Revision as of 22:27, 12 June 2008 by SandroHawke
Please justify/explain all edits to this page, in your "edit summary" text.
00:00:00 <sandro> PRESENT: IanH, bmotik, Ivan, Evan_Wallace, MarkusK, uli, msmith, Sandro, bcuencagrau, calvanese, Achille, Peter_Patel-Schneider, Zhe, alanru, bijan, JeffPan, clu, m_schnei 16:59:55 <RRSAgent> RRSAgent has joined #owl 16:59:55 <RRSAgent> logging to http://www.w3.org/2008/06/11-owl-irc 16:59:55 <Zakim> bmotik, you need to end that query with '?' 17:00:01 <bmotik> Zakim, who is here? 17:00:01 <Zakim> On the phone I see IanH, bmotik (muted) 17:00:02 <Zakim> On IRC I see RRSAgent, MarkusK, ivan, ewallace, IanH, pfps, Zakim, bmotik, sandro, trackbot 17:00:16 <uli> uli has joined #owl 17:00:32 <ivan> zakim, dial ivan-voip 17:00:36 <Zakim> ok, ivan; the call is being made 17:00:39 <Zakim> +Ivan 17:00:40 <Zakim> +Evan_Wallace 17:00:44 <msmith> msmith has joined #owl 17:00:47 <Zakim> +??P4 17:00:54 <IanH> zakim, who is here? 17:00:54 <Zakim> On the phone I see IanH, bmotik (muted), Ivan, Evan_Wallace, ??P4 17:00:55 <Zakim> On IRC I see msmith, uli, RRSAgent, MarkusK, ivan, ewallace, IanH, pfps, Zakim, bmotik, sandro, trackbot 17:01:10 <Zakim> +??P5 17:01:13 <uli> zakim, ??P5 is me 17:01:13 <Zakim> +uli; got it 17:01:20 <uli> zakim, mute me 17:01:20 <Zakim> uli should now be muted 17:01:24 <Zakim> +msmith 17:01:36 <sandro> RRSAgent, pointer? 17:01:36 <RRSAgent> See http://www.w3.org/2008/06/11-owl-irc#T17-01-36 17:01:45 <bcuencagrau> bcuencagrau has joined #owl 17:01:46 <Zakim> +Sandro 17:01:47 <calvanese> calvanese has joined #owl 17:01:55 <IanH> zakim, who is here? 17:01:55 <Zakim> On the phone I see IanH, bmotik (muted), Ivan, Evan_Wallace, MarkusK, uli (muted), msmith, Sandro 17:01:57 <Zakim> On IRC I see calvanese, bcuencagrau, msmith, uli, RRSAgent, MarkusK, ivan, ewallace, IanH, pfps, Zakim, bmotik, sandro, trackbot 17:04:00 <ivan> scribe: Markus 17:04:14 <IanH> zakim, who is here? 17:04:14 <Zakim> On the phone I see IanH, bmotik (muted), Ivan, Evan_Wallace, MarkusK, uli (muted), msmith, Sandro, bcuencagrau, calvanese 17:04:16 <Zakim> On IRC I see calvanese, bcuencagrau, msmith, uli, RRSAgent, MarkusK, ivan, ewallace, IanH, pfps, Zakim, bmotik, sandro, trackbot 17:04:29 <ewallace> Who has the machine that goes "ping"? 17:04:48 <> Topic: Admin 17:04:48 <MarkusK> Subtopic: Agenda amendments 17:04:48 <MarkusK> no agenda amendments 17:04:54 <Zakim> +[IBM] 17:05:08 <MarkusK> Subtopic: Previous minutes 17:06:07 <MarkusK> Ian: can someone approve previous minutes? 17:06:02 <uli> they aren't pretty 17:06:05 <pfps> the previous minutes were not acceptable yesterday 17:06:22 <msmith> msmith has changed the topic to: http://www.w3.org/2007/OWL/wiki/Teleconference.2008.06.11/Agenda 17:06:33 <MarkusK> Ian: Minutes may need new mechanism to be prepared. 17:06:46 <MarkusK> Sandro: The old partial minutes have confused people. 17:07:05 <MarkusK> Link to unformatted minutes: http://www.w3.org/2007/OWL/wiki/Chatlog_2008-06-04 17:07:31 <MarkusK> Sandro: the scribe should edit the IRC log and it should be reformatted later on request. 17:07:36 <pfps> no fixes were performed on the minutes at all - 17:08:01 <pfps> what is "the right place"? 17:08:03 <MarkusK> Ian: previous minutes cannot be accepted yet since not many people saw the final version 17:08:55 <MarkusK> Subtopic: F2F3 registration 17:08:55 <MarkusK> Ian: registration for F2F3 still should be completed 17:09:06 <MarkusK> Topic: Action item review 17:09:06 <MarkusK> Subtopic: Action 148 17:09:58 <MarkusK> Achille: Review of RDF mapping document completed last week. I found them to be OK, though primer still needs to be updated. My only point was in the syntax document. 