Re: draft response for LC comment 32 CO1

I agree. We shouldn't attempt to turn the spec into a general  
exposition that tries to explain everything to everyone -- that way  
lies madness (and an *even longer* spec).

I think that the current response is good enough.

Ian


On 24 Feb 2009, at 14:43, Peter F. Patel-Schneider wrote:

> I worry about diluting the sections of the Profiles document with all
> this extra stuff.
>
> peter
>
>
> From: Alan Ruttenberg <alanruttenberg@gmail.com>
> Subject: Re: draft response for LC comment 32 CO1
> Date: Mon, 23 Feb 2009 01:01:07 -0500
>
>> While I think the general answer to Chime's comment is clear, I think
>> we would do well to add a couple of specific examples that  
>> demonstrate
>> a problem that someone might realistically run in to (ideally
>> constructed after a conversation with someone like Chime) to the
>> profiles document, and mention in the response that we intend to do
>> so.
>>
>> -Alan
>>
>> On Fri, Feb 20, 2009 at 5:01 AM, Peter F. Patel-Schneider
>> <pfps@research.bell-labs.com> wrote:
>>> From: Ivan Herman <ivan@w3.org>
>>> Subject: Re: draft response for LC comment 32 CO1
>>> Date: Fri, 20 Feb 2009 09:33:06 +0100
>>>
>>>> Well... although what you say is technically true, I would not  
>>>> find it
>>>> unreasonable if some characteristic _examples_ would be added to  
>>>> the
>>>> profile document for both cases that Chime refers to. This tough
>>>> rebuttal is a little bit to brisk for my taste, and we should  
>>>> have some
>>>> more explanation in our documents in my view. (Whether this is  
>>>> in the
>>>> profile or somewhere else like the primer is of course a  
>>>> different issue!)
>>>>
>>>> I just came up with some things that we could add:
>>>>
>>>> - For his first comment, we could, eg, refer to the fact that  
>>>> arbitrary
>>>> RDF graphs may include statements on the otherwise reserved  
>>>> vocabulary
>>>> with all kinds of funny consequences, or that they would allow  
>>>> defining
>>>> functional datatype properties whose consequences are a bit  
>>>> unpredictable.
>>>
>>> Fine.  I'll put in something like:
>>>
>>> Arbitrary RDF graphs can include constructs that have surprising
>>> consequences.  The reasons for these are many and varied, including
>>> effects on the "syntax" of OWL 2.  Because there are so many ways in
>>> which the rules could incorrect, the working group has decided  
>>> not to
>>> exactly characterize how the incorrectness arises, but instead has
>>> included a mention that arbitrary RDF graphs can affect the
>>> underpinnings of OWL, as follows:
>>>
>>>  For ontologies satisfying the syntactic constraints described in
>>>  Section 4.2, a suitable rule-based implementation will have  
>>> desirable
>>>  computational properties; for example, it can return all and  
>>> only the
>>>  correct answers to certain kinds of query (see Section 4.3 and
>>>  [Conformance]). Such an implementation can also be used with  
>>> arbitrary
>>>  RDF graphs. In this case, however, these properties no longer  
>>> hold —
>>>  in particular, it is no longer possible to guarantee that all  
>>> correct
>>>  answers can be returned*, for example if the RDF graph uses the
>>>  built-in vocabulary in unusual ways.*
>>>
>>>> - An example of the (missing) consequences due to the missing RDF 
>>>> (S)
>>>> axiomatic triples and some of the entailement rules is that
>>>>
>>>> rdf:_i rdf:type rdfs:ContainerMembershipProperty . (axiomatic rule)
>>>> meaning that
>>>> rdf:_i rdfs:subPropertyOf rdfs:member . (rdfs12 entailement rule)
>>>>
>>>> If these were around, an OWL RL user could handle container  
>>>> membership
>>>> more easily with those rules.
>>>
>>> I don't know whether *these* triples and rules provide an example  
>>> of the
>>> missing inferences.  Perhaps Boris can comment or provide an  
>>> example.
>>>
>>>> Ivan
>>>
>>> peter
>>>
>>>
>>>> Peter F. Patel-Schneider wrote:
>>>>> [Draft Response for LC Comment 32:] CO1
>>>>>
>>>>> Dear Chimezie,
>>>>>
>>>>> Thank you for your message
>>>>>   http://lists.w3.org/Archives/Public/public-owl-comments/ 
>>>>> 2009Jan/0039.html
>>>>> on the OWL 2 Web Ontology Language last call drafts.
>>>>>
>>>>> Arbitrary RDF graphs can include constructs that have surprising
>>>>> consequences.  The reasons for these are many and varied,  
>>>>> including
>>>>> effects on the "syntax" of OWL 2.  Because there are so many  
>>>>> ways in
>>>>> which the rules could incorrect, the working group has decided  
>>>>> not to be
>>>>> more explicit in the introduction to OWL 2 RL.
>>>>>
>>>>> There are an infinite number of RDFS axiomatic triples, so  
>>>>> including
>>>>> them all in the OWL 2 RL rules does not directly lead to an  
>>>>> effective
>>>>> rule implementation.  There are some RDFS rules that produce
>>>>> consequences that are not relevant to the conclusions  
>>>>> guaranteed by
>>>>> Theorem PR1.  Listing all the "deficiencies" is not  
>>>>> particularly easy,
>>>>> and would probably only confuse the issue.  The working group has
>>>>> therefore decided not to be more explicit in the preamble to  
>>>>> Theorem
>>>>> PR1.
>>>>>
>>>>> Please acknowledge receipt of this email to
>>>>> <mailto:public-owl-comments@w3.org> (replying to this email should
>>>>> suffice). In your acknowledgment please let us know whether or  
>>>>> not you
>>>>> are satisfied with the working group's response to your comment.
>>>>>
>>>>> Regards,
>>>>> Peter F. Patel-Schneider
>>>>> on behalf of the W3C OWL Working Group
>>>>>
>>>>> No trees are known to have been harmed in the preparation of this
>>>>> response.
>>>>>
>>>>
>>>> --
>>>>
>>>> Ivan Herman, W3C Semantic Web Activity Lead
>>>> Home: http://www.w3.org/People/Ivan/
>>>> mobile: +31-641044153
>>>> PGP Key: http://www.ivan-herman.net/pgpkey.html
>>>> FOAF: http://www.ivan-herman.net/foaf.rdf
>>>

Received on Monday, 2 March 2009 19:15:56 UTC