Warning:
This wiki has been archived and is now read-only.
Chatlog 2009-04-22
From OWL
See original RRSAgent log and preview nicely formatted version.
Please justify/explain all edits to this page, in your "edit summary" text.
<scribenick> PRESENT:bcuencagrau, bmotik, ewallace, alanr, IanH, MarkusK_, Christine, zimmer, Rinke, pfps, sandro,bijan, msmith, Michael Schneider,uli, Achille, Jie, Zhe <scribenick> CHAIR: Alan Ruttenberg <Christine> REGRETS: Jeff Pan 16:52:50 <RRSAgent> RRSAgent has joined #owl 16:52:50 <RRSAgent> logging to http://www.w3.org/2009/04/22-owl-irc 16:52:55 <Rinke> Zakim, this is owl 16:52:55 <Zakim> Rinke, I see SW_OWL()1:00PM in the schedule but not yet started. Perhaps you mean "this will be owl". 16:53:02 <Rinke> Zakim, this will be owl 16:53:02 <Zakim> ok, Rinke; I see SW_OWL()1:00PM scheduled to start in 7 minutes 16:53:16 <Rinke> RRSAgent, make records public 16:55:44 <Christine> Christine has joined #owl 16:56:28 <Zakim> SW_OWL()1:00PM has now started 16:56:29 <Zakim> +??P4 16:56:44 <MarkusK_> MarkusK_ has joined #owl 16:57:04 <Zakim> SW_OWL()1:00PM has ended 16:57:05 <Zakim> Attendees were 16:57:21 <Zakim> SW_OWL()1:00PM has now started 16:57:22 <Zakim> +Peter_Patel-Schneider 16:57:29 <IanH> IanH has joined #owl 16:58:07 <Zakim> +??P5 16:58:08 <Zakim> -??P5 16:58:08 <Zakim> +??P5 16:58:12 <IanH> Well, in two minutes presumably 16:58:12 <Christine> Zakim, ??P5 is me 16:58:12 <Zakim> +Christine; got it 16:59:08 <Zakim> +IanH 16:59:10 <alanr> alanr has joined #owl 16:59:30 <Zakim> +??P8 16:59:33 <Rinke> zakim, ??P8 is me 16:59:33 <Zakim> +Rinke; got it 16:59:34 <alanr> zakim, who is here? 16:59:35 <Zakim> On the phone I see Peter_Patel-Schneider, Christine, IanH, Rinke 16:59:37 <Zakim> On IRC I see alanr, IanH, MarkusK_, Christine, RRSAgent, Zakim, zimmer, Rinke, pfps, sandro, trackbot 16:59:37 <Rinke> zakim, mute me 16:59:37 <Zakim> Rinke should now be muted 16:59:39 <IanH> IanH has changed the topic to: http://www.w3.org/2007/OWL/wiki/Teleconference.2009.04.22/Agenda 16:59:46 <Zakim> +??P9 16:59:51 <IanH> Zakim, this will be owlwg 16:59:51 <Zakim> ok, IanH; I see SW_OWL()1:00PM scheduled to start in 1 minute 17:00:01 <IanH> RRSAgent, make records public 17:00:05 <ewallace> ewallace has joined #owl 17:00:32 <bmotik> bmotik has joined #owl 17:00:56 <bcuencagrau> bcuencagrau has joined #owl 17:01:13 <IanH> ScribeNick: cgolbrei 17:01:21 <alanr> zakim, who is here? 17:01:21 <Zakim> I notice SW_OWL()1:00PM has restarted 17:01:22 <Zakim> On the phone I see Peter_Patel-Schneider, Christine, IanH, Rinke (muted), ??P9, +3539149aaaa, jar, bmotik 17:01:24 <Zakim> On IRC I see bcuencagrau, bmotik, ewallace, alanr, IanH, MarkusK_, Christine, RRSAgent, Zakim, zimmer, Rinke, pfps, sandro, trackbot 17:01:28 <alanr> zakim, jar is alanr 17:01:28 <Zakim> +alanr; got it 17:01:49 <Zakim> +bmotik.a 17:02:03 <IanH> zakim, who is here? 17:02:03 <Zakim> On the phone I see Peter_Patel-Schneider, Christine, IanH, Rinke (muted), ??P9, +3539149aaaa, alanr, bmotik, bmotik.a 17:02:05 <Zakim> On IRC I see