Warning:
This wiki has been archived and is now read-only.
Chatlog 2008-10-08
From OWL
See original RRSAgent log and preview nicely formatted version.
Please justify/explain all edits to this page, in your "edit summary" text.
16:59:04 <scribenick> PRESENT: IanH, bijan (muted), bmotik (muted), msmith, Sandro, uli (muted), pfps, ratnesh, Zhe, baojie, Rinke, ewallace, bcuencagrau, alan_ruttenberg, Achille, Christine, Elisa, m_schnei, Carsten 17:02:42 <IanH> ScribeNick: Zhe 17:02:56 <IanH> zakim, who is here? 17:02:56 <Zakim> On the phone I see IanH, bijan (muted), bmotik (muted), msmith, Sandro, uli (muted), Peter_Patel-Schneider, ratnesh, Zhe, baojie, Rinke 17:02:59 <Zakim> On IRC I see Rinke, baojie, ewallace, bcuencagrau, uli, Zhe, bmotik, RRSAgent, Zakim, msmith, IanH, ratnesh, bijan, pfps, alanr, sandro, trackbot 17:03:06 <Rinke> rrsagent, pointer? 17:03:06 <RRSAgent> See http://www.w3.org/2008/10/08-owl-irc#T17-03-06 17:03:22 <Rinke> rrsagent, make records public 17:03:24 <Zhe> Topic: Admin 17:03:35 <IanH> zakim, who is here? 17:03:35 <Zakim> On the phone I see IanH, bijan (muted), bmotik (muted), msmith, Sandro, uli (muted), Peter_Patel-Schneider, ratnesh, Zhe, baojie, Rinke 17:03:37 <Zakim> On IRC I see Rinke, baojie, ewallace, uli, Zhe, bmotik, RRSAgent, Zakim, msmith, IanH, ratnesh, bijan, pfps, alanr, sandro, trackbot 17:03:58 <Zakim> +[IBM] 17:04:29 <Achille> Achille has joined #owl 17:04:36 <Zhe> Agenda amendments. 17:04:36 <Zhe> IanH: add issue 127 to the agenda 17:04:45 <Achille> Zakim, IBM is me 17:04:45 <Zakim> +Achille; got it 17:04:47 <bcuencag> bcuencag has joined #owl 17:04:59 <Zakim> +Evan_Wallace 17:05:02 <cgolbrei> cgolbrei has joined #owl 17:05:17 <Zakim> +??P15 17:05:21 <Zhe> IanH: other suggestions? 17:05:37 <uli> they are a bit laconic, but what can we do 17:05:38 <pfps> previous minutes are acceptable 17:05:45 <Zhe> Proposed: accept previous minutes Oct 1, 2008 17:05:49 <cgolbrei> zakim, +??P15 is cgolbrei 17:05:49 <Zakim> sorry, cgolbrei, I do not recognize a party named '+??P15' 17:05:58 <Elisa> Elisa has joined #owl 17:06:03 <Zhe> Resolved: previous minutes accepted. 17:06:11 <uli> zakim, ??P15 is cgolbrei 17:06:11 <Zakim> +cgolbrei; got it 17:06:12 <Zhe> Topic: Action item status 17:06:24 <Zhe> IanH: SKOS comments 17:06:31 <cgolbrei> zakim, ??P15 is cgolbrei 17:06:31 <Zakim> I already had ??P15 as cgolbrei, cgolbrei 17:06:44 <pfps> q+ 17:06:58 <IanH> q? 17:07:17 <IanH> Alan? 17:07:20 <Zhe> IanH: action-189 review RDF Mapping 17:07:34 <Zhe> IanH: suggest to Alan to drop it 17:07:44 <pfps> suggest dropping 189 as overtaken by events 17:07:59 <Zakim> +??P19 17:08:01 <Zhe> Alan: I will do the review. So just leave it open. It does not have to be done before publishing. 17:08:20 <Bernardo> Bernardo has joined #owl 17:08:23 <Zhe> IanH: action-202: Alan Ruttenberg to have another try at punning proposal in the light of discussion with peter and come up with test cases 17:08:23 <trackbot> http://www.w3.org/2007/OWL/tracker/actions/202 17:08:40 <Zakim> -??P19 17:08:44 <Zhe> Alan: Boris has a proposal worth discussion 17:08:47 <bijan> It's never going to get done 17:08:52 <Zhe> Resolved: close Action-202 17:08:56 <IanH> q? 17:08:57 <bijan> Mooted by events 17:09:01 <pfps> q- 17:09:02 <Zakim> +??P19 17:09:08 <Bernardo> Zakim, ??P19 is me 17:09:08 <Zakim> +Bernardo; got it 17:09:11 <Zhe> IanH: Action-174: Bijan Parsia to provide an rdf serialization for his rich annotation proposal 17:09:32 <Zhe> IanH: Action-207: Sandro Hawke to keep rdf:text publication on track 17:09:32 <trackbot> http://www.w3.org/2007/OWL/tracker/actions/207 17:09:34 <IanH> q? 17:10:13 <Zakim> +Elisa_Kendall 17:10:14 <IanH> q? 17:10:23 <Bernardo> I did 17:10:24 <IanH> q? 17:10:55 <Zhe> IanH: under the impression that not much has happended 17:11:03 <Zhe> sandro: it is moving. 