17:11:34 <MarkusK> Ian: The above refered to Action 148, which was still open, though not mentioned in the agenda 17:12:04 <MarkusK> Ian: Action 148 completed 17:12:23 <MarkusK> Ian: Boris' Action 131 shall be deferred until later 17:12:41 <MarkusK> Subtopic: Action 42 17:13:11 <MarkusK> Bijan: Action is ongoing, I hope to have it done by next week 17:13:36 <MarkusK> Subtopic: Action 147 17:13:45 <MarkusK> Ian: A detailed review has been provided already. 17:14:00 <IanH> zakim, who is here? 17:14:00 <Zakim> On the phone I see IanH, bmotik (muted), Ivan, Evan_Wallace, MarkusK, uli (muted), msmith, Sandro, bcuencagrau (muted), calvanese (muted), Achille, Peter_Patel-Schneider, Zhe, jar, 17:14:03 <Zakim> ... ??P8, JeffPan 17:14:04 <Zakim> On IRC I see alanr, bijan, JeffPan, JeffP, Zhe, Achille, calvanese, bcuencagrau, msmith, uli, RRSAgent, MarkusK, ivan, ewallace, IanH, pfps, Zakim, bmotik, sandro, trackbot 17:14:05 <MarkusK> Ian: But Michael Schneider is not on the call to comment. 17:14:53 <MarkusK> Boris: I did look at Michael's review, though without checking all details. 17:14:53 <IanH> zakim, who is here? 17:14:53 <Zakim> On the phone I see IanH, bmotik, Ivan, Evan_Wallace, MarkusK, uli (muted), msmith, Sandro, bcuencagrau (muted), calvanese (muted), Achille, Peter_Patel-Schneider, Zhe, jar, ??P8, 17:14:56 <Zakim> ... JeffPan 17:14:57 <Zakim> On IRC I see alanr, bijan, JeffPan, JeffP, Zhe, Achille, calvanese, bcuencagrau, msmith, uli, RRSAgent, MarkusK, ivan, ewallace, IanH, pfps, Zakim, bmotik, sandro, trackbot 17:15:29 <MarkusK> Boris: The points Michael Schnieder made appear to be minor suggestions for the most part, and I will try to implement the easy comments first. I will come back with any non-obvious comments for further discussion. 17:16:19 <MarkusK> Ian: Then let us close Action 147 for the moment. 17:16:43 <IanH> zakim, who is here? 17:16:44 <Zakim> On the phone I see IanH, bmotik (muted), Ivan, Evan_Wallace, MarkusK, uli (muted), msmith, Sandro, bcuencagrau (muted), calvanese (muted), Achille, Peter_Patel-Schneider, Zhe, 17:16:47 <Zakim> ... alanr, ??P8, JeffPan 17:16:48 <Zakim> On IRC I see m_schnei, alanr, bijan, JeffPan, JeffP, Zhe, Achille, calvanese, bcuencagrau, msmith, uli, RRSAgent, MarkusK, ivan, ewallace, IanH, pfps, Zakim, bmotik, sandro, 17:16:50 <Zakim> ... trackbot 17:16:57 <MarkusK> Ian: ??P8 and jar are unidentified participants who must be identified. 17:17:13 <MarkusK> Bijan: I fixed this now 17:16:57 <MarkusK> (all participants have then been identified) 17:17:10 <bijan> zakim, who is here? 17:17:10 <Zakim> On the phone I see IanH, bmotik (muted), Ivan, Evan_Wallace, MarkusK, uli (muted), msmith, Sandro, bcuencagrau (muted), calvanese (muted), Achille, Peter_Patel-Schneider, Zhe, 17:17:13 <Zakim> ... alanr, bijan, JeffPan 17:17:14 <Zakim> On IRC I see m_schnei, alanr, bijan, JeffPan, JeffP, Zhe, Achille, calvanese, bcuencagrau, msmith, uli, RRSAgent, MarkusK, ivan, ewallace, IanH, pfps, Zakim, bmotik, sandro, 17:17:16 <Zakim> ... trackbot 17:17:34 <MarkusK> Subtopic: Action 150 17:17:48 <MarkusK> Ian: Jie Bao is not here to comment, so Action 150 remains open until next week. 17:18:34 <m_schnei> I have finished action 147 17:18:20 <MarkusK> Topic: Issues 17:18:20 <MarkusK> Subtopic: Issue 124 17:18:46 <MarkusK> Ian: Issue 124 appeared to be a rather obvious fix for the OWL Full semantics 17:19:14 <MarkusK> Boris: Yes, we can change the mapping to address that issue. 