bcuencagrau, bmotik, ewallace, alanr, IanH, MarkusK_, Christine, RRSAgent, Zakim, zimmer, Rinke, pfps, sandro, trackbot 17:02:15 <Zakim> + +1.603.897.aabb 17:02:20 <uli> uli has joined #owl 17:02:28 <bmotik> Zakim, bmotik.a is bcuencagrau 17:02:28 <Zakim> +bcuencagrau; got it 17:02:31 <Zakim> +Evan_Wallace 17:02:32 <bmotik> Zakim, mute me 17:02:33 <Zakim> bmotik should now be muted 17:02:35 <bcuencagrau> Zakim, bmotik.a is bcuencagrau 17:02:35 <Zakim> sorry, bcuencagrau, I do not recognize a party named 'bmotik.a' 17:02:44 <Zhe> Zhe has joined #owl 17:02:54 <bijan> bijan has joined #owl 17:03:00 <bcuencagrau> bcuencagrau has joined #owl 17:03:00 <zimmer> zakim, +1.603.897.aabb is me 17:03:00 <Zakim> +zimmer; got it 17:03:08 <bcuencagrau> Zakim, mute me 17:03:08 <Zakim> bcuencagrau should now be muted 17:03:09 <alanr> zakim, who is here? 17:03:09 <Zakim> On the phone I see Peter_Patel-Schneider, Christine, IanH, Rinke (muted), MarkusK_, +3539149aaaa, alanr, bmotik (muted), bcuencagrau (muted), zimmer, Evan_Wallace 17:03:13 <Zakim> On IRC I see bcuencagrau, bijan, Zhe, uli, bmotik, ewallace, alanr, IanH, MarkusK_, Christine, RRSAgent, Zakim, zimmer, Rinke, pfps, sandro, trackbot 17:03:14 <bmotik> -q 17:03:35 <Zakim> +??P16 17:03:36 <alanr> zakim, who is here? 17:03:37 <Zakim> On the phone I see Peter_Patel-Schneider, Christine, IanH, Rinke (muted), MarkusK_, +3539149aaaa, alanr, bmotik (muted), bcuencagrau (muted), zimmer, Evan_Wallace, ??P16 17:03:40 <Zakim> On IRC I see bcuencagrau, bijan, Zhe, uli, bmotik, ewallace, alanr, IanH, MarkusK_, Christine, RRSAgent, Zakim, zimmer, Rinke, pfps, sandro, trackbot 17:03:41 <Zakim> +??P17 17:03:42 <uli> zakim, ??P16 is me 17:03:43 <Zakim> +uli; got it 17:03:47 <uli> zakim, mute me 17:03:47 <Zakim> uli should now be muted 17:03:58 <bijan> zakim, ??P17 is me 17:03:58 <Zakim> +bijan; got it 17:04:33 <msmith> msmith has joined #owl 17:04:33 <Christine> Topic: Admin 17:04:34 <pfps> alanr: agenda amendments are in the agenda 17:04:59 <Zakim> +msmith 17:05:16 <pfps> minutes look OK 17:05:24 <IanH> alanr: will take amended agenda items at appropriate place in agenda 17:05:43 <alanr> PROPOSED: accept minutes of (15 April) 17:05:46 <Christine> Christine has joined #owl 17:05:51 <Rinke> +1 17:05:52 <IanH> +1 17:05:54 <pfps> +1 17:06:00 <alanr> RESOLVED: accept minutes of (15 April) 17:06:02 <ewallace> +1 17:06:13 <bijan> +1 17:06:23 <Christine> Topic: Documents and Reviewing 17:07:09 <IanH> q+ 17:07:15 <alanr> ack IanH 17:07:51 <IanH> http://www.w3.org/2007/OWL/wiki/Round_6#Publicity_Tracking 17:08:29 <Christine> Ianh: send publicty please 17:08:51 <Zakim> +Sandro 17:08:51 <alanr> q? 17:08:53 <sandro> sorry I'm late..... 17:09:40 <Christine> Subtopic: LC Working Drafts 17:09:41 <Achille> Achille has joined #owl 17:10:07 <pfps> q+ 17:10:12 <Christine> Ianh: minimise changes at this point 17:10:15 <alanr> ack pfps 17:10:19 <Zakim> +[IBM] 17:10:27 <bijan> q+ 17:10:36 <alanr> ack bijan 17:10:45 <Achille> zakim, IBM is me 17:10:45 <Zakim> +Achille; got it 17:11:14 <alanr> q? 17:11:23 <IanH> q+ 17:11:28 <alanr> ack IanH 17:12:45 <IanH> bijan: I wanted to make some changes to XML syntax 17:13:05 <IanH> ian: do it off line and upload after LC period is over 17:13:06 <uli> what about the tables? 