[rdfweb-dev] namespace versioning

Ian Davis iand at internetalchemy.org
Wed Jul 28 15:02:51 UTC 2004

On 28/07/2004 08:00, Julian Bond wrote:
> That wasn't the problem. The problem was 3 similar versions of the same 
> namespace being used in different foaf files. And in theory as
>  http://www.perceive.net/schemas/20021119/relationship/relationship.rdf#
> predates the following, it knows nothing about
>  http://www.perceive.net/schemas/20031015/relationship/relationship.rdf#
>  http://purl.org/vocab/relationship/
>  http://vocab.org/relationship/rel-vocab-20040308.html

> so there's no way any inference rules in the RDF Schema and OWL could be 
> used to translate the first two into the last without the application 
> being told they are equivalent. Is there a forward and backward chain 
> between all these of dcterm:replaces and dcterm:isReplacedBy? Should 
> (must) there be for any namespace that goes through multiple versions?

if http://purl.org/vocab/relationship/friendOf is declared owl:sameAs
then an owl aware application could add the former property whenever it 
saw the latter and vice versa.

The whole point of the relationship vocabulary update was to fix some 
problems that had arisen with the namespace. The current URI is intended 
to be as permanent as possible. See http://vocab.org/policies/availability

Is your specific issue that the current relationship schema only 
declares http://purl.org/vocab/relationship/friendOf to be owl:sameAs 
http://www.perceive.net/schemas/relationship/friendOf ?

If so, propose the equivilences you need here (assuming Dan doesn't mind 
  relationship vocab changes being discussed here). I only added 
equivilences with the previous version of the vocabulary. You don't even 
have to wait for me or Eric to make changes - just write the triples you 
need as a simple schema and make your application read that as well as 
the standard relationship one.

The cool thing about rdfs and owl is that they enable you to make these 
mappings machine readable and reduce the burden on the application writer.


More information about the foaf-dev mailing list