w3.org>
>
>I am fairly new to rdf/xml/daml+oil and ontology building, so I may be
>missing the obvious answer, however:
>
>Throughout our ontologies, we reference rdf, rdfs, and daml in our namespace
>declarations. I have noticed that the namespaces don't always reference the
>same URL.

If this is true of RDF and RDFS then its probably a mistake.  Neither the 
RDF or the RDFS namespaces have changed in a while.  DAML has been through 
a number of versions and the namespace changed as it evolved.  This is 
correct behaviour.

There has also been an issue with DAML where there has been confusion over 
whether the namespace includes a .daml extension.  It doesn't.

>  Some of the referenced URLs no longer exist, if they ever did.

There is no requirement in RDF and RDFS that a namespace URI be 
dereferencable, though they often are.  The RDFS namespace is 
dereferencible.  I can't remember whether the RDF one is or not.


>Is there a standard/suggested way to deal with changing URLs for new
>rdf/rdfs/daml specs?

One trick, favoured by some and not by others is to use entities, e.g. 
define an entity &rdf; to correspond to the namespace uri reference and use 
that wherever the namespace might be used.

An issue with this approach is that entities are a dtd concept and some 
folks favour the obsolesence of dtd's.   By the time that happens there is 
a good chance that RDF will have been extended to accept qnames in 
attribute values, so a simple search and replace should fix that.

>Am I going to need to alter them throughout our ontology?
>
>I was considering creating a schema that points to the relevant URLs outside
>our ontology and then referencing the new schema everywhere else.

I'm not sure I see how that would work.

Brian
Received on Tuesday, 2 April 2002 14:33:18 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 21 September 2012 14:16:30 GMT