RE: OWL Comment: have long CR period for OWL, or move owl:oneOf, owl: have Value to OWL Full

Hi Jim

Thank you for your response [1].  From the point of view of the formal
process I accept your response and don't want to pursue this point any
further.

However, I would just like to check my understanding of the situation.

You say:

> First, we have dropped the discussion of a "Complete OWL DL 
> Consistency Checker" from the Test document.  We believe this is 
> consistent with your request
> 

My interpretation of this is that you do not expect a complete consistency
checker for OWL DL to exist; in particular, you do not require such a
complete implementation in order to advance the proposal.

You add:

>   The WG has been made aware of implementations of OWL DL 
> that include 
> both inverseOf and oneOf and which seem to be performing well in 
> practice.  The working group will definitely consider their status 
> and usability before deciding on our schedule with respect to 
> Candidate Recommendation and Proposed Recommendation.

The HP representative on the working group (Jeremy) isn't aware of such
implementations; I've asked him to find out more information. However, in
view of your earlier comments I assume that these implementations aren't
complete i.e. there are other OWL DL features which aren't supported in
these implementations.

If it is the case that there are complete implementations of OWL DL that
have acceptable performance and are available on terms compatible with the
W3C patent policy then I would be very pleased to hear it, and you can
safely ignore any of these comments.  However, given your comments earlier I
assume that the intended status of OWL DL is

* it is decidable
* there are implementations of a number of different proper subsets.



From a purely HP-centric point of view (concerned with developing practical
tools and applications of and for the semantic web) this isn't particularly
interesting to us. 

OWL Full is interesting: we would expect a number of implementations to
support proper subsets of OWL Full and agree on the correctness of any
reachable entailments. We expect to be able to share OWL Full documents
amongst different users and for them to have common understanding of the
semantics of these documents. We would expect special-purpose reasoners to
be able to provide proofs for ontologies developed in particular (proper)
subsets of OWL full - we would eventually hope to be able to share these
proofs amongst different users.

OWL Lite is interesting: we would expect a number of implementations of
reasoners for OWL Lite. We would expect to be able to share OWL lite
documents amongst different users and for them to reach the same conclusions
about the consequences of these documents.

But OWL DL is "merely" a subset of OWL Full for which one would not expect
to get full reasoning support, comparable in usability terms to any other
fragment of OWL Full.  It does have the property that any ontology written
in a proper subset of OWL DL has a fighting chance of having reasoning
support, and so could therefore be described, roughly speaking, as the
minimal unimplementable subset of OWL Full.

I would therefore expect us (again, this is an HP-centric comment) to
concentrate on OWL FUll and OWL lite, and largely ignore OWL DL.

I'd welcome any (informal) comments on this

Martin Merry


[1]
http://lists.w3.org/Archives/Public/public-webont-comments/2003Jun/0023.html
>

Received on Wednesday, 25 June 2003 06:16:41 UTC