17:11:10 <Zhe> IanH: we will then call the action done. 17:11:20 <IanH> q? 17:11:26 <m_schnei> m_schnei has joined #owl 17:11:26 <Zhe> IanH: action-227: Alan Ruttenberg to email to Elisa and other interested person about metamodel 17:11:26 <trackbot> http://www.w3.org/2007/OWL/tracker/actions/227 17:11:40 <Zhe> alanr: it is done 17:11:45 <pfps> timeframe? 17:12:01 <IanH> q? 17:12:05 <Zhe> Alanr: is it feasible to get it done before F2F4 17:12:33 <pfps> the clock is ticking quite fast here 17:12:42 <Zhe> Elisa: Not sure. I can get the latest models from them. Some portion depends on availability of other people 17:12:35 <ewallace> What tool can load this? 17:12:44 <IanH> q? 17:13:02 <Zhe> IanH: action-217: Jie Bao to get to the RIF to ensure that RDF changes are done properly 17:13:02 <trackbot> http://www.w3.org/2007/OWL/tracker/actions/217 17:13:32 <Zhe> Jie: I need one more week 17:13:45 <Zhe> IanH: Are you confident it can be done by next week? I will then update the due date. 17:13:55 <Zakim> +[IPcaller] 17:14:05 <Zhe> Topic: Brief discussion on F2F4 17:14:09 <m_schnei> zakim, [IPcaller] is me 17:14:09 <Zakim> +m_schnei; got it 17:14:13 <m_schnei> zakim, mute me 17:14:13 <Zakim> m_schnei should now be muted 17:14:34 <Zhe> IanH: Now there is a draft agenda. We need to get it out later today. We may come to it if anyone has comments/suggestion on the agenda 17:15:06 <IanH> q? 17:15:20 <Zhe> Topic: Reviewing and Publishing 17:15:36 <Zakim> +[IPcaller] 17:15:38 <Zhe> sandro: There is one more thing on the todo list. We need to fix broken links. Anyone wants to fix broken links? 17:15:42 <m_schnei> zakim, [IPcaller] is me 17:15:42 <Zakim> +m_schnei; got it 17:15:47 <m_schnei> zakim, mute me 17:15:47 <Zakim> m_schnei should now be muted 17:15:57 <bmotik> q+ 17:16:00 <bmotik> Zakim, unmute me 17:16:00 <Zakim> bmotik should no longer be muted 17:16:00 <IanH> q? 17:16:15 <Zhe> Boris: what links are broken? in reference ? 17:16:22 <Zhe> sandro: There are about 6~7 of them. Most are references. Just fix on the wiki, and run check links. 17:16:50 <Zhe> Boris: I can do that 17:17:20 <Zhe> IanH: we are good to go then 17:17:39 <Zhe> sandro: I think so. I am generating another version now. Not sure if web master can publish today. 17:17:41 <alanr> q+ 17:17:47 <bmotik> Zakim, mute me 17:17:47 <Zakim> bmotik should now be muted 17:17:50 <IanH> q? 17:17:55 <Zhe> IanH: The expectation is to get it ready for pulibshing today. 17:18:00 <bmotik> q- 17:18:09 <IanH> ack alanr 17:18:19 <Zhe> alanr: We are missing one editor comment on the conformance. 17:18:29 <m_schnei> q+ 17:18:33 <IanH> q? 17:18:36 <Zhe> IanH: I am not convinced when looking at the comment. Not sure if it is a problem. 17:18:54 <m_schnei> zakim, unmute me 17:18:54 <Zakim> m_schnei should no longer be muted 17:18:55 <IanH> q? 17:19:10 <Zakim> -Alan 17:19:23 <bmotik> Zakim, mute me 17:19:23 <Zakim> bmotik was already muted, bmotik 17:19:50 <Zhe> m_schnei: datatype map conformance requires at least all of the datatype from OWL 2 datatype maps 17:20:43 <Zhe> IanH: If you only use a subset, it does not make you inconsistent. 17:21:06 <IanH> q? 17:21:17 <Zhe> m_schnei: I will have another look then. 17:21:26 <Zhe> IanH: Maybe I will put in an editor note anyway. If we agree it is ok, then we can take it out. I will do it right after the tele conf 17:21:42 <IanH> q? 17:21:44 <m_schnei> zakim, mute me 17:21:44 <Zakim> m_schnei should now be muted 17:22:05 <m_schnei> works for me 17:22:06 <IanH> q? 17:22:13 <bmotik> q+ to ask a qustion to sandro 17:22:13 <m_schnei> q- 17:22:18 <IanH> q? 17:22:20 <bmotik> Zakim, unmute me 17:22:20 <Zakim> bmotik should no longer be muted 17:22:26 <IanH> ack bmotik 17:22:31 <Zakim> bmotik, you wanted to ask a qustion to sandro 17:22:31 <IanH> q? 17:22:33 <Zhe> bmotik: Sandro, just a brief question, is it just the Profiles needs fixing? 