17:20:00 <MarkusK> Ian: So we can propose to resolve Issue 124. 17:20:21 <IanH> PROPOSED: Resolve Issue 124 as per http://www.w3.org/mid/0EF30CAA69519C4CB91D01481AEA06A096B434@judith.fzi.de 17:20:28 <bmotik> +1 17:20:32 <IanH> +1 17:20:36 <Zhe> +1 17:20:37 <bcuencagrau> +1 17:20:40 <pfps> +1 17:20:42 <sandro> 0 (sorry, not up to speed on issue) 17:20:46 <ivan> +1 17:20:47 <JeffPan> 0 17:20:49 <uli> +1 17:20:50 <msmith> +1 17:20:55 <ewallace> +1 17:20:56 <bijan> +1 17:21:12 <IanH> RESOLVED: Resolve Issue 124 as per http://www.w3.org/mid/0EF30CAA69519C4CB91D01481AEA06A096B434@judith.fzi.de 17:21:17 <clu> Sorry for being late. 17:18:20 <MarkusK> Subtopic: Issue 104 17:21:46 <MarkusK> Ian: Issue 104 was discussed in many emails, and many people raised concerns. 17:23:02 <MarkusK> Boris: One issue is that reification and collections are the only ones that make sense to take out of the reserved vocabulary. 17:23:46 <MarkusK> Boris: I sent a proposal for having a shadow-vocabulary for OWL, but this met little approval. 17:23:52 <alanr> The proposal was to define our own terminology for the serialization. 17:24:16 <MarkusK> Boris: the technical problem is that we have no ObjectProperty/DatatypeProperty puning, hence vocabulary like rdf:first are not easy to define in OWL: one would have to asign a fixed type. 17:24:44 <alanr> We need not assign a type in the language - leave it to the modeler. 17:25:14 <MarkusK> Boris: rdf:first may then get many types in different applications and I thus propose to not allow it to be used in OWL DL vocabulary. Otherwise modelling could become very messy. 17:25:15 <bijan> Or a shadow vocabulary 17:25:43 <MarkusK> AlanR: There were various proposals to address this. One was to have an OWL shadow vocabulary, such as owl:first, to resolve possible typing conflicts. The other proposal was to admit rdf:first and leave typing to modellers. One would then need to use OWL Full if conflicting types for collection properties would occur. 17:26:20 <m_schnei> in my mail I argue for not treat lists at all in OWL DL, so people may declare it to be either a data or a object property, if they wish 17:26:52 <m_schnei> The property rdf:first would then be just an URI like any other 17:27:12 <msmith> q+ to ask what use case shadow vocab addresses 17:27:37 <alanr> "messy" is subjective 17:28:14 <bijan> Subjective considerations aren't invalid (as we've seen :)) 17:27:41 <MarkusK> Boris: I do think that having an explicit type for rdf:first would be no good modelling practice. Ontologies should not contain declarations for such properties, and I would rather like to not have it. 17:27:53 <alanr> then owl full 17:27:59 <m_schnei> a shadow vocabulary for lists can be used /always/ by users - no need to say anything about it 17:28:37 <alanr> shadow was for internal use 17:28:38 <bijan> Interop 17:28:52 <MarkusK> MSmith: I exactly agree with Boris. I see no use case for a shadow vocabulary though. 17:29:05 <alanr> yes 17:29:15 <alanr> reverse mapping 17:29:18 <alanr> fixes this 17:29:31 <alanr> not 17:29:50 <alanr> billions and billions 17:29:53 <msmith> yes 17:29:54 <alanr> served 17:30:15 <bmotik> By the way, I have just fixed the mapping document regarding owl:datatypeComplementOf. 17:30:19 <MarkusK> Ivan: Existing OWL ontologies may already use RDF constructs, so it is not clear that we even have an option for disallowing that now in OWL 2. 