ISSUE-153: Comments on SKOS namespace change question

Hi,

I am currently doing some SKOS implementations in the form of an API  
and a SKOS editor kit for Protege 4.  I thought it might be useful to  
let you know the implications of a namespace change from a tools point  
of view.

My tools import and use the SKOS data model directly, this means  
knowledge of SKOS namespaces is defined somewhere in the source code.  
However, the SKOS semantics are taken directly from those defined in  
the data model and handled by standard reasoners, I get this from the  
SKOS ontology you publish on the web.  If you change the semantics at  
any point, my tools still work and always give the right answer  
according the published spec, it's up to the user to be aware of what  
the SKOS semantics entail. So for this reason I would be in favor of  
keeping the old namespaces and just changing the semantics.  Changing  
the namespaces will break everything. Obviously i can just update my  
code to handle the new namespaces without too much hassle, but then  
you will introduce a backward compatibility headache for me.

A simple example would be, in the editor I want it to handle files  
that use both the old and new namespaces. But what would you expect to  
happen if you said "add broader" in the tool, should it use the old or  
the new namespaces by default? I don't want to implement a "old" SKOS  
mode and a "new" SKOS mode... it's messy and confusing I think!

By keeping the old namespaces you avoid any implementors having to  
worry about old and new SKOS. If you go with the new namespaces I am  
inclined to just not support the old ones as they were defined before  
the SKOS final spec was released.

Cheers,
Simon


Simon Jupp
University of Manchester
simon.jupp@manchester.ac.uk
http://www.cs.man.ac.uk/~sjupp/



On 2 Oct 2008, at 10:22, Sean Bechhofer wrote:

>
>
> Dear Tim,
>
> Thanks again for your comments which have been raised as ISSUE-153.  
> The Working
> Group's issue tracker system is online at:
>
> http://www.w3.org/2006/07/SWD/track/
>
> We hope to respond by Friday 10th October.
>
> Kind regards,
>
> Sean Bechhofer
>
> On 30 Sep 2008, at 20:52, Tim Berners-Lee wrote:
>
>>
>>
>> Reading http://www.w3.org/TR/2008/WD-skos-reference-20080829/ ,
>>
>> I sympathize very much with the people who want to keep the  
>> namespace the same.
>> We're trying to get critical mass and cutting some of the data off  
>> and letting it float away by itself it is costly. I don't generate  
>> skos myself, but I have come across it.
>>
>> Some people think it's important. I strongly suggest giving new  
>> names (within the same namespace)  to the five things which have  
>> changed, especially if they're rather obscure.
>>
>> You have of course to evaluate the damage if you were to just make  
>> an erratum to skos 2005.
>>
>> Did you in 2005 make a say about the change rules for skos 2005
>> I suggest to use new names and not change the namespace.
>>
>> Tim
>>
>>
>>
>>
>
> --
> Sean Bechhofer
> School of Computer Science
> University of Manchester
> sean.bechhofer@manchester.ac.uk
> http://www.cs.manchester.ac.uk/people/bechhofer
>
>
>
>

Received on Thursday, 2 October 2008 15:53:07 UTC