RE: working around the identity crisis

Hi Benjamin, all,

> regarding SKOS, would you like to see support of this "slash 
> plus hash"
> approach for concept URIs in SKOS authoring tools? IOW, would you
> encourage glossary authors to create concept URIs of that 
> form? (Or am I
> asking that a too early stage? My current plan was to use 
> slash'd concept
> URIs..)

Well this is exactly my problem ... I'm currently drafting new documentation
for SKOS.  Although I do think the SKOS docs should remain politically
neutral about URI form, the reality is that whatever style gets used in the
examples will get copied.

And the other thing is, folks new to the semantic web wanting to use RDF for
their thesauri are looking for clear recommendations, or at the least a
clear and simple explanation of the cost/benefit for each option, so they
can make a good decision without having to spend a lot of effort, and I'm
not yet in a position to give them either. 

This really is the crux issue, because assigning URIs to thesaurus concepts
is the first step to using the semantic web, and if that first step is
hellishly complex, well that's what they call a 'stumbling block'.

Also, choices regarding URI form have long term consequences, which need to
be carefully considered before a decision is made, to ensure among other
things that the vocabulary remains usable in at least the medium term.

Regarding hash vs. slash, it would be OK if one camp was saying 'X is
better, Y isn't so good' and the other vice versa, but I am getting clear,
unequivocal and opposing statements saying 'do not do X'.   

What I would really like is a clear and up-to-date case from each camp
stating (as briefly and objectively as possible) reasons why a [hash/slash]
URI *should not* be used to identify a thesaurus concept.

I'm heading into the abyss here, aren't I ... ;)  Thing is, I just don't see
a way around it yet.

Cheers,

Al.






> 
> cheers,
> benjamin
> 
> >Al.
> 
> --
> Benjamin Nowack
> 
> Kruppstr. 100
> 45145 Essen, Germany
> http://www.bnode.org/
> 
> 
> 

Received on Tuesday, 16 November 2004 19:43:41 UTC