17:30:52 <MarkusK> AlanR: the shadow vocabulary would be for our own serialisation, not for users. Restricting uses of rdf:first in a certain way may still be better than not allowing it at all. 17:32:17 <msmith> I am now confused about the intended use of the "shadow vocabulary" 17:32:32 <ivan> msmith, you are not the only one:-) 17:33:05 <alanr> idea is that we use owl:first, owl:next, owl:nil in our serialization 17:33:29 <MarkusK> Boris: switching from RDF lists to something else in serialisation seems to be no good idea. Many ontologies are also already using RDF lists. 17:33:30 <ivan> +1 to boris 17:33:29 <MarkusK> Boris: But I think it is not a major backwards compatibility issue. 17:33:54 <alanr> I'd like the backwards compatibility case spelled out clearly, please 17:34:07 <bijan> I think mike was looking for the utility of the shadow vocabulary 17:34:15 <bijan> I wonder that so many people can't understand mike :) 17:35:25 <MarkusK> Boris: I proposed the shadow vocabulary, so that users are freed of some burdon reinventing vocabulary for standard tasks. 17:35:38 <ivan> +1 again to Boris 17:35:29 <alanr> Can't introduce it to the RDFS users - they are the ones that we want to bring in to the fold 17:35:41 <MarkusK> Boris: But the drawback is that we get into modelling discussions here. We cannot really reason about lists logically. 17:36:09 <alanr> q+ to say we are *not* getting into modelling. We are getting out of the way of modelers. 17:36:34 <MarkusK> Boris: the list is just a part of data, not a true semantic construct. 17:36:23 <alanr> no no 17:36:29 <alanr> we are trying to let more rdf be owl 17:36:41 <alanr> they can all be defined as annotation properties if need be 17:37:12 <MarkusK> Bijan: I second Boris' concern on the lack of utility of the list vocabulary. I often suggest to users to not employ RDF collections or containers in OWL ontologies. In our practical experience, users were willing to use a custom vocabulary to model lists. It is usually easy to migrate RDF lists to some custom vocabulary. 17:37:29 <alanr> this is clearly getting in to the modeling business - we think it is bad modeling, so we will forbid it? 17:38:10 <m_schnei> do i understand this right? rdf:first as an annotation property? and if a reasoner throws away all annotations? then there are a lot of unconnected nodes around. :) 17:38:28 <bmotik> Alan, we will prohibit the usage of rdf:List for technical reasons that are not negligible. People can model lists using their own vocabulary if they want. 17:38:47 <alanr> I don't see the technical reasons as being very much 17:38:57 <alanr> just not convinced 17:39:06 <bijan> "seeing" is subjective :) 17:39:15 <sandro> what about promoting a standard list vocabulary? 17:39:37 <MarkusK> Pfps: Where exactly are RDF lists used in OWL ontologies? 17:39:47 <MarkusK> Ivan: Mostly in the serialisation of OWL. 17:39:41 <sandro> (owl:ObjectList and owl:DataList, etc.... ) 17:39:43 <alanr> if it can be done unambiguously then great! 17:39:48 <bijan> sandro, without proper list semantics? 17:39:55 <sandro> no, with. 17:40:00 <bmotik> What is the proper list semantics? 17:40:06 <bijan> Now we're out of first order logic at least 17:40:07 <bmotik> Lists cannot be modeled semantically! 17:40:08 <m_schnei> there are no RDF lists customly used in OWL DL, since this is not allowed :) 17:40:18 <bmotik> Yes, in FOL, I meant. 17:40:19 <bijan> Since normally lists are well founded and defined with transitive closure 17:40:27 <alanr> any ontology that uses owl2:first is owl 2 17:40:45 <alanr> every ontology that is owl2 is serialized with mention of owl2:first 17:40:51 <msmith> the times I have seen lists in OWL, the type of items is also restricted, which would require specializing any "standard" shadow vocabulary 17:40:54 <ivan> :-) 17:41:19 <Zakim> alanr, you wanted to say we are *not* getting in to modeling. We are getting out of the way of modelers. 