17:22:49 <Zhe> sandro: they are all in profiles doc. 17:22:33 <Zakim> +Alan 17:22:39 <IanH> q? 17:22:46 <alanr> sorry - had to attend to something 17:23:10 <alanr> q+ to ask if someone is taking notes 17:23:14 <Zhe> sandro: It is the fragments (part after URL #) 17:23:24 <IanH> q? 17:23:35 <bmotik> Zakim, mute me 17:23:35 <Zakim> bmotik should now be muted 17:23:36 <alanr> q- 17:23:40 <IanH> q? 17:23:42 <Zhe> IanH: I will add the note to the conformance 17:24:05 <Zhe> IanH: Is there anything else with publishing progress? Done then. 17:24:22 <IanH> q? 17:24:22 <Zhe> IanH: Review Manchester Syntax doc 17:24:34 <Zhe> IanH: Do we want people to review? 17:24:35 <IanH> q? 17:24:48 <Zhe> pfps: It has been reviewed. 17:24:56 <IanH> q? 17:24:57 <Zhe> alanr: I am not toally done. Should be done today. 17:25:03 <IanH> q? 17:25:35 <pfps> There are 2.5 reviews for Manchester. 17:25:46 <pfps> One significant comment - using labels instead of names. 17:25:50 <Zhe> Zhe has joined #owl 17:26:40 <IanH> q? 17:26:41 <Rinke> Review comment from AlanRuttenberg 05:10, 8 October 2008 (UTC) 17:26:41 <Rinke> The use of labels to to replace URIs is central to productive use of Manchester syntax when URIs are not meaningful, as is common in many ontologies. In addition it is often recommended that, in general, URIs not have meaninful information encoded in their strings (see e.g. http://www.w3.org/DesignIssues/Axioms.html). Given this and the goal of making the Manchester syntax readable and user friendly, this specification should say precisely how to use labels in 17:26:42 <Zhe> IanH: Regarding M Syntax, is it anything we can usefully discuss? 17:27:00 <m_schnei> I don 17:27:02 <Zhe> pfps: comment M Syntax use URI as name. Suggestion is to use rdfs:label as the name. 17:27:08 <IanH> q? 17:27:11 <alanr> q+ 17:27:21 <bijan> I don't understand this 17:27:21 <IanH> q? 17:27:25 <IanH> ack alanr 17:27:36 <m_schnei> I don't understand this, too 17:27:51 <Zhe> alanr: In the Protege, many ontologies use opaque ids or URIs. It would be more user friendly if lables are used. I think we know how to do it. I have sent email about how. 17:28:34 <IanH> q? 17:28:39 <pfps> rdfs:label annotation value, not rdfs:comment annotation value 17:28:40 <uli> could you post a link to the email, Alan? 17:29:00 <IanH> q? 17:29:11 <Zhe> alanr: when you look at a class definition, it is not understandable from P4 (Protege) 17:29:14 <IanH> q? 17:29:17 <Zhe> uli: Is this an OWL or a P4/tools issue? 17:29:19 <bijan> q+ 17:29:22 <uli> zakim, unmute me 17:29:22 <Zakim> uli should no longer be muted 17:29:24 <pfps> q+ 17:29:36 <alanr> I think so 17:29:38 <Zhe> uli: are we sure this is really owl issue, but not P4 issue. 17:29:41 <IanH> q? 17:29:42 <alanr> we are defining the format 17:29:46 <bijan> zakim, unmute me 17:29:46 <Zakim> bijan should no longer be muted 17:29:48 <Zhe> IanH: I don't know 17:29:50 <IanH> ack uli 17:29:50 <Zakim> uli, you wanted to ask whether this is an OWL or a P4/tools issue 17:29:55 <m_schnei> I thought that I have used Manchester syntax many times in Topbraid, and never found something missing 17:29:57 <IanH> ack bijan 17:29:57 <uli> zakim, mute me 17:29:58 <Zakim> uli should now be muted 17:30:08 <IanH> q? 17:30:30 <alanr> q+ 17:30:49 <Zhe> bijan: This could be handled by smart editors. It is possible not to make this change. 17:31:03 <IanH> q? 17:31:05 <bijan> zakim, mute me 17:31:05 <Zakim> bijan should now be muted 17:31:53 <Zhe> pfps: This is largely due to editors using different presentation methods 17:32:00 <IanH> q? 17:32:01 <pfps> q- 17:32:06 <IanH> ack alanr 17:32:22 <Zhe> alanr: the motivation to have M syntax is to have an accessible, useful syntax 17:32:26 <bijan> Which is does 17:32:29 <pfps> q+ 17:32:39 <bijan> q+ 17:32:47 <IanH> q? 17:32:54 <IanH> ack pfps 17:33:00 <Zhe> pfps: we introduce a new, weird way to damage ontologies 17:33:10 <alanr> why do we need it? 17:33:24 <alanr> if just for the primer, not justified, imo 17:33:31 <bijan> Oo, that's a good point (Peter's) 17:33:38 <IanH> q? 17:33:41 <bijan> zakim, unmute me 17:33:41 <Zakim> bijan should no longer be muted 17:33:42 <IanH> ack bijan 17:33:48 <pfps> q- 17:33:57 <Zhe> bijan: the current version does meet all criteria Alan said. It is unclear to me that this technique needs to be incorporated 17:34:39 <IanH> q? 17:34:43 <bijan> zakim, mute me 17:34:43 <Zakim> bijan should now be muted 17:35:01 <bijan> q+ 17:35:08 <Carsten> Carsten has joined #owl 17:35:10 <IanH> ack alanr 17:35:10 <Zhe> alanr: If the sole use is for primer, then there is no need for publishing. 