Revision as of 11:36, 4 September 2008 by Uli
Please justify/explain all edits to this page, in your "edit summary" text.
00:00:00 <scribenick> PRESENT: Peter_Patel-Schneider, msmith, MartinD (muted), uli (muted), IanH, Sandro, bcuencagrau, JeffP, bmotik (muted), ivan, ewallace, baojie, Zhe, m_schnei, bparsia 00:00:00 <scribenick> REGRETS: Markus Krötzsch, Elisa Kendall, Achille Fokoue, Evan Wallace 00:00:00 <scribenick> CHAIR: IanH 16:45:50 <RRSAgent> RRSAgent has joined #owl 16:45:50 <RRSAgent> logging to http://www.w3.org/2008/09/03-owl-irc 16:46:26 <IanH> IanH has changed the topic to: http://www.w3.org/2007/OWL/wiki/Teleconference.2008.09.03/Agenda 16:46:50 <IanH> Zakim, this will be owlwg 16:46:50 <Zakim> ok, IanH; I see SW_OWL()1:00PM scheduled to start in 14 minutes 16:47:03 <IanH> RRSAgent, make records public 16:55:24 <uli> uli has joined #owl 16:56:15 <msmith> msmith has joined #owl 16:58:07 <Zakim> SW_OWL()1:00PM has now started 16:58:12 <Zakim> +msmith 16:58:15 <Zakim> +Peter_Patel-Schneider 16:58:39 <pfps> pfps has joined #owl 16:59:24 <Zakim> + +0190827aaaa 16:59:28 <MartinD> zakim, aaaa is me 16:59:28 <Zakim> +MartinD; got it 16:59:32 <Zakim> +??P4 16:59:35 <MartinD> zakim, mute me 16:59:35 <Zakim> MartinD should now be muted 16:59:40 <uli> zakim, ??P4 is me 16:59:40 <Zakim> +uli; got it 16:59:44 <uli> zakim, mute me 16:59:44 <Zakim> uli should now be muted 16:59:55 <uli> scribenick uli 16:59:56 <sandro> sandro has joined #owl 17:00:04 <bcuencagrau> bcuencagrau has joined #owl 17:00:36 <JeffP> JeffP has joined #owl 17:00:39 <uli> scribenick: uli 17:00:44 <Zakim> +IanH 17:00:57 <IanH> zakim, who is here? 17:00:57 <Zakim> On the phone I see Peter_Patel-Schneider, msmith, MartinD (muted), uli (muted), IanH 17:01:00 <Zakim> On IRC I see JeffP, bcuencagrau, sandro, pfps, msmith, uli, RRSAgent, Zakim, IanH, bmotik, MartinD, ewallace, trackbot 17:01:06 <Zakim> +Sandro 17:01:10 <uli> ScribeNick: uli 17:01:15 <ivan> ivan has joined #owl 17:01:40 <Zakim> +??P13 17:01:44 <bcuencagrau> Zakim, ??P13 is me 17:01:44 <Zakim> +bcuencagrau; got it 17:02:01 <IanH> zakim, who is here? 17:02:01 <Zakim> On the phone I see Peter_Patel-Schneider, msmith, MartinD (muted), uli (muted), IanH, Sandro, bcuencagrau 17:02:03 <Zakim> On IRC I see ivan, JeffP, bcuencagrau, sandro, pfps, msmith, uli, RRSAgent, Zakim, IanH, bmotik, MartinD, ewallace, trackbot 17:02:05 <Zakim> +StuartTaylor 17:02:16 <Zakim> +??P15 17:02:16 <JeffP> zakim, StuartTaylor is me 17:02:17 <Zakim> +JeffP; got it 17:02:22 <bmotik> Zakim. ??P15 is me 17:02:32 <bmotik> Zakim, mute me 17:02:32 <Zakim> bmotik should now be muted 17:02:34 <IanH> zakim, who is here? 17:02:34 <Zakim> On the phone I see Peter_Patel-Schneider, msmith, MartinD (muted), uli (muted), IanH, Sandro, bcuencagrau, JeffP, bmotik (muted) 17:02:36 <Zakim> On IRC I see ivan, JeffP, bcuencagrau, sandro, pfps, msmith, uli, RRSAgent, Zakim, IanH, bmotik, MartinD, ewallace, trackbot 17:02:59 <ivan> zakim, code? 17:02:59 <Zakim> the conference code is 69594 (tel:+1.617.761.6200 tel:+33.4.89.06.34.99 tel:+44.117.370.6152), ivan 17:02:59 <uli> sure 17:03:04 <baojie> baojie has joined #owl 17:03:07 <Zhe> Zhe has joined #owl 17:03:29 <uli> Topic: Admin 17:03:29 <uli> subtopic: Agenda Amendments 17:03:34 <uli> none 17:03:44 <Zakim> +Danny 17:03:48 <uli> subtopic: Previous minutes 17:03:50 <ivan> zakim, Danny is ivan 17:03:50 <Zakim> +ivan; got it 17:03:51 <Zakim> +baojie 17:03:54 <pfps> minutes look fine to me 17:04:18 <uli> IanH: minutes accepted 17:04:22 <Zakim> +Zhe 17:04:27 <Zhe> zakim, mute me 17:04:27 <Zakim> Zhe should now be muted 17:04:33 <uli> subtopic: Pending actions 17:04:43 <pfps> q+ 17:04:55 <IanH> q? 17:04:59 <IanH> ack pfps 17:05:35 <uli> pfps: action 182 and 