17:41:51 <MarkusK> Alanr: I do not think that we truly need to use a shadow vocabulary. A shadow vocabulary would be closed, hence one can check for occurrences of this vocabulary to decide if a serialisation belongs to OWL 2. OWL should allow lists as data, since people adopt them due to their syntactic simpliciy in Turtle and SPARQL. 17:42:32 <bijan> http://www.daml.org/services/owl-s/1.1/generic/ObjectList.owl 17:42:40 <pfps> if you don't use owl2 vocabulary but do use lists in an object sense, are you in owl1 then? 17:43:08 <sandro> q+ to support Alan 17:43:15 <alanr> michael is not here, he supports 17:44:17 <MarkusK> AlanR: Michael may also support my position, but is not on the call. I am certainly not convinced by the current arguments against it. 17:44:53 <MarkusK> MarkusK: There was some confusion here. I have not voiced any oppinion on this issue, I am scribing. 17:45:05 <alanr> sorry - my mistake 17:45:59 <alanr> http://bibliontology.com/ 17:46:22 <MarkusK> Sandro: can we have a straw poll to get some impression here? At least to measure general motivation in the group. Proposed strawpoll: It would be nice, if we can find a workable technical solution, to support RDF List in OWL DL. 17:46:05 <ewallace> +1 to Sandro's suggestion 17:46:37 <msmith> alanr, what part of that ontology? 17:46:10 <alanr> exactly 17:46:50 <alanr> hunting - discussion was in email 17:46:58 <MarkusK> Ian: still some speakers on the queue first 17:47:15 <alanr> http://bibliontology.com/#term_contributorList 17:47:39 <MarkusK> Boris: the technical questions seem to be rather severe. Changing the vocabulary is not a trivial change. Using things such as rdf:first in conjunction with OWL constructs such as nominals may have unexpected/complex consequences both in OWL DL and in OWL Full. 17:49:18 <m_schnei> In OWL Full, there isn't any restriction on the use of rdf:first, anyway 17:49:39 <MarkusK> Bijan: I see a user need for expressing lists, but we can leave it to implementors and future work to properly solve that 17:49:08 <alanr> Could we get documentation on how Pellet accomodates? 17:49:35 <bijan> See the code :) 17:50:11 <alanr> yes, will check code, but if you could give a hint, that would be greatly appreciated ;-) 17:51:30 <bijan> I don't know off hand 17:50:02 <IanH> STRAWPOLL: It would be nice, if we can find a workable technical solution, to support RDF List in OWL DL. 17:50:13 <sandro> +1 17:50:15 <alanr> +1 17:50:16 <bijan> That's the straw poll? 17:50:22 <msmith> +1 it would be nice. it doesn't seem feasible 17:50:25 <Achille> +1 17:50:26 <JeffPan> 0- 17:50:28 <pfps> +1, in the same sense that it would be nice to have rules, self-knowledge, etc., etc. 17:50:31 <m_schnei> +1 (would be nice, but no shaddow vocab) 17:50:35 <ewallace> +0 17:50:35 <Zhe> +1 would be nice 17:50:37 <ivan> +1 would be nice... 17:50:37 <uli> +1 but not too optimistic 17:50:41 <bmotik> +1 it would be nice, but I strongly doubt we can solve this 17:50:43 <alanr> +1 to transmutation 17:50:46 <bcuencagrau> 0 17:50:49 <bijan> +0 but I wouldn't mind free puppies either 17:50:49 <clu> 0 17:51:08 <pfps> who is going to bell this cat? 17:51:13 <bijan> Er... if someone wants to , they should do so 17:51:14 <alanr> Can we get a list of issues to start? 