17:35:12 <pfps> q+ 17:35:24 <bijan> zakim, unmute me 17:35:24 <Zakim> bijan should no longer be muted 17:35:53 <Zhe> IanH: just a few more comments and then we need to move on 17:35:54 <m_schnei> +1 to ian, I first have to understand the basic problem 17:36:05 <Zhe> bijan: it is 100% a spec. 17:36:08 <alanr> I was reacting to Peter's comment. I propose to publish it as a note to supplement primer, meets our goal and primer requirement. 17:36:39 <pfps> q- 17:36:44 <pfps> Bijan covered my points. We need more compelling argument, if we don't have this feature then the value is gone 17:36:56 <alanr> qq+ 17:36:57 <alanr> q+ 17:37:04 <bijan> zakim, mute me 17:37:04 <Zakim> bijan should now be muted 17:37:13 <IanH> ack alanr 17:37:27 <bijan> I'm fine debating the feature 17:37:30 <Zhe> alanr: it seems quite easy to deal with it 17:37:37 <bijan> I strongly object to the esclation of the significance of it 17:37:38 <Zhe> IanH: We will come back to it. 17:37:52 <Zhe> IanH: Quick reference guide status? Due by end of Oct? 17:37:52 <alanr> I strongly object to your strong objection ;-) 17:38:11 <pfps> there are some reviews in already, but not all 17:38:21 <IanH> q? 17:38:28 <bijan> I object on the grounds that the drama is counterproductive and based on clear inaccuracy (e.g., that it's not a spec). 17:38:34 <Zhe> Elisa: we did get some feedback. Jie is going to respond. That card has been used many times. Feedback is fantastic. Need restructuring and more work. Peter has some good suggestions on re-org. We will try it. Hopefully we can get it done in the few weeks. Our goal is to complete a revision/re-org by f2f 17:39:56 <uli> I would offer to have a look at the design before you implement it 17:40:58 <alanr> uli: email was http://lists.w3.org/Archives/Public/public-owl-wg/2008Sep/0248.html 17:40:59 <IanH> q? 17:41:07 <uli> Elisa, if i wouldn't think that this would be a useful doc I wouldn't have offered 17:41:22 <baojie> Original card by Li and Tim: http://ebiquity.umbc.edu/resource/html/id/94/ 17:41:25 <Zakim> +??P26 17:41:25 <Zhe> IanH: leave it to you and uli to talk offine and work together 17:41:34 <Carsten> zakim, ??pp26 is me 17:41:34 <Zakim> sorry, Carsten, I do not recognize a party named '??pp26' 17:41:34 <IanH> q? 17:41:42 <Carsten> zakim, ??p26 is me 17:41:42 <Zakim> +Carsten; got it 17:41:45 <Carsten> zakim, mute me 17:41:45 <Zakim> Carsten should now be muted 17:41:45 <Zhe> Elisa: we can set a call to have Jie, Elisa, Uli to talk through re-org issues 17:41:46 <uli> sure 17:41:54 <IanH> q? 17:42:11 <Zhe> Topic: Issue discussions 17:42:44 <Zhe> IanH: proposal to resolve issue 130: Conformance, warnings, errors 17:42:48 <trackbot> http://www.w3.org/2007/OWL/tracker/issues/130 17:43:26 <IanH> q? 17:43:38 <Zhe> IanH: anyone against resolving this issue? 17:44:28 <pfps> PROPOSED: close issue 130 as in T&C 17:44:41 <bmotik> +1 17:44:42 <pfps> +1 17:44:44 <m_schnei> +1 (FZI) 17:44:44 <msmith> +1 17:44:45 <sandro> +1 17:44:46 <ewallace> +1 17:44:46 <alanr> +1 17:44:47 <IanH> +1 17:44:47 <bijan> +1 17:44:47 <uli> _1 17:44:48 <Zhe> Zhe: +1 17:44:48 <Rinke> +1 17:44:54 <Achille> +1 17:44:56 <uli> +1 that is 17:44:57 <Elisa> +1 17:45:06 <IanH> RESOLVED: close issue 130 as in T&C 17:45:30 <Zhe> IanH: Issue discussions. lift issue 127 17:45:34 <Zhe> IanH: issue-127: documents contain bits of nary datatype but these are not yet in OWL 2 17:45:34 <trackbot> http://www.w3.org/2007/OWL/tracker/issues/127 17:45:46 <bijan> q+ 17:45:49 <bijan> zakim, unmute me 17:45:50 <Zakim> bijan should no longer be muted 17:45:53 <Zhe> IanH: what are we going to do if we don't have nary in the SPEC? 17:45:56 <IanH> q? 17:46:00 <IanH> ack bijan 17:46:15 <alanr> we have had discussion 17:46:18 <Zhe> bijan: We have hooks right now 17:46:26 <Zhe> IanH: What are the hooks? 