17:13:49 <IanH> alanr: accessibility audit? 17:14:00 <IanH> bijan: will try to have it done before CR 17:14:38 <sandro> I'm surprised an empty alt tag is preferable to a brief explanation (eg "UML diagram corresponding to grammar Foo") 17:14:56 <IanH> alanr: bijan will add an action for about 1 month out to do accessibility audit 17:15:27 <IanH> bijan: alt tag provides replacement text; may not be needed 17:15:32 <bijan> Action on: bijan to do accessibility audit in one month 17:15:32 <trackbot> Sorry, couldn't find user - on 17:15:35 <alanr> q? 17:15:37 <sandro> bijan: empty alt tags are, I think, the right thing when the diagrams provide no additional information. 17:15:41 <IanH> alanr: Bijan can report back in 1 month 17:15:50 <Christine> Action: Bijan to about 1 month out to do accessibility audit 17:15:50 <trackbot> Created ACTION-331 - About 1 month out to do accessibility audit [on Bijan Parsia - due 2009-04-29]. 17:16:09 <Christine> Subtopic: First Public or Ordinary Working Drafts 17:16:12 <IanH> alanr: LC till 12 May; short period; tell people they need to comment ASAP 17:16:36 <alanr> qrg, nf&r, primer, doc overvier, manch, datarange extension 17:16:57 <IanH> alanr: status reports on these docs 17:17:06 <Christine> Alanr: QRG ? 17:17:08 <Rinke> NB The Wiki frontpage needs to be updated with links to the correct versions of documents (it currently lists only the old WD's) 17:17:16 <pfps> q+ 17:17:22 <alanr> q? 17:17:25 <alanr> ack pfps 17:17:34 <IanH> pfps: QRG being pulled in several directions 17:17:54 <IanH> ... not sure I can see path forward 17:18:06 <bijan> Common stuff rather than comprehensive 17:18:09 <IanH> ... recent changes going in wrong direction -- e.g., getting longer 17:18:11 <alanr> q? 17:18:21 <Christine> +q 17:18:27 <bijan> I share a lot of pov with peter 17:18:35 <IanH> Me too 17:18:48 <bijan> I concur with Alan on reasonable layout possibility 17:18:59 <bijan> q+ to say something on layout 17:19:03 <alanr> q? 17:19:17 <alanr> ack Christine 17:19:30 <IanH> christine: what process will we use for QRG? 17:19:39 <IanH> ... many different suggestions/directions 17:19:39 <Christine> Alanr: needs review 17:19:43 <alanr> ack bijan 17:19:43 <Zakim> bijan, you wanted to say something on layout 17:20:21 <IanH> bijan: concerned that to achieve right layout may require substantive reworking of content 17:20:31 <IanH> ... at least in order to make CSS reasonable 17:20:54 <IanH> ... review of layout might be next thing to do 17:21:07 <uli> peter, "as we TC"? 17:21:13 <baojie> baojie has joined #owl 17:21:40 <pfps> I'll review QRG 17:21:40 <IanH> alanr: let's have a quick review now -- Peter? 17:21:56 <IanH> ... drafts of layout may be useful 17:22:19 <IanH> ... perhaps using graffle(?) or some such 17:22:31 <IanH> bijan: I would use CSS 17:22:38 <IanH> ... there are decent tools 17:23:04 <sandro> zakim, who is on the call? 17:23:04 <Zakim> On the phone I see Peter_Patel-Schneider, Christine, IanH, Rinke (muted), MarkusK_, +3539149aaaa, alanr, bmotik (muted), bcuencagrau (muted), zimmer, Evan_Wallace, uli (muted), 17:23:07 <Zakim> ... bijan, msmith, Sandro, Achille 17:23:14 <bijan> q+ 17:23:24 <IanH> alanr: Peter, will you do a review? 