183 have empty bodies 17:05:41 <IanH> q? 17:05:58 <uli> IanH: something should be done 17:05:59 <baojie> +q 17:06:11 <uli> pfps: or we say now that they are done 17:06:37 <uli> IanH: we agree that action 182 and 183 are done, even though their bodies are empty 17:06:48 <IanH> q? 17:06:58 <IanH> ack baojie 17:07:09 <uli> baojie: there is an incomplete version on the wiki 17:07:35 <uli> IanH: asks for a pointer to this version 17:07:42 <IanH> q? 17:07:42 <pfps> q+ 17:08:03 <uli> I will run down the corridor and remind bijan 17:08:32 <baojie> An incomplete pdf of Quick Reference Guide: http://www.w3.org/2007/OWL/wiki/Image:Owl2-refcard_2008-08-19.pdf 17:08:47 <uli> back! 17:08:57 <uli> i think so 17:09:13 <uli> IanH: action 150 17:09:22 <IanH> q? 17:09:28 <pfps> q- 17:09:34 <uli> baojie: we have come to a conclusion regarding owl:internationalizedString / rif:text, so it should be done - we changed the specification, I post a link: 17:09:52 <baojie> http://lists.w3.org/Archives/Public/public-owl-wg/2008Sep/0019.html 17:10:42 <uli> IanH: can you come forward with a proposal re. internationalized string? 17:10:44 <bmotik> q+ 17:10:50 <bmotik> Zakim, unmute me 17:10:50 <Zakim> bmotik should no longer be muted 17:10:58 <IanH> q? 17:11:07 <IanH> ack bmotik 17:11:18 <uli> bmotik: I think there is a draft with the basics 17:11:24 <baojie> preliminary spec: http://www.w3.org/2007/OWL/wiki/InternationalizedStringSpec 17:11:45 <m_schnei> m_schnei has joined #owl 17:11:56 <pfps> what is the status of the wiki page, and what should happen to it? 17:12:01 <uli> IanH: who take care of looking at this spec and see how we modify ours? 17:12:06 <IanH> q? 17:12:15 <bmotik> q+ 17:12:17 <uli> ACTION: bmotik to modify OWL spec accordingly 17:12:17 <trackbot> Sorry, couldn't find user - bmotik 17:12:25 <Zakim> +??P21 17:12:28 <bmotik> ACTION: bmotik2 to modify OWL spec accordingly 17:12:28 <trackbot> Created ACTION-206 - Modify OWL spec accordingly [on Boris Motik - due 2008-09-10]. 17:12:36 <m_schnei> zakim, ??P21 is me 17:12:36 <Zakim> +m_schnei; got it 17:12:40 <m_schnei> zakim, mute me 17:12:40 <Zakim> m_schnei should now be muted 17:12:41 <bmotik> q+ 17:12:43 <IanH> q? 17:12:45 <bcuencagrau> Zakim, mute me 17:12:45 <Zakim> bcuencagrau should now be muted 17:12:53 <uli> pfps: it would be odd if, in our spec, we would point to a wiki page 17:13:16 <uli> sandro: we could publsih the (content of) wiki as a working draft 17:13:32 <uli> IanH: as a RIF or as an OWL publication? 17:13:34 <ivan> can be a joined 17:13:35 <IanH> q? 17:13:41 <sandro> sandro: I think it's OKAY as long we're only making the reference from a WD (pre-LC). Maybe we should make it a WD? 17:13:42 <IanH> ack bmotik 17:13:43 <ivan> q+ 17:14:17 <IanH> q? 17:14:19 <IanH> ack ivan 17:14:28 <uli> bmotik: we make the draft a WD and then reference it 17:14:54 <IanH> q? 17:15:04 <uli> ivan: I had a look at this and it looks as if its publication shouldn't cause any problems. 17:15:18 <uli> ivan: we can even have a joint RIF/OWL publication 17:15:44 <bparsia> bparsia has joined #owl 17:16:04 <uli> ACTION: sandro to take this publication plan forward 17:16:04 <trackbot> Created ACTION-207 - Take this publication plan forward [on Sandro Hawke - due 2008-09-10]. 