Re: When did SKOS namespace change, and why?

On 17 Jun 2008, at 09:32, Bernard Vatant wrote:

>
> Thomas
>
>> Hi Bernard,
>>
>> 2008/05/skos is not downwards compatible with 2004/02/skos!
>> E.g.skos:subject is discontinued, and there is no class TopConcept  
>> any more.
> OK. Let me explain my use case and concern. We've been using SKOS  
> in Mondeca for quite a while, among other things, for outcoming  
> workflow towards search engines (Antidot, in cc).
> This is not for toys, it's implemented in production workflow with  
> major customers. See e.g., http://itis1.antisearch.net/fpnrportal/.  
> You won't see it, but there is SKOS under the hood of the faceted  
> classification. And also we've been importing vocabularies  
> published in SKOS, like GEMET which you know well.
> Consider we have been as careful as to use only SKOS features we  
> considered stable, and actually they are, (so none of the above),  
> namely we've been using only so far in this workflow.
>
> skos:Concept
> skos:ConceptScheme
> skos:inScheme
> skos:hasTopConcept
> skos:prefLabel
> skos:broader
> skos:narrower
> skos:definition
>
> All of those have been there since the early SKOS drafts, even well  
> before it was on the W3C track, and they are still there in the new  
> release, but with a new namespace. Why?
> The semantics of those has not changed, or has it?

Bernard

In some cases, the semantics /have/ changed. For example (and I  
appreciate that this in itself may be an issue for discussion),  
skos:broader and skos:narrower are no longer defined to be transitive  
(as was the case in the 2004 vocabulary). Some additional discussion  
on this question was here:

http://lists.w3.org/Archives/Public/public-swd-wg/2008Apr/0032.html

> Thanks for the pointer, answering my previous post. Reading http:// 
> www.w3.org/2008/05/06-swd-minutes.html#skosnamespace where the  
> decision was taken, I keep being puzzled at e.g., (Guus speaking) :  
> "people using the old namespace would migrate without problem" and  
> "anyway as long as the old namespace is there, nothing is made  
> incoherent". What does that mean?

I think what Guus means by the second statement is that if the 2004/ 
skos document is available, vocabularies using it can still be  
interpreted consistently. If we change the 2004/skos vocab in situ,  
this may no longer be the case.

Cheers,

	Sean

--
Sean Bechhofer
School of Computer Science
University of Manchester
sean.bechhofer@manchester.ac.uk
http://www.cs.manchester.ac.uk/people/bechhofer

Received on Tuesday, 17 June 2008 09:51:08 UTC