17:23:28 <Zakim> +baojie 17:23:28 <IanH> pfps: yes 17:23:41 <trackbot> Try using a different identifier, such as family name or username (eg. ppatelsc, phaase) 17:24:05 <Christine> Action: pfps to quick review of QRG 17:24:06 <trackbot> Created ACTION-332 - Quick review of QRG [on Peter Patel-Schneider - due 2009-04-29]. 17:24:32 <IanH> alanr: Jie, any comments on QRG? 17:24:47 <pfps> q+ 17:24:48 <Christine> Jie: will change table layout and links to othet docs 17:24:57 <sandro> I'm torn by this new Link column --- I really like it for on-line usage, but it's a waste of space when printed. 17:25:05 <bijan> display none 17:25:11 <alanr> q? 17:25:15 <alanr> ack bijan 17:25:22 <sandro> (So it falls into the camp, for me, of things to hide when printing. which Bijan had spoken against.) 17:25:25 <uli> +1 to Sandro 17:26:25 <sandro> http://www.w3.org/TR/2009/WD-owl2-quick-reference-20090421/ 17:26:29 <sandro> http://www.w3.org/2007/OWL/wiki/Quick_Reference_Guide 17:26:44 <IanH> bijan: wiki version very different from WD 17:26:44 <bijan> http://www.w3.org/TR/2008/WD-owl2-quick-reference-20081202/ 17:26:57 <sandro> bijan, that's the old one. 17:26:58 <alanr> ack pfps 17:27:04 <bijan> Ok! Thanks! 17:27:27 <bijan> Sandro, so the main web page is wrong? 17:27:29 <uli> +1 to discuss *where* we want the primer to go 17:27:50 <sandro> No, Bijan, your point is still valid. The CSS seems totally different between the wiki and yeterday's publicaiton. 17:27:53 <uli> q+ 17:27:57 <bijan> Yeah 17:27:58 <uli> zakim, unmute me 17:27:58 <Zakim> uli should no longer be muted 17:27:59 <alanr> ack uli 17:28:02 <Rinke> But the main page is *also* wrong! 17:28:08 <bijan> Yes :) 17:28:34 <Christine> uli: should discuss the general directions not the details 17:28:52 <bijan> I agree that we haven't changed our decision 17:29:01 <IanH> q+ 17:29:07 <uli> zakim, unmute me 17:29:07 <Zakim> uli was not muted, uli 17:29:12 <uli> zakim, mute me 17:29:12 <Zakim> uli should now be muted 17:29:26 <bijan> I can't live with 4 pages 17:29:29 <alanr> ack Ianh 17:29:34 <uli> why could it be impossible to live with 2 pages? 17:29:48 <bijan> 4 pages would be a change in our decision 17:29:56 <sandro> PROPOSED; We make QRG fit onto two pages. 17:30:58 <uli> and we agree that the QRG doesn't need to be "complete" 17:31:01 <uli> ? 17:31:01 <bijan> q+ 17:31:09 <alanr> ack bijan 17:31:25 <sandro> http://www.w3.org/TR/2009/WD-owl2-new-features-20090421/ 17:33:21 <Christine> +q 17:33:31 <alanr> ack Christine 17:33:40 <bijan> I'm happy to subsume my follow up to christine's response as a review as long as that doesn't imply I've conceded on it 17:34:12 <Rinke> I can review as well 17:34:12 <bijan> q+ 17:34:21 <alanr> ack bijan 17:35:02 <Christine> bijan: still disagrees appendix 17:35:05 <alanr> q? 17:35:55 <Christine> +q 17:36:09 <alanr> ack Christine 17:37:33 <ewallace> What? 17:37:34 <Christine> +q 17:37:47 <alanr> ack Christine 17:38:17 <IanH> christine: why is much more work required for NF&R? 17:38:20 <bijan> It's obviously false that NF&R is more work than all of the other documents 17:38:24 <bijan> Easy to show 17:38:30 <alanr> not necessary, bijan 17:38:38 <IanH> ... doesn't see need for multiple syntaxes 17:38:45 <bijan> What's not necessary? Refuting falsehoods? 