W3C home > Mailing lists > Public > public-owl-wg@w3.org > January 2008

Re: Fragments discussion

From: Deborah L. McGuinness <dlm@ksl.stanford.edu>
Date: Wed, 30 Jan 2008 01:47:38 -0500
Message-ID: <47A01D8A.9050009@ksl.stanford.edu>
To: Ian Horrocks <ian.horrocks@comlab.ox.ac.uk>
CC: "Web Ontology Language ((OWL)) Working Group WG" <public-owl-wg@w3.org>

We have a candidate here who is speaking during part of the telecon time 
later today so i will miss some of the meeting.
Thus, just in case the fragments discussion is not when i am on, i 
wanted to post the things i care most about with respect to fragments.

1 - i would like to see backwards compatibility with OWL-Lite.
2 - i know of a number of users who would like to use some fragment(s) 
of OWL and i think it is important to have at least one (and hopefully 
not 10s of) option(s) for useful fragment(s).
2a - i would like to have at least one fragment have first class status 
and thus it seems to me that at least one is rec track.
3 - i know of the most need for something like the fragment jeremy and 
jim and asking for and thus, this seems more compelling (although i 
admit to not really knowing how to quantify this notion well).

Ian Horrocks wrote:
> In an effort to get us thinking and to structure our discussion we 
> came up with the following list of questions/topics (see 
> http://www.w3.org/2007/OWL/wiki/Teleconference.2008.01.30/Discussion). 
> Although we don't expect to make any firm decisions, we do hope to 
> reach agreement on some general directions.
> Note that as this is a large and as yet relatively little discussed 
> topic we have allowed 45 minutes. This may mean postponing some or all 
> issue related matters to next week.
>     *  There are (at least) two aspects to this discussion: Language 
> Fragments, defined in terms of (restrictions on) the (structural) 
> syntax, and Conformance levels, defined in terms of implementation 
> behaviour.
>           o Do we understand and agree with this distinction?
>     * "Rule based" fragments such as OWLPrime.
>           o Review of current status.
>           o What are the language fragment and conformance level issues?
>     * OWL-Lite
>           o Do we want to retain a/the OWL-Lite?
>           o Is there a backwards compatibility issue?
>           o How would it relate to other fragments?
>     * Number of fragments
>           o Should we limit the number of fragments?
>           o If so, why and to how many?
>           o Are some fragments more or less compelling than others 
> (e.g., in terms of implementer experience and utility)?
>     * Documentation
>           o Should a/the tractable fragments document be REC track?
>           o Is the existing tractable fragments document appropriate?
>           o Do we need additional user facing documentation for the 
> fragments?
> Regards,
> Ian and Alan
Received on Wednesday, 30 January 2008 06:47:55 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:42:02 UTC