17:46:29 <IanH> q? 17:46:42 <Zhe> bijan: For example, we can have a list of property restrictions and then we can have nary predicate, onProperties can take a list. That is the base level. 17:46:43 <alanr> q+ to ask instead what is the status of the n-ary 17:46:53 <IanH> q? 17:47:07 <IanH> q? 17:47:08 <m_schnei> datatype complements also 17:47:23 <Zhe> alanr: Those are the kinds of things I was referring to 17:47:25 <bmotik> q+ 17:47:31 <IanH> ack alanr 17:47:31 <Zakim> alanr, you wanted to ask instead what the status of the n-ary 17:47:55 <bmotik> Zakim, unmute me 17:47:55 <Zakim> bmotik should no longer be muted 17:48:08 <IanH> ack bmotik 17:48:27 <Zhe> bmotik: To be precise, the hooks is the datarange class 17:48:41 <IanH> q? 17:48:46 <Zhe> IanH: What if we don't have it? 17:49:00 <alanr> We have extensibility of datatype map 17:49:06 <m_schnei> q+ 17:49:23 <bmotik> Zakim, mute me 17:49:23 <Zakim> bmotik should now be muted 17:49:34 <m_schnei> zakim, unmute me 17:49:34 <Zakim> m_schnei should no longer be muted 17:49:35 <IanH> q? 17:50:05 <bmotik> bmotik: The Syntax document contains an explanation of what the hooks are in Section 7 and Section 8.4 17:50:29 <Zhe> m_schnei: I want to add that the hook is also on compelmentOf 17:50:31 <m_schnei> zakim, mute me 17:50:31 <Zakim> m_schnei should now be muted 17:50:38 <IanH> q? 17:50:40 <m_schnei> zakim, unmute me 17:50:40 <Zakim> m_schnei should no longer be muted 17:50:47 <Zhe> IanH: Would it be a problem if all have arity 1 17:50:51 <IanH> ack m_schnei 17:51:07 <Zhe> m_schnei: maybe fine 17:51:15 <alanr> q? 17:51:16 <IanH> q? 17:51:16 <m_schnei> zakim, mute me 17:51:17 <Zakim> m_schnei should now be muted 17:51:48 <ewallace> So why don't we just put N-ary in OWL 2? 17:52:09 <alanr> If there is progress made and hope for a spec by f2f, then I suggest we postpone in anticipation. 17:52:13 <msmith> I believe Bijan is referring to this document http://www.w3.org/2007/OWL/wiki/Data_Range_Extension:_Linear_Equations 17:52:17 <Zhe> bijan: Have a paper on how to compute satisfiability. I don't know who else objects the hooks. 17:52:30 <IanH> q? 17:52:36 <alanr> q+ 17:52:42 <IanH> q? 17:53:11 <IanH> q? 17:53:16 <IanH> ack alanr 17:53:17 <Zhe> alanr: I am happy to know the progress. I object because we don't want to add things that are not understandble by reasoners. 17:53:34 <bijan> q+ to ask if this is datatypes 17:53:41 <IanH> q? 17:54:17 <IanH> q? 17:54:59 <msmith> discussion at f2f3 http://www.w3.org/2007/OWL/meeting/2008-07-29#N__2d_ary_datatype 17:55:00 <Zhe> bijan: there were objections (from Boris) on difficulty ground to add nary datatype 17:55:10 <m_schnei> q+ 17:55:16 <IanH> q? 17:55:21 <IanH> ack bijan 17:55:21 <Zakim> bijan, you wanted to ask if this is datatypes 17:55:34 <bmotik> I look forward to learning about your results! 17:55:38 <Zhe> bijan: I hope to convince Boris to include it after working out details. We already have extension point for data type 17:55:41 <m_schnei> zakim, unmute me 17:55:41 <Zakim> m_schnei should no longer be muted 17:55:43 <IanH> q? 17:55:59 <alanr> over objection 17:56:15 <alanr> and still under discussion 17:56:42 <IanH> q? 17:57:00 <alanr> +1 17:57:15 <IanH> q? 17:57:18 <Zhe> IanH: Maybe we can postpone once we make further progress with document 17:57:19 <alanr> (and is hopeful that document will be completed) 17:58:00 <Zhe> m_schnei: I think I remember from last F2F, putting concrete n-ary datatype in has big impact on implementors 17:58:07 <sandro> [ Apologies, I need to run off to another meeting. Enjoy.... ] 17:58:11 <bijan> Yes, we're there :) 17:58:17 <Zakim> -Sandro 17:58:20 <alanr> q+ 17:58:24 <bijan> The conformance problem would be solved by pointing to an additional recommendation 17:58:34 <bijan> Yes! 17:58:42 <IanH> q? 17:58:47 <m_schnei> zakim, mute me 17:58:47 <Zakim> m_schnei should now be muted 17:59:01 <IanH> ack m_schnei 17:59:05 <IanH> ack alanr 17:59:31 <Zhe> alanr: I object to leave it completely open in the document 17:59:33 <uli> I don't think that this will happen, Alan 17:59:41 <alanr> uli: great! 