17:39:17 <ewallace> Yes. 17:39:44 <Christine> +q 17:39:48 <sandro> I'm flipping through NF&R and only seeing FS. 17:39:56 <Christine> no 2 syntaxes 17:40:07 <bijan> I'm happy refuting or things being struck from the record. Neither seems bad. 17:40:32 <pfps> Some LCC asked for multiple syntaxes throughout. 17:40:35 <alanr> ack Christine 17:40:35 <sandro> Ah..... Now I see some triples in an example. 17:40:35 <ewallace> A couple of people asked for RDF/XML examples 17:41:08 <sandro> eg the example in http://www.w3.org/TR/2009/WD-owl2-new-features-20090421/#F4:_Self_Restriction does have RDF triples. 17:41:56 <bijan> 7th May is fine for me 17:42:04 <Christine> +q 17:42:15 <alanr> Ack christine 17:42:39 <IanH> q+ 17:42:43 <alanr> ack IanH 17:43:02 <Rinke> I repeat my offer to review NF&R 17:43:14 <Rinke> by next week 17:43:19 <ewallace> thanks Rinke 17:43:56 <Rinke> Action: Rinke to review NF&R 17:43:56 <trackbot> Created ACTION-333 - Review NF&R [on Rinke Hoekstra - due 2009-04-29]. 17:44:56 <Christine> * NF&R will have the record of number of reviews 17:46:33 <alanr> q? 17:46:53 <bijan> q+ 17:46:58 <alanr> ack bijan 17:48:33 <Christine> bijan: have time to work on the Primer (as editor) 17:49:38 <pfps> just some typos to fix in DO 17:49:54 <Christine> Ian : overview in good shape 17:50:15 <Christine> ... in LC shape 17:50:23 <Zakim> +[IPcaller] 17:50:31 <Zakim> -[IPcaller] 17:50:46 <Christine> pfps: Manchester syntax ready 17:51:11 <schneid> schneid has joined #owl 17:52:04 <Zakim> +??P21 17:52:10 <schneid> zakim, ??P21 is me 17:52:10 <Zakim> +schneid; got it 17:52:14 <schneid> zakim, mute me 17:52:14 <Zakim> schneid should now be muted 17:53:19 <uli> schneid, and I thought that German trains would never get cancelled! 17:53:37 <alanr> q? 17:54:00 <Christine> Subtopic : Schedule going forward 17:54:13 <ewallace> me/ is not sure we captured anything about Datarange extension doc status 17:54:17 <bijan> q+ to ask about CR critieria 17:54:36 <alanr> ack bijan 17:54:36 <Zakim> bijan, you wanted to ask about CR critieria 17:55:21 <sandro> q+ 17:55:33 <msmith> i.e., *some* test cases must precede CR 17:56:28 <alanr> ack sandro 17:56:58 <bijan> +1 to polish! 17:57:02 <bijan> And to approving a slew 17:57:20 <bijan> q+ 17:57:34 <alanr> ack bijan 17:57:42 <sandro> sandro: Let's call the impractical tests "extra credit" like webont did. 17:57:51 <msmith> q+ 17:57:56 <alanr> ack msmith 17:58:04 <schneid> I think "Extra Credit" existed for some Full test cases, don't know about DL 17:58:40 <bijan> q+ 17:58:43 <schneid> And I also think that it was very arbitrary what was extra credit and what not 17:58:57 <bijan> q- 17:59:05 <IanH> EC tests also existed for DL; e.g., very large sat problems 17:59:38 <bijan> q+ 17:59:48 <alanr> ack biajn 17:59:55 <IanH> q+ 18:00:01 <IanH> ack bijan 18:00:01 <alanr> ack bijan 18:01:14 <alanr> ack IanH 18:01:32 <bijan> Yeah 18:02:09 <schneid> One reasoner per test case? cool criterion ;-) 18:02:12 <sandro> right -- every non-extra-credit test had to be passed by at least two reasoners. 