[Pellet-users] Pellet problem

[Bijan Parsia]

On Wed, August 15, 2007 12:48 pm, Martin O'Connor <martin.oconnor at stanford.edu> said:

> Bijan Parsia wrote:
> 
>>>>These are the SWRL and SWRL built-ins namespaces as defined by the
>>>>SWRL
>>>>Submission. Unfortunately, there are no actual OWL files at these
>>>>locations
>>>>
>>>>
>>
>>I find that *fortunate*.
>>
>>And I would count doing so as a mistake anyway.
>>
> Why?

Because I don't think they should be imported. I don't think "ontologies" with regard to basic syntax are a good idea. If I had my druthers, they wouldn't exist at all.

> These 'missing' ontologies have caused a lot of user confusion over
> the past few years.

Because of the bad practice of suggesting they import them.

> They require custom configuration for every OWL tool
> that wants to import these ontologies.

No OWL tool *should* want to import them. They shouldn't (and don't) import owl.owl. (If they did, every such ontology would automatically be in OWL Full.)

This route of syntax extension is, IMHO, fundamentally broken.

> Redirecting to the web site of a
> project that has been dead for three years always seemed odd.

I was a bit too elliptical in my reply. My objection is to importing them in the first place.

> One of the
> authors of the SWRL Submission was happy enough to send an email to the
> W3C to request the ontologies be moved there (they ignored him too).

If you are going to import them, then having them at a decent spot is, of course, important. Just don't import them :)

[snip]

Cheers,
Bijan.