17:16:21 <uli> (I chose sandro already - he said 'yes' first) 17:16:26 <uli> wellcome, ivan 17:16:35 <Zakim> +??P22 17:16:46 <bparsia> zakim, ??p22 is me 17:16:46 <Zakim> +bparsia; got it 17:16:50 <bparsia> zakim, mute me 17:16:50 <Zakim> bparsia should now be muted 17:16:57 <msmith> q+ 17:17:02 <IanH> q? 17:17:03 <uli> IanH: action 192 re. UNA and OWL QL has been done as seen in an email 17:17:07 <IanH> ack msmith 17:17:14 <uli> msmith: yes, we can close that one 17:17:29 <pfps> The consensus should result in a discussion / resolution agenda item for next week. 17:17:44 <IanH> q? 17:17:54 <uli> IanH: action 202 must wait for next week, as must 172 17:18:09 <uli> IanH: I will chase Achille re. 172 17:18:13 <bparsia> I've had no action joy this week 17:18:32 <uli> IanH: action 168 has been on for some time 17:18:37 <IanH> q? 17:18:41 <bparsia> zakim, unmute me 17:18:41 <Zakim> bparsia should no longer be muted 17:18:59 <uli> q+ 17:19:05 <IanH> q? 17:19:20 <bparsia> zakim, mute me 17:19:20 <Zakim> bparsia should now be muted 17:19:24 <IanH> q? 17:19:25 <uli> zakim, unmute me 17:19:25 <Zakim> uli should no longer be muted 17:19:29 <IanH> ack uli 17:20:07 <bparsia> works for me! 17:20:23 <bparsia> zakim, unmute me 17:20:23 <Zakim> bparsia should no longer be muted 17:20:27 <IanH> q? 17:20:38 <uli> bparsia: have done some testing, am waiting for Robert 17:21:01 <uli> uli: perhaps we should see whether there is some w3c official route and not bother Robert 17:21:23 <uli> bparsia: there are some easy problems, e.g., diagrams not alt-ed correctly 17:21:33 <IanH> q? 17:21:49 <uli> sandro: doesn't know of official w3c 'route' 17:22:15 <uli> bparsia: we could do a proper accessibility audit 17:22:32 <uli> IanH: so action 168 remains on you? 17:22:46 <bparsia> zakim, mute me 17:22:46 <Zakim> bparsia should now be muted 17:22:52 <uli> bparsia: couldn't we move it to a general "to-do" list? 17:22:56 <uli> IanH: ok, will do 17:22:58 <bparsia> agreed 17:23:01 <IanH> q? 17:23:14 <bparsia> works for me 17:23:15 <uli> IanH: action 170 is mooted by events 17:23:33 <IanH> q? 17:23:34 <uli> IanH: action 174? 17:23:37 <bparsia> zakim, unmute me 17:23:37 <Zakim> bparsia should no longer be muted 17:23:52 <bparsia> zakim, mute me 17:23:52 <Zakim> bparsia should now be muted 17:23:53 <uli> bparsia: actually yes, bit also might be mooted shortly 17:23:58 <bparsia> yep 17:24:08 <uli> IanH: ok, so we move it by 1 week 17:24:17 <uli> subtopic: Reviewing 17:24:26 <uli> IanH: I saw already some reviews 17:24:32 <m_schnei> yes, thanks for the reviews so far! 17:24:37 <uli> IanH: anybody else? 17:24:40 <pfps> perhaps the review page could be updated as reviews come in? 17:24:47 <uli> IanH: reviews are due on september 8, in 5 days 17:24:52 <bmotik> Zakim, mute me 17:24:52 <Zakim> bmotik should now be muted 17:24:57 <IanH> q? 17:25:00 <bmotik> I just muted me 17:25:04 <uli> zakim, mute me 17:25:04 <Zakim> uli should now be muted 17:25:08 <bmotik> myself 17:25:45 <uli> IanH: a slight problem with the profiles document, other docs should be able to be reviewed by september 8 17:25:58 <pfps> q+ 17:26:11 <IanH> q? 17:26:16 <uli> IanH: the SKOS people have their SKOS reference out for last call 17:26:34 <uli> pfps: I have already produced a review for the SKOS semantics document 17:26:48 <uli> IanH: and this is different from the reference? 17:26:49 <m_schnei> only the SKOS ref is in LC 17:26:58 <ivan> http://www.w3.org/TR/skos-reference 17:27:02 <ivan> SKOS Reference 17:27:14 <IanH> q? 17:27:15 <uli> pfps: forget - I meant powder! 17:27:19 <IanH> ack pfps 17:27:38 <m_schnei> q+ 17:27:42 <uli> IanH: so, volunteers to review LC draft for SKOS reference? 17:27:42 <pfps> -1 17:27:43 <m_schnei> zakim, unmute me 17:27:43 <Zakim> m_schnei should no longer be muted 17:28:15 <m_schnei> zakim, mute me 17:28:15 <Zakim> m_schnei should now be muted 17:28:18 <uli> m_schnei: I started to do a personal look-through, but only with OWL full glasses on, and would prefer to keep it that way 17:28:23 <JeffP> I could try 17:28:23 <ivan> +q 17:28:25 <ivan> q+ 17:28:26 <uli> IanH: anybody else? 17:28:40 <IanH> ack m_schnei 17:28:43 <m_schnei> q- 17:28:51 <IanH> ack ivan 17:28:56 <uli> ivan: the major issue is related to the annotation discussion -- where are we with ours? 17:29:08 <IanH> q? 17:29:14 <m_schnei> but does skos refer to owl 2 at all? 17:29:43 <uli> Ivan: all the rest isn't really complicated, but we should check on issues around annotations 17:30:07 <uli> IanH: ok, I will send emails around to likely suspects 17:30:15 <uli> IanH: F2F4 17:30:18 <m_schnei> true, skos:related and skos:broaderTransitive are intended to be disjoint properties 17:30:30 <uli> subtopic: F2F4 17:30:49 <uli> IanH: you need to book early if you want to profit from special rate 17:31:17 <m_schnei> i found a hotel for about 70EUR in the neighbourhood :) 17:31:29 <uli> sandro: 'special rate' is insane, I suggest to look around in the neighbourhood 17:31:47 <uli> IanH: or you can look around on the internet? 17:32:11 <uli> sandro: but then you don't contribute to the meeting room rates 17:32:38 <IanH> http://www.w3.org/2007/OWL/wiki/F2F4_People 17:32:40 <uli> IanH: and don't forget to register to TPAC 17:32:49 <sandro> s/insane/shockling high, esp in US$/ 17:33:06 <uli> sandro, we can remove all the above 17:33:32 <uli> http://www.w3.org/2002/09/wbs/35125/TPAC2008/ 17:33:33 <bparsia> Perhaps a link to tpac from the f2f4 page? 17:33:35 <sandro> REGISTER HERE: http://www.w3.org/2002/09/wbs/35125/TPAC2008/ 17:33:39 <ivan> there is a link on the wiki page, too 17:34:07 <uli> topic: Issues 17:34:07 <uli> subtopic: Issues 131 and 116 17:34:54 <uli> IanH: Issue 131, 141 and 130 seem to be related, a bit more to discuss on 130. 17:34:56 <IanH> q? 17:35:10 <m_schnei> q+ 17:35:14 <m_schnei> zakim, unmute me 17:35:14 <Zakim> m_schnei was not muted, m_schnei 17:35:16 <uli> IanH: but with the drafts we have in the wiki, perhaps we can resolve 131 and 141 17:35:54 <uli> m_schnei: I am perfectly happy with proposal for 116 17:36:11 <m_schnei> zakim, mute me 17:36:11 <Zakim> m_schnei should now be muted 17:36:24 <uli> IanH: any other opinions? 17:36:38 <Zhe> q+ 17:36:42 <IanH> q? 17:36:44 <m_schnei> q- 17:36:45 <Zhe> zakim, unmute me 17:36:45 <Zakim> Zhe should no longer be muted 17:36:52 <uli> IanH: I have discussed this earlier with Alan, and he seems ok 17:37:02 <uli> i can't hear you, Zhe 17:37:29 <IanH> q? 17:37:36 <IanH> ack zhe 17:37:37 <Zhe> I thought we are waiting for RPI's response on unification idea 17:37:48 <uli> baojie: I didn't follow this discussion closely 17:38:37 <bparsia> +1 to move forward and let people react 17:38:38 <uli> IanH: I have discussed these with Jim, and seems to be fine and he will review the document anyway. 17:38:39 <JeffP> reasonable 17:38:53 <uli> sorry, Zhe, baojie, I couldn't tell who was talking 17:39:04 <Zhe> np 17:39:59 <sandro> from my notes "Alan: Close issue-131 by saying we're happy with the current structure of Profiles. There's one semantics for OWL RL, which the OWL Full semantics...." 17:40:22 <uli> PROPOSED: resolve issue 131 and 116 as per Ian's email http://lists.w3.org/Archives/Public/public-owl-wg/2008Aug/0288.html 17:40:35 <bmotik> +1 17:40:40 <bcuencagrau> +1 17:40:41 <bparsia> +1 17:40:46 <sandro> Sandro: we're still haggling about conformance, which is no longer connected here. 17:40:56 <uli> thanks, sandro 17:41:13 <m_schnei> +1 (FZI) 17:41:33 <pfps> +1 (ALU) 17:41:35 <msmith> +1 17:41:49 <uli> we could be more precise saying "under 1 in Ian's email" 17:41:57 <sandro> +1 (with us being clear that CONFORMANCE is not addressed here) 17:42:01 <IanH> +1 17:42:04 <Zhe> +1 17:42:06 <uli> +1 17:42:16 <baojie> +1 17:42:16 <ivan> +1 17:42:22 <MartinD> +1 17:42:28 <uli> RESOLVED: resolve issue 131 and 116 as per Ian's email http://lists.w3.org/Archives/Public/public-owl-wg/2008Aug/0288.html 17:42:28 <JeffP> +1 17:43:21 <IanH> q? 17:43:24 <uli> IanH: can we have a similar resolution wrt 116? 