17:51:15 <MarkusK> Sandro: maybe a follow-up straw poll on the amount of resources to invest in the issue would be useful 17:51:43 <m_schnei> for me, the question is, whether we can just say /nothing/ about RDF lists, and it would work 17:52:11 <MarkusK> Ian: are there volunteers for trying to solve the problem? 17:51:44 <alanr> I will volunteer if Michael Schneider will 17:52:22 <m_schnei> I don't know what to volunteer for, but if it sounds good, I will do it ;-) 17:52:22 <msmith> I can talk to you about Pellet 17:52:36 <m_schnei> please say in irc, alan! 17:52:56 <alanr> yes 17:52:56 <MarkusK> Ian: AlanR and Michael Schneider shall pursue the issue 17:53:15 <MarkusK> Alanr: I will start by compiling a list of concerns that were raised so far. 17:53:47 <alanr> action: Alan to work with M_schnei to collect, propose how to address issues in making rdf list vocabulary 17:53:47 <trackbot> Created ACTION-159 - Work with M_schnei to collect, propose how to address issues in making rdf list vocabulary [on Alan Ruttenberg - due 2008-06-18]. 17:54:16 <m_schnei> OK, people, I need to stop for about 15 minutes. My battery is down. Sorry! 17:54:23 <m_schnei> bye for now! 17:54:18 <MarkusK> Ian: discussion on "at risk" features such as EasyKeys, especially regarding whether or not such features should appear in the spec with an appropriate disclaimer. Any comments? 17:53:57 <bijan> nary! 17:56:12 <MarkusK> Bijan: Many discussions are not finished yet at the current state, and we still need to gather more information. We need to decide on the current status of each feature, and we can add our concerns to the spec to gather feedback. "At risk" comments are not a problem, I would like the features to be in the spec in general. 17:57:05 <bijan> http://www.w3.org/2007/OWL/wiki/Easy_Keys#Spec_Proposal 17:57:09 <sandro> +1 Bijan add them now, no need for "At Risk" 17:57:46 <bijan> q+ to mention spec work 17:57:48 <MarkusK> Boris: I do not like to add feartures to the spec now that we may remove later on. I would prefer to first do some investigations, and then start modifying the spec. 17:58:43 <MarkusK> Bijan: For EasyKeys, extending the spec should not be hard, since the existing text is almost ready for use in the spec. I agree that implementation experiences are good, but adding the features to the spec would still help to gather more feedback. 18:00:37 <bmotik> My main comment is that it is not only the structural spec that changes: most of the documents will need to change. 18:00:59 <bijan> Boris, yes, I'm working on bits for rdf mapping and sematncis as well 17:59:23 <MarkusK> Ian: do you generally consider these features to be modular? 17:59:41 <MarkusK> Bijan: EasyKeys and Top/Bottom properties both seem to be modular. 17:59:48 <MarkusK> Boris: I agree. 17:59:48 <bijan> yep 17:59:53 <MarkusK> Ian: I will prepare a straw poll. 18:00:36 <IanH> STRAWPOLL: easy keys and top and bottom roles should be added to spec with comment that they could be removed later if implementation experience is negative. 18:00:45 <bmotik> But I can live with that 18:00:50 <ivan> +1 18:00:57 <sandro> +1 18:00:58 <Achille> +1 18:00:59 <uli> +1 18:01:00 <Zhe> +1 18:01:00 <clu> +1 18:01:00 <bijan> +1 18:01:01 <ewallace> +1 18:01:05 <msmith> +1 18:01:05 <alanr> +1 18:01:06 <bmotik> -0 18:01:09 <JeffPan> 0- 18:01:10 <bcuencagrau> 0 18:01:18 <pfps> +0 18:01:28 <alanr> is -0 = 0- ? 18:01:44 <MarkusK> Ian: so it seems that it is OK for the group to add both with some comment. 18:01:38 <bmotik> Frankly, we don't need a comment. 18:01:39 <alanr> which can also be read as "is someone looking" 18:02:07 <bijan> A joint action? 