17:59:48 <trackbot> http://www.w3.org/2007/OWL/tracker/issues/144 17:59:51 <m_schnei> m_schnei: I remember the idea was that we put the hooks in the core OWL 2 language, and then (either we or a different WG) specifies certain n-ary datatypes (comparisons, etc.) as "standard extensions" 17:59:48 <Zhe> IanH: issue-144: missing base triple in serialization of axioms with annotations. 17:59:53 <IanH> q? 18:00:01 <m_schnei> q+ 18:00:04 <IanH> q? 18:00:05 <bijan> Let's dump reificatioN! 18:00:05 <m_schnei> zakim, unmute me 18:00:05 <Zakim> m_schnei was not muted, m_schnei 18:00:07 <pfps> no change from last time :-) 18:00:12 <IanH> ack m_schnei 18:00:36 <Zhe> m_schnei: Not having the base triple will cause some syntatic non-monotonicity. After adding annotations will remove original assertions, it is not a problem for FULL semantics because base triple will be re-created. 18:02:04 <Zhe> m_schnei: However, it will be a problem for SPARQL, you have to re-create it 18:01:58 <IanH> q? 18:02:03 <bmotik> q+ 18:02:16 <IanH> q? 18:02:17 <bmotik> Zakim, unmute me 18:02:18 <Zakim> bmotik should no longer be muted 18:02:19 <m_schnei> zakim, mute me 18:02:19 <Zakim> m_schnei should now be muted 18:02:21 <IanH> ack bmotik 18:02:34 <Zhe> bmotik: It still does not solve the problem if the triple is not there. adding a reification rule. 18:03:53 <Zhe> ... tools can always add it. 18:04:14 <Zhe> ... we should say tool should put related triples together in serialization, to address efficiency problem 18:04:46 <Zhe> ... for triple stores 18:02:34 <alanr> what tool would remove the triple? 18:03:08 <alanr> q+ 18:03:11 <IanH> q? 18:03:35 <m_schnei> q+ 18:03:56 <IanH> q? 18:04:12 <IanH> ack alanr 18:04:29 <Zhe> alanr: having a de-reification rule is not feasible. Two versions of ontologies will require two versions of query 18:04:43 <bijan> q+ to ask about this "common scenarios" 18:04:57 <bmotik> q+ 18:05:02 <IanH> q? 18:05:06 <m_schnei> zakim, unmute me 18:05:06 <Zakim> m_schnei should no longer be muted 18:05:10 <uli> when would we then have trouble, Alan? 18:05:13 <IanH> ack m_schnei 18:06:11 <bijan> zakim, unmute me 18:06:11 <Zakim> bijan was not muted, bijan 18:06:14 <Zhe> m_schnei: why not just put it in? 18:06:15 <m_schnei> zakim, mute me 18:06:15 <Zakim> m_schnei should now be muted 18:06:24 <Zhe> bijan: i don't know 18:06:28 <IanH> ack bijan 18:06:28 <Zakim> bijan, you wanted to ask about this "common scenarios" 18:06:31 <IanH> q? 18:06:54 <msmith> m_schnei, just putting it in breaks the RDF -> functional mapping 18:06:57 <m_schnei> q+ 18:07:01 <alanr> its not just axioms. it's entity annotations as well 18:07:27 <IanH> q? 18:08:05 <IanH> q? 18:08:06 <bijan> zakim, mute me 18:08:07 <Zakim> bijan should now be muted 18:08:08 <IanH> ack bmotik 18:08:17 <alanr> q+ 18:08:18 <Zhe> bijan: it is new and not well supported by RDF. we still need some smart technique to process it 18:08:45 <Zhe> bmotik: cannot distinguish axiom wo annotation and axiom with annotation 18:08:54 <IanH> q? 18:08:55 <uli> Bijan, in the above, did "it" mean "OWL"? 18:09:07 <bijan> Axiom annotations 18:09:08 <alanr> they should not be! 18:09:15 <bijan> +1 to boris 18:09:29 <msmith> e.g., SubClassOf(A B) and SubClassOf( Annotation( dc:creator "msmith" ) A B) 18:09:41 <uli> thanks, Bijan 18:09:55 <alanr> msmiths - smileys mess up your message 18:10:07 <IanH> q? 18:10:19 <m_schnei> zakim, unmute me 18:10:19 <Zakim> m_schnei should no longer be muted 18:10:30 <IanH> ack m_schnei 18:11:13 <IanH> q? 18:11:22 <msmith> it would add axioms that didn't exist 18:11:42 <bmotik> q+ 18:11:47 <Zhe> m_schnei: I don't see the problem 18:11:55 <IanH> q? 18:11:56 <m_schnei> zakim, mute me 18:11:56 <Zakim> m_schnei should now be muted 18:11:56 <bmotik> Consider the following axioms: 18:11:58 <IanH> ack alanr 18:12:01 <bmotik> SubClassOf( A B ) 18:12:03 <bmotik> and 18:12:12 <msmith> q+ 18:12:15 <bmotik> SubClassOf( Label("bla") A B ) 18:12:32 <bmotik> The first gets translated into one triple only 18:12:45 <Zhe> alanr: I am confused. it does not even make sense to have an un-annotated axiom and an annotated version in one ontology 18:12:48 <bmotik> The second gets translated into the four triples only. 18:13:00 <uli> Alan, we won't be able to prevent these "duplicate axioms" and I don't think we should! 