17:43:53 <uli> PROPOSED: resolve issue 116 as per Ian's email http://lists.w3.org/Archives/Public/public-owl-wg/2008Aug/0288.html 17:44:01 <pfps> +1 17:44:02 <bmotik> +1 17:44:09 <sandro> +1 17:44:11 <sandro> :-) 17:44:29 <JeffP> :-) 17:44:35 <ivan> this just makes the point that we really really resolved it 17:44:53 <uli> IanH: rules generating literals in subject position 17:45:06 <IanH> Q? 17:45:10 <IanH> q? 17:45:12 <uli> IanH: issue 141 17:45:13 <Zhe> q+ 17:45:24 <IanH> ack zhe 17:45:28 <uli> IanH: this is already made clear in the document 17:45:57 <ivan> not predicate but subject position 17:46:08 <uli> Zhe: just to make sure: if this "literal in subject position" happens, what do we do? 17:46:42 <uli> IanH: the rule sets works on a generalization of triples 17:46:46 <IanH> q? 17:47:09 <uli> Zhe: what is the best approach to avoid generation of "illegal rfd triples"? 17:47:32 <JeffP> They are triples but not RDF triples 17:47:39 <uli> IanH: we already say in the spec that these are "generalized" triples, so this is ok and you won't see these since you can't ask for them 17:48:03 <ivan> q+ 17:48:05 <uli> Zhe: I see - so I guess it's fine 17:48:11 <IanH> q? 17:48:13 <ivan> http://www.w3.org/TR/rif-rdf-owl/#End_Notes 17:48:14 <IanH> ack ivan 17:48:19 <uli> ivan: this is editorial - the above is a note regarding the same problem which could be added to the document 17:48:28 <pfps> As far as the basic conformance is concerned, there is no way to tell if the system is generating these generalized triples. 17:49:08 <m_schnei> one implication is that you get with generalized triples every entailment which you got before (without) 17:49:12 <IanH> q? 17:49:44 <uli> PROPOSED: resolve issue 141 as per Peter's email http://lists.w3.org/Archives/Public/public-owl-wg/2008Aug/0203.html 17:49:46 <pfps> +1, surprise :-) 17:49:46 <JeffP> +1 17:49:48 <bmotik> +1 17:49:48 <bijan> +1 17:49:49 <bcuencagrau> +1 17:49:50 <uli> +1 17:49:52 <IanH> +1 17:49:54 <MartinD> +1 17:49:54 <m_schnei> +1 (FZI) 17:49:58 <ivan> +1 17:50:03 <Zhe> +1 17:50:06 <msmith> +1 17:50:34 <sandro> +1 17:50:38 <baojie> +1 17:51:03 <uli> RESOLVED: resolve issue 141 as per Peter's email http://lists.w3.org/Archives/Public/public-owl-wg/2008Aug/0203.html 17:51:22 <IanH> q? 17:51:33 <IanH> q? 17:51:33 <uli> subtopic: Issue 130 17:51:35 <uli> IanH: for issue 130, we have a proposal http://www.w3.org/2007/OWL/wiki/Conformance 17:51:39 <IanH> q? 17:51:49 <sandro> q+ 17:51:53 <IanH> q? 17:52:05 <uli> IanH: so, can we resolve it like this next week? 17:52:06 <bmotik> Zakim, mute me 17:52:06 <Zakim> bmotik was already muted, bmotik 17:52:52 <uli> sandro: I still see the issue that Michael raised, and I would like a simple solution to this 17:52:56 <IanH> q? 17:52:59 <uli> sandro, which problem is this? 17:53:04 <m_schnei> q+ 17:53:08 <sandro> ack sandro 17:53:45 <IanH> ack sandro 17:54:04 <uli> IanH: perhaps sandro has overlooked the precise meaning of this, i.e., that reasoners cannot flip flop between answers 17:54:28 <m_schnei> zakim, unmute me 17:54:28 <Zakim> m_schnei should no longer be muted 17:54:29 <uli> sandro: perhaps the problem isn't so bad 17:54:32 <IanH> q? 17:54:38 <IanH> ack m_schnei 17:54:47 <uli> m_schnei: all I wanted with my remark was to explicate this 17:54:51 <sandro> q? 17:54:54 <uli> m_schnei, what? 17:55:12 <sandro> m_schnei: I just wanted it documented 17:55:32 <IanH> q? 17:55:37 <m_schnei> m_schnei: I want to clarify that I just want to have this conformance behaviour made explicit, I do *not* deny this 17:55:59 <uli> IanH: we should say that, all conformant systems should always agree on their answer 17:56:17 <uli> sandro: what about negative entailments? Do we need another reasoner for them? 17:56:37 <IanH> q? 17:56:53 <IanH> q? 17:57:15 <IanH> q? 17:57:26 <m_schnei> you cannot always say from "false" that the converse is true, in particular not under OWA 17:57:36 <IanH> q? 17:57:39 <uli> sandro: oracle wasn't interested in negative/theorem 1 checks 17:57:50 <sandro> Sandro: Are people going to implement the theorem-1 check? 