18:03:05 <bmotik> I don't think we need to discuss this on the call. 18:03:25 <bijan> q+ 18:04:25 <bmotik> Can you summarize the issues? 18:05:41 <alanr> ack bijan 18:05:42 <IanH> q+ 18:05:46 <alanr> ack IanH 18:06:18 <IanH> http://www.w3.org/2007/OWL/wiki/Responses_to_Last_Call_Comments 18:07:01 <Christine> Ianh: has created a new section with new table for LC responses 18:07:02 <Christine> Topic: Issues 18:09:06 <alanr> PROPOSED: close issue 56 with no action 18:09:07 <IanH> +1 18:09:09 <pfps> +1 18:09:11 <alanr> +1 18:09:13 <ewallace> +1 18:09:13 <Rinke> +1 18:09:13 <schneid> +1 18:09:16 <pfps> +56 :-) 18:09:16 <sandro> +1 18:09:19 <bcuencagrau> +1 18:09:21 <MarkusK_> +1 18:09:22 <Zhe> +1 18:09:30 <alanr> RESOLVED: close issue 56 with no action 18:09:37 <bijan> +1 18:09:47 <Christine> Topic: Implementation and Test Cases 18:10:15 <Zakim> -Evan_Wallace 18:11:20 <msmith> q+ 18:11:32 <alanr> ack msmith 18:11:33 <pfps> +1 to adding to *this* table 18:12:11 <pfps> the test results page is post-facto, the implementations page is for use now 18:12:22 <sandro> q+ 18:12:25 <alanr> ack sandro 18:13:21 <bijan> SHER is built on Pellet 18:13:31 <bijan> So I don't know if it's really a separate implementation 18:13:34 <IanH> q+ 18:13:37 <bijan> Though they have a custom EL reasoner 18:13:41 <IanH> ack IanH 18:14:00 <bijan> q+ 18:15:13 <msmith> +1 to no crap 18:15:23 <alanr> q? 18:15:27 <alanr> ack bijan 18:16:16 <alanr> implementations listed in on http://www.w3.org/2007/OWL/wiki/Implementations should be ones that say they will handle OWL 2 18:16:29 <sandro> Maybe next to each implementation we can include a characterization of how much of the test suite it passes. 18:16:37 <sandro> q+ 18:16:49 <alanr> ack sandro 18:17:16 <msmith> or even just a link to the test results page 18:18:21 <bijan> See uli's list o' reasoners: http://owl.cs.manchester.ac.uk/reasoners.html 18:19:22 <bijan> q+ 18:19:50 <MarkusK_> Also see http://semanticweb.org/wiki/Category:Reasoner 18:20:55 <bijan> If comprehensive means at least 1 test mentioning every feature I don't object. :) 18:21:25 <sandro> *shrug* 18:21:28 <msmith> q+ on this 18:21:34 <bijan> q- 18:21:35 <alanr> ack bijan 18:21:39 <alanr> ack msmith 18:21:39 <Zakim> msmith, you wanted to comment on this 18:22:23 <msmith> I agree and we'll keep track of this 18:22:49 <msmith> q+ 18:22:52 <alanr> http://www.w3.org/2007/OWL/wiki/Test_Suite_Status 18:22:55 <alanr> ack msmith 18:23:16 <msmith> http://www.w3.org/2003/08/owl-systems/test-results-out 18:24:41 <bijan> q+ 18:24:45 <alanr> ack bijan 18:28:03 <msmith> q+ 18:28:23 <alanr> ack msmith 18:28:49 <bijan> And with RDF it's extra hard 18:28:54 <pfps> Consider a FS to RDF translator that "shares tails". It would, in some sense, preserve meaning, but it is broken because it can't be reverse translated. 