18:13:00 <IanH> q? 18:13:02 <m_schnei> yes, thats fine 18:13:08 <IanH> ack bmotik 18:13:09 <bijan> q+ to support distinguishing it 18:13:11 <Zhe> bmotik: what do you mean by adding annotation 18:13:37 <Zhe> ... P4 will retract and add a new one 18:13:47 <Zhe> ... however it is beyond the point 18:13:41 <m_schnei> at least, the original semantics have to be the same after a roundtrip through RDF 18:13:51 <alanr> so you have 4 annotations and you add a fifth and you "retract" the 4 annotations and add a *new* axiom with 5 annotations? 18:13:52 <bijan> I wrote SubClassOf( A B ). Then I merge with an ontology that had SubClassOf(Label("bla") A B) 18:14:03 <m_schnei> q+ 18:14:08 <IanH> q? 18:14:09 <bijan> SubClassOf(Label("Bla" A B ) 18:14:09 <uli> e.g., we could have "the same axiom" from different imports and that have been created by different people 18:14:21 <Zhe> bmotik: from a pure definition's perspective, it can happen. It will be strange to forbid it. 18:14:21 <bijan> It'd be nice to notice that there are two! 18:14:22 <alanr> They should be considered the *same* axiom 18:14:29 <msmith> q- 18:14:35 <alanr> q+ 18:14:38 <bijan> E.g., that one was written by me (without a label). 18:14:40 <IanH> q? 18:14:42 <m_schnei> q- 18:14:47 <bijan> The other one wasn't (and has a label) 18:14:55 <bijan> zakim, unmute me 18:14:55 <Zakim> bijan should no longer be muted 18:15:12 <IanH> q? 18:15:17 <IanH> ack bijan 18:15:17 <Zakim> bijan, you wanted to support distinguishing it 18:15:20 <uli> how far would we go? E.g., how much normalization/rewriting would we consider to decide "equality of axioms"? 18:15:20 <IanH> q? 18:15:26 <alanr> if you compare you should see that there is an added annotation to *the* axiom 18:15:44 <IanH> q? 18:15:48 <uli> +1 to Bijan 18:15:49 <alanr> Second case is not at risk 18:15:51 <Zhe> q+ 18:16:18 <bmotik> q+ 18:16:37 <bijan> I didn't undersatnd what alan said was "very clear" 18:16:37 <IanH> ack alanr 18:16:49 <bijan> I meant *merge* not import 18:16:51 <m_schnei> q+ 18:16:53 <bijan> That's why I said *merge* 18:17:04 <bijan> You are kidding 18:17:09 <m_schnei> we have an imports closure 18:17:25 <bijan> I merge ontologies all the time 18:17:29 <IanH> q? 18:17:30 <bijan> Cut and paste 18:17:32 <pfps> q+ to ask why the "know" 18:17:33 <bijan> i've written tools to do it 18:17:36 <bijan> P4 does it 18:17:50 <IanH> q? 18:17:55 <uli> I have done it and seen people doing it (the cut and paste) 18:18:00 <m_schnei> ok, I will write down my points and send it to the list 18:18:05 <pfps> q- 18:18:09 <m_schnei> q- 18:18:19 <IanH> ack Zhe 18:18:27 <bijan> Refactor>>Extract/Remove Axioms will move sets of axioms from one ontology to the other 18:18:48 <uli> Zhe: I have been stressing efficiency for a long time, I don't know whether Boris's suggestion (on putting relevant triples together) is feasible in practice. 18:18:53 <pfps> a number of people have produced countering efficiency claims 18:19:01 <IanH> q? 18:19:37 <IanH> Zhe: why is it useful to distinguish axioms with and without annotations? 18:19:52 <alanr> I am the same when a few skin cells slough off 18:20:07 <alanr> one can certainly redefine identity appropriately 18:20:10 <IanH> q? 18:20:15 <IanH> ack bmotik 18:20:35 <uli> Alan, I think your comparisons don't work 18:20:41 <alanr> because? 