17:57:53 <uli> Zhe: flexibility for user is a good thing 17:58:30 <uli> Zhe: it will be difficult to tell which rules are bottleneck, so theorem 1 check could be useful 17:58:51 <IanH> q? 17:58:57 <uli> Zhe: I don't know yet what exactly we will implement, but we may implement it 17:59:28 <bijan> SHOULD! 17:59:28 <uli> IanH: for the test, should we strengthen 'may' to 'should'? 17:59:34 <IanH> q? 17:59:37 <ivan> q+ 17:59:58 <sandro> ack ivan 18:00:18 <IanH> q? 18:00:27 <bijan> I'll call at MUST 18:00:33 <uli> ivan: I would prefer 'may' since otherwise the implementor load is too high 18:01:04 <bijan> zakim, unmute me 18:01:04 <Zakim> sorry, bijan, I do not know which phone connection belongs to you 18:01:14 <ivan> q+ 18:01:16 <bparsia> q+ 18:01:17 <uli> sandro: we shouldn't allow reasoners to say 'false' unless it's really false 18:01:19 <IanH> q? 18:01:36 <bparsia> zakim, unmute me 18:01:36 <Zakim> bparsia should no longer be muted 18:01:39 <m_schnei> zhe, even if you only implement a /partial/ /forward/ chainer, then you have an implicit entailment checker: just look in the resulting inference graph and only say "yes", if some entailment is in, and say "no" otherwise 18:01:39 <uli> sandro: we should call that part 'must' and otherwise, use 'unknown' 18:01:42 <bparsia> +1 to sandro's must proposal 18:02:03 <sandro> sandro: How about you MUST do theorem-1 checking before returning FALSE, BUT you can return UNKNOWN if you don't want to do that checking. 18:02:27 <bparsia> zakim, mute me 18:02:27 <Zakim> bparsia should now be muted 18:02:38 <uli> bparsia: I like sandro's suggestion - having this check available will enhance interoperability, and the 'unknown' option is a good compromise 18:02:57 <IanH> q? 18:03:03 <ivan> ack bparsia 18:03:12 <IanH> ack ivan 18:03:16 <uli> IanH: but if we change to "must", then we must explain what implementors could do who wouldn't want to implement the test 18:03:42 <sandro> sandro: absolutely -- we need text here which makes sense to people without thinking it all through at this level. 18:03:54 <IanH> q? 18:03:57 <IanH> q? 18:04:03 <uli> ivan: from Zhe's presentation in Manchester, how would the 'must' work with this? 18:04:04 <Zhe> q+ 18:04:35 <uli> IanH: tricky since we talk about entailments, but we are also interested in queries. So a false is then a 'no, it's really not in the query' 18:04:36 <IanH> q? 18:04:44 <IanH> ack zhe... 18:04:55 <sandro> Ian: in real life, people do query answering. So the "false" is kind of like not answering the query. 18:05:22 <uli> Zhe: I would prefer 'may' since 'should' or 'must' would be a burden 18:05:41 <uli> IanH: but sandro's proposal also allow you to return 'unknown' and this gives us more honesty: 'false' really means false! 18:06:14 <IanH> q? 18:06:15 <bparsia> q+ 18:06:28 <bparsia> (to answer this) 18:06:32 <bparsia> zakim, unmute me 18:06:32 <Zakim> bparsia was not muted, bparsia 18:06:36 <IanH> q? 18:06:36 <uli> Zhe: but in a forward chaining system, where could be return such an 'unknown'? 18:06:39 <IanH> ack bparsia 18:06:48 <sandro> ack bparsia 18:07:14 <uli> bparsia: on load time, or in the query 18:07:15 <IanH> q? 18:07:39 <uli> sandro: so, user asks query 'q', and didn't get a certain result - does this mean that rules couldn't find this result or that it shouldn't be in answer? 18:08:10 <IanH> q? 18:08:14 <m_schnei> q+ 18:08:26 <uli> Zhe: but how would 'unknown' be helpful there? 