18:02:08 <MarkusK> Boris: Is that a resolution? Shall we have an action? 18:02:12 <bijan> We'd need a resolution. 18:02:17 <MarkusK> Ian: I think so. 18:02:25 <alanr> absolutely 18:02:51 <MarkusK> Bijan: we first need an official resolution. 18:03:34 <MarkusK> AlanR: we can also consider that again before the next publication. 18:03:34 <IanH> PROPOSED: add easy keys and top and bottom roles to the spec; review when we get to next publication round 18:03:49 <bijan> +1 18:03:54 <alanr> +1 18:03:55 <bmotik> +1 18:04:00 <Zhe> +1 18:04:02 <bcuencagrau> +1 18:04:05 <IanH> +1 18:04:06 <JeffPan> +1 18:04:09 <uli> +1 18:04:13 <ivan> +1 18:04:21 <msmith> +! 18:04:24 <sandro> +1 18:04:29 <msmith> +1 18:04:33 <pfps> +0 18:04:34 <ewallace> +1 18:04:36 <clu> +1 18:04:42 <IanH> RESOLVED: add easy keys and top and bottom roles to the spec; review when we get to next publication round 18:03:42 <bmotik> ACTION: bmotik2 to Add easy keys and Top and Bottom role to the spec 18:03:42 <trackbot> Created ACTION-160 - Add easy keys and Top and Bottom role to the spec [on Boris Motik - due 2008-06-18]. 18:05:08 <MarkusK> Subtopic: Issue 109 18:05:24 <MarkusK> Ian: are there new technical arguments on that or shall we just vote? 18:06:26 <MarkusK> AlanR: Ivan and Bijan might be able to reach an agreement. 18:06:49 <MarkusK> Ivan: Bijan and I tried to compile all pros and cons to support the decision. The working group now has to consider these and make a decision. 18:07:16 <IanH> http://lists.w3.org/Archives/Public/public-owl-wg/2008Jun/0031.html 18:07:44 <MarkusK> Ian: The email is rather long, so we should give people the chance of reading the email. Who read it already? 18:07:42 <ivan> i did 18:07:44 <uli> I did 18:07:45 <pfps> me 18:07:45 <bijan> I did 18:07:49 <alanr> me 18:07:59 <sandro> I didn't. :-( 18:08:00 <bijan> Michael did 18:08:01 <Achille> I did not 18:08:03 <ewallace> I didn't 18:08:06 <Zhe> am reading it now 18:08:07 <bcuencagrau> I didn't 18:08:16 <JeffPan> I just did but didn't finish 18:08:29 <alanr> action to all who haven't read it? 18:08:29 <trackbot> Sorry, couldn't find user - to 18:08:36 <m_schnei> m_schnei has joined #owl 18:09:12 <MarkusK> Ian: I will postpone that to next week, and every participant next week should be prepared to vote on that issue, i.e. should have read the email. So the issue will be voted on next week. 18:09:13 <alanr> is BIJAN one of the options? 18:09:23 <bijan> Ivan did a good job with the email 18:09:24 <alanr> bijan :someValuesFrom 18:10:48 <MarkusK> Subtopic: Issues 21 and 24 18:11:03 <MarkusK> Ian: are we ready to try to resolve those? 18:10:45 <bmotik> +1 to resolve these issues 18:11:40 <MarkusK> AlanR: I think we should announce it on the agenda for next week 18:12:17 <MarkusK> Pfps: when putting the issues on the agenda, the concrete proposals should be made explicit as well, especially for Issue 24 18:12:01 <alanr> 24 reject, no inconsistencies 18:12:41 <MarkusK> Ian: Alan and I will prepare a wording for both proposals 18:12:51 <MarkusK> Subtopic: Issue 111 18:12:51 <MarkusK> (User intent signaling) 18:13:11 <sandro> zakim, who is on the call? 18:13:11 <Zakim> On the phone I see IanH, bmotik, Ivan, Evan_Wallace, MarkusK, uli (muted), msmith, Sandro, bcuencagrau (muted), calvanese (muted), Achille, Peter_Patel-Schneider, Zhe, alanr, 18:13:14 <Zakim> ... bijan, JeffPan, clu (muted), m_schnei (muted) 18:13:59 <MarkusK> AlanR: I thought of the case where someone writes an ontology that needs to be interpreted correctly, e.g. for exchanging messages. The