18:20:45 <msmith> +1 to boris, the current definition of structural consistency is very nice from a software implementation perspective 18:21:03 <uli> if we have the same axiom from different authors in different files, we might care 18:21:16 <m_schnei> we don't care about perfect roundtrip anymore, remember! 18:21:18 <alanr> not an issue when in different files 18:21:23 <m_schnei> we have "semantic" roundtripping 18:21:26 <bijan> I care about roudntripping 18:21:28 <IanH> q? 18:21:34 <bijan> And preserving it as much as possible 18:21:35 <uli> oups, i forgot 18:21:43 <alanr> if merging two files and you want to preserve source file, then annotate with provenance 18:21:46 <uli> but still, when you copy and paste, then you need this 18:21:54 <uli> this distinction, i mean 18:21:56 <alanr> you can do anything with cut and paste 18:22:07 <IanH> STRAWPOLL: we should add the base triple? 18:22:09 <alanr> it's a matter of comparing what's priority 18:22:10 <pfps> -1 down with base triples 18:22:23 <Zhe> Zhe: +1 with base triples 18:22:27 <msmith> -1 18:22:28 <Bernardo> -1 to base triples 18:22:28 <bijan> -1 18:22:31 <m_schnei> +1 to base triple (we do not care about roundtripping since 6 months or so) 18:22:32 <baojie> 0 (need to think more) 18:22:32 <Achille> 0 18:22:34 <bmotik> -1 to base triples 18:22:34 <alanr> +1 to base triples 18:22:38 <uli> -1 18:22:41 <ewallace> 0 18:22:47 <Rinke> -0 18:22:47 <IanH> -1 18:22:49 <ratnesh> -1 18:22:50 <Carsten> 0 18:22:52 <alanr> (one of Sandro or Ivan would vote +1, I expect) 18:23:01 <pfps> q+ 18:23:09 <pfps> q- 18:23:14 <bmotik> q+ 18:23:19 <Zhe> IanH: we are kind of split 18:23:25 <IanH> q? 18:23:31 <Carsten> 7:2 = split in the middle? 18:23:38 <Zhe> bmotik: if I see a convincing way to roundtrip it 18:23:40 <alanr> I'm not guessing 18:23:48 <msmith> 7:3 I think 18:23:55 <Carsten> sorry 18:24:05 <uli> 8:3? 18:24:08 <msmith> 8:3 18:24:26 <Zhe> IanH: the point is that it is not just one person against the rest 18:24:38 <alanr> different if we count by institution ;-) 18:24:55 <Zhe> IanH: issue-137: Table 4 in RDF mapping introduces incompatibility with OWL 1 18:24:56 <trackbot> http://www.w3.org/2007/OWL/tracker/issues/137 18:24:57 <IanH> q? 18:25:03 <alanr> q+ 18:25:04 <bmotik> -q 18:25:07 <bmotik> q- 18:25:14 <IanH> q? 18:25:19 <IanH> ack alanr 18:26:03 <bijan> Er... that's not a use case 18:26:16 <bijan> "RDF not an XML solution" just isn't a use case 18:26:45 <IanH> q? 18:27:00 <Zhe> alanr: there is no reason to have additional inclusion mechnism other than owl import 18:27:56 <IanH> q? 18:28:04 <Zhe> bijan: i don't agree with it. I am not convinced. 18:28:32 <IanH> q? 18:28:33 <Zhe> bijan: use xml include 18:28:37 <IanH> q? 18:28:42 <IanH> q? 18:28:53 <Zhe> alanr: not happy with it. 18:29:16 <pfps> n-triples as a rec? where is it coming from? 18:29:25 <bijan> n-triples is already a rec 18:29:26 <bijan> q+ 18:29:31 <IanH> q? 18:29:54 <IanH> q? 18:30:16 <alanr> yes - turtle 18:30:19 <IanH> q? 18:30:27 <alanr> it's not a turtle problem 18:30:33 <IanH> q? 18:30:38 <IanH> ack bijan 18:30:43 <alanr> we can do this in owl, folks don't like the solution 18:31:03 <alanr> so there is a compromise offered 18:31:06 <bijan> I don't see any movement will happen...we'll problaby not get consensus 18:31:12 <Zakim> -Evan_Wallace 18:31:14 <Zakim> -msmith 18:31:15 <Zakim> -bijan 18:31:17 <Zakim> -Bernardo 18:31:21 <Rinke> stop don't go! 18:31:35 <pfps> F2F agenda looks good 18:31:36 <Zhe> IanH: F2F4 agenda? anyone looked at it? 18:31:37 <uli> Yes, But i need to look again 18:31:38 <Rinke> +1 18:31:46 <m_schnei> not yet looked at it 18:31:51 <Zakim> -baojie 18:31:55 <bijan> you'll have to explain (in email) how the compromise isn't just your position ...I don't see how it's different 18:32:12 <alanr> my position was to fix the mapping to handle it 18:32:19 <Zhe> IanH: please get back to me with your comments in the next hour. 18:32:20 <alanr> peter offered to have some inclusion mechanism instead 18:32:35 <uli> oh, yes: I was wondering whether the second session of Day 2 is really reserved to repairs...this seems really long