18:08:27 <IanH> q? 18:09:07 <sandro> sandro: on query results, systems should include a flag saying whether complete reasoning was done or not. that's the equivalent of this false/unknown thing in the conformance definition. 18:09:09 <uli> bparsia: with SPARQL owl, i looked at racerPro and Sher, and there it is important as well to have a mechanism to indicate to the user how complete you are 18:09:24 <m_schnei> zakim, unmute me 18:09:24 <Zakim> m_schnei was not muted, m_schnei 18:09:24 <IanH> q? 18:09:29 <bparsia> zakim, mute me 18:09:29 <Zakim> bparsia should now be muted 18:09:34 <IanH> ack m_schnei 18:09:38 <IanH> q? 18:10:05 <uli> m_schnei, I can't understand you 18:10:52 <sandro> m_schnei: you have to at least implement the full ruleset, and have it not FOL entailed, before you can return FALSE 18:10:55 <uli> heavy breathing 18:11:14 <sandro> (I have a response to m_schnei, but .... maybe I'll save it.) 18:11:45 <IanH> q? 18:11:51 <uli> IanH: using 'unknown' would be a mechanism to indicate to the user that the results to a query may be partial 18:12:06 <uli> Zhe: i don't see the additional value 18:12:26 <bparsia> q+ 18:12:31 <IanH> q? 18:12:33 <uli> IanH: it prevents implementors from having unsound systems and calling them conformant 18:12:34 <m_schnei> m_schnei: you are only allowed to say "False", if the entailment does not exist w.r.t. the /complete/ ruleset. so the NULL reasoner is not allowed. An implementer MAY go beyond the whole ruleset, up to the complete full semantics 18:13:03 <bparsia> zakim, unmute me 18:13:03 <Zakim> bparsia should no longer be muted 18:13:17 <uli> sandro: I would like to have a flag that distinguishes complete from incomplete reasoners 18:13:40 <uli> sandro: but can any OWL RL rule implementation ever be conformant? 18:13:53 <m_schnei> the /ruleset/ is the lower bound of RL conformance 18:13:58 <IanH> q? 18:14:08 <uli> IanH: sure - they are *sound*, we only talk about non-entailments, cases where things are *not* returned 18:14:16 <IanH> q? 18:14:19 <IanH> ack bparsia 18:14:25 <uli> sandro: and then you could use theorem 1 to find complete cases 18:14:31 <m_schnei> btw, if the ruleset entails something, then you can savely say "True", because then OWL Full would produce the same entailment 18:14:49 <sandro> ian: Theorem 1 gives you the completeness guarantee -- it says that if the ontology looks like this, complete-rule-reasoning is complete-ontology-reasoning. 18:15:08 <uli> bparsia: users from bioontology really value complete reasoning, and so we should be able to signal this 18:15:35 <ivan> q+ 18:15:37 <bparsia> zakim, mute me 18:15:37 <Zakim> bparsia should now be muted 18:15:39 <IanH> q? 18:15:41 <uli> IanH: let's take the discussion on-line, implement the suggested modifications and discuss next week 18:15:57 <sandro> q+ to ask if query answering should be covered in Conformance 18:16:07 <sandro> q- ivan 18:16:12 <uli> ivan: i would still like to see the consequences for an implementation being written down 18:16:27 <IanH> q? 18:16:32 <IanH> ack sandro 18:16:32 <Zakim> sandro, you wanted to ask if query answering should be covered in Conformance 18:16:47 <Zhe> q+ 18:17:01 <uli> sandro: let's write it down - but where do we write about query answering? In the conformance document? 18:17:04 <IanH> ack zhe 18:17:21 <bparsia> I'd be open to flagging it as "depeding on implementor feedback" 18:17:33 <bparsia> I'd rather have the stronger and weaken, then do the weaker and then strengthen 18:18:02 <uli> IanH: the tricky bit is the dependency between profiles and conformance: we can't review profiles before we fixed conformance 18:18:04 <bparsia> zakim, unmute me 18:18:04 <Zakim> bparsia should no longer be muted 18:18:10 <IanH> q? 18:18:53 <bparsia> zakim, mute me 18:18:53 <Zakim> bparsia should now be muted 18:19:03 <uli> bparsia: why don't we make conformance really strict (so that poking holes in it is easier) and then review them together 18:19:07 <IanH> q? 18:19:16 <uli> sandro: who updates the draft? 