idea of "intents" is not so clear in some other use cases. Some ontology building tools may support only a particular profile of OWL 2, but should they add an intent to their ontologies? The main use case I really see is when one strictly requires specific conclusions to be drawn and specifies an intent for that. 18:15:35 <alanr> Sandro, were you not concerned about this one too? 18:15:50 <sandro> absolutely 18:17:09 <alanr> one question is whether I must use wsdl for message 18:18:19 <bijan> alanr, if not wsdl, some description, perhaps english 18:15:16 <MarkusK> Bijan: I raised this issue but would like to withdraw it now. I agree that the use of "intents" is not always clear. Having intents in ontologies may eventually create more noise. You can always do the work-around of including an OWL Full tautology in your DL, if you want to tell people it's OWL Full. Overall, the issue appears to get more complicated than first expected. Thus I propose to defer that until we may have more experiences. 18:18:40 <m_schnei> conventions might arise outside the WG, where people use an rdfs:comment on the ontology header which tells the profile 18:19:22 <alanr> what about dl versus r? 18:19:56 <MarkusK> Sandro: there are non-entailments for OWL-R that are entailed by other OWL versions. 18:20:12 <bmotik> Not at the OWL R DL side: if you're in OWL R DL, then the entailements coincide with OWL 2 DL 18:20:26 <m_schnei> OWL R Full is both a /syntactic/ AND a /semantic/ subset of Full 18:20:37 <MarkusK> Bijan: I thought that OWL-R would be a syntactic fragment that entails all consequences that the larger fragments would entail. If this is not true for OWL-R Full, I would consider this a bug in OWL-R Full 18:20:48 <alanr> we discussed this in detail at the last f2f 18:21:00 <alanr> so patient... 18:21:01 <bijan> Then I didn't understand it in detail at the last f2f 18:21:06 <bijan> Are there examples? 18:21:41 <pfps> sounds good to me 18:22:07 <m_schnei> q+ to explain OWL R Full vs. OWL Full 18:22:10 <MarkusK> Ian: In OWL-R Full one can state arbitrary DL statements, on account of being "Full", but it would not entail the DL consequences. 18:22:30 <MarkusK> Bijan: I would consider OWL-R Full to be broken then 18:22:47 <MarkusK> Uli: there seems to be a potential misunderstanding here. Sandro asked whether OWL Full ontologies should always signal this. Bijan referred to the option of signalling OWL Full if the interpretation as OWL Full is considered crucial. 18:23:35 <alanr> what about owl-r full versus owl-full 18:24:19 <MarkusK> Michael Schneider: OWL-R Full still lacks some syntactic features of OWL Full. I can give an example for the semantic differences, but it is probably better to give it by email. 18:24:38 <pfps> q+ to ask what syntactic features are missing in OWL-R Full 18:25:17 <MarkusK> Pfps: What syntax is not in OWL-R Full then? 18:25:21 <bijan> Obviously we don't have a clear understanding! 18:25:24 <MarkusK> Michael Schneider: Nominals are not in!? 18:25:45 <MarkusK> Ian: Syntactically nominals are allowed 18:26:03 <MarkusK> Michael Schneider: Indeed 18:25:52 <bcuencagrau> The OWL-R Full specification is telling you what you are allowed to entail 18:26:12 <pfps> OWL R Full allows *all* RDF graphs, but does not provide any "extra" semantics for several constructs that are in OWL (even in OWL DL). 18:26:33 <bijan> Eek! 18:26:34 <m_schnei> yes, I was a bit confused at the moment :) 18:26:39 <bijan> That wasn't my undersatnding 18:26:11 <MarkusK> Topic: Any other business? 18:26:11 <MarkusK> No other business.