18:19:38 <uli> ACTION: IanH to update the conformance document with 'unkown' 18:19:38 <trackbot> Sorry, couldn't find user - IanH 18:20:12 <IanH> q? 18:20:29 <bparsia> zakim, unmute me 18:20:29 <Zakim> bparsia should no longer be muted 18:20:31 <m_schnei> I already saw the distinct "ox" namespace in the POWDER semantics ;-) 18:21:07 <uli> subtopic: Issue 109 18:21:22 <IanH> q? 18:21:36 <uli> bparsia: it would be good to not have to change namespaces 18:22:11 <IanH> q? 18:22:22 <uli> sandro: can we have a pointer to this 18:22:26 <bparsia> zakim, mute me 18:22:26 <Zakim> bparsia should now be muted 18:22:59 <bmotik> q+ 18:23:01 <bmotik> Zakim, unmute me 18:23:01 <Zakim> bmotik should no longer be muted 18:23:03 <uli> subtopic: issue 138 18:23:06 <IanH> ack bmotik 18:23:28 <ivan> q+ 18:23:29 <msmith> q+ 18:23:29 <bmotik> Zakim, mute me 18:23:30 <IanH> q? 18:23:30 <uli> bmotik: let's use owl:datetime since the datatype is different from the xsd one 18:23:31 <Zakim> bmotik should now be muted 18:23:32 <bparsia> +1 to boris 18:23:58 <uli> ivan: [procedural] didn't we want to ask xsd people about that? 18:24:23 <uli> IanH: didn't sandro want to edit this message from peter? 18:24:36 <pfps> Sandro sent a message, but didn't ask for any action. 18:24:53 <pfps> I'm willing to edit the document, I guess. 18:25:07 <pfps> ?? 18:25:16 <IanH> q? 18:25:19 <uli> IanH: I observe confusion -- pfps, can you edit the mail and send it? To make it more punchy? 18:25:27 <ivan> ack ivan 18:25:39 <uli> sandro: it should say more clearly what they should do. 18:25:42 <msmith> q? 18:25:52 <IanH> q? 18:26:08 <IanH> q? 18:26:09 <uli> IanH: would their answer have any influence of what we do about datetime namespace 18:26:13 <IanH> ack msmith 18:26:49 <uli> msmith: bmotik convinced me that xsd and owl datetime are really different, so perhaps we don't need to waste time by asking them? 18:26:52 <bparsia> +1 18:26:52 <IanH> q? 18:27:01 <bmotik> It already is owl:dateTime. 18:27:02 <IanH> q? 18:27:05 <pfps> +epsilon 18:27:07 <uli> IanH: so msmith suggest to just go ahead with owl:datetime? 18:27:17 <bmotik> I used owl:dateTime in anticipation of this discussion. There is an editorial comment about it. 18:27:24 <pfps> OK by me 18:27:30 <uli> ivan: we should keep the issue open, but use owl:datetime 18:27:40 <bmotik> q+ 18:27:43 <bmotik> Zakim, unmute me 18:27:43 <Zakim> bmotik should no longer be muted 18:27:44 <IanH> q? 18:27:49 <IanH> ack bmotik 18:27:49 <sandro> ack bmotik 18:28:04 <bmotik> Zakim, mute me 18:28:04 <Zakim> bmotik should now be muted 18:28:21 <pfps> +1 18:28:23 <uli> bmotik: we already use owl:datetime, so we can't do anything else on this now 18:28:33 <uli> IanH; AOB? 18:28:33 <Zhe> q+ 18:28:41 <IanH> q? 18:28:47 <IanH> ack Zhe 18:28:54 <uli> Zhe: i want to open an issue about base triples? 18:29:16 <uli> IanH: you raised it, and it is now open, and we can discuss this next week 18:29:24 <uli> IanH: AOB? 18:29:28 <JeffP> thanks, bye 18:29:32 <Zakim> -bmotik 18:29:33 <Zhe> bye 18:29:33 <uli> meeting is closed, thanks 18:29:33 <Zakim> -msmith 18:29:35 <IanH> bye 18:29:36 <Zakim> -JeffP 18:29:37 <Zakim> -Peter_Patel-Schneider 18:29:37 <Zakim> -Zhe 18:29:37 <msmith> msmith has left #owl 18:29:38 <Zakim> -bparsia 18:29:39 <Zakim> -bcuencagrau 18:29:40 <Zakim> -IanH 18:29:41 <Zakim> -uli 18:29:41 <Zakim> -baojie 18:29:41 <sandro> thanks, Ian. :-) 18:29:43 <Zakim> -ivan 18:29:49 <Zakim> -m_schnei 18:29:51 <Zakim> -MartinD 18:29:57 <MartinD> MartinD has left #OWL