18:29:04 <bijan> Or moves annotations around 18:29:55 <msmith> Sounds good. Thanks. 18:29:55 <sandro> I'm thinking some kind of round-trip, or a->b->b kind of testing.... 18:30:07 <Zakim> -msmith 18:30:08 <Rinke> bye 18:30:08 <Zakim> -Peter_Patel-Schneider 18:30:08 <MarkusK_> bye 18:30:08 <Zakim> -bmotik 18:30:09 <uli> bye 18:30:10 <Zakim> -Sandro 18:30:12 <Zakim> -MarkusK_ 18:30:12 <uli> uli has left #owl 18:30:13 <Zakim> -alanr 18:30:13 <Zakim> -Achille 18:30:14 <Zakim> -baojie 18:30:15 <Zakim> -Rinke 18:30:16 <Zakim> -IanH 18:30:16 <bijan> But you still need a criterion for sameness, sandro 18:30:19 <bijan> THat's the challenge here 18:30:19 <Zakim> -schneid 18:30:22 <Zakim> - +3539149aaaa 18:30:23 <Zakim> -Christine 18:30:24 <Zakim> -bcuencagrau 18:30:26 <Zakim> -bijan 18:30:29 <Zakim> -uli 18:30:32 <Zakim> -zimmer 18:30:33 <Zakim> SW_OWL()1:00PM has ended 18:30:34 <Zakim> Attendees were Peter_Patel-Schneider, Christine, IanH, Rinke, +3539149aaaa, bmotik, alanr, bcuencagrau, Evan_Wallace, MarkusK_, zimmer, uli, bijan, msmith, Sandro, Achille, baojie, 18:30:37 <Zakim> ... [IPcaller], schneid 18:30:46 <alanr> for the rdf mapping testing that a particular set of triples is produced, modulo bnode mapping 18:31:08 <alanr> jena has an "isomorphic to" method. 18:31:12 <bijan> For obnoxious little technical reasons (e.g., that cause roundtripping not to be guarunteed) it's obnoxious to write the test 18:31:20 <bijan> Butyou can have some triple variation in some cases, yes? 18:31:29 <sandro> character-by-character sameness, Bijan. That's not going to work with a->b->a, but it might with a->b1->a2->b2, where b1 and b2 have to be the same. 18:31:32 <bijan> A proper test suite would have to allow for that 18:31:50 <bijan> character by character doesn't work for XML even :) 18:32:08 <alanr> on bnodes, and if in the case where there is explicit order, but structurally order doesn't matter 18:32:09 <bijan> You could impost, maybe, a canonicalization step 18:32:20 <alanr> I agree - test should propery test :) 18:32:35 <alanr> s/propery/properly/ 18:32:44 <bijan> If there's non-determinism 18:33:03 <alanr> I don't remember any in the rdf mapping other than the issue of list ordering I mention 18:33:11 <bijan> Such that two different implementations could, at their discretion, produce different number of triples (for exmaple) 18:33:15 <alanr> that's fs->rdf I'm talking about 18:33:21 <bijan> then your test has to account for unionOf those answers 18:33:27 <alanr> I don't know of any such 18:33:32 <bijan> If not, then we don't need entailment but can use isomorphisic of the graphs 18:33:37 <bijan> Or structural equivalence 18:33:48 <alanr> we don 18:33:57 <alanr> 'we don't have tests for structural equivalence currently 18:34:08 <bijan> WE don't need *tests* for st eq 18:34:09 <alanr> and structural equiv isn't preserved in the rdf mapping 18:34:40 <bijan> Then but you say RDF2FS2RDF=> same graph 18:35:11 <alanr> not necessarily, because of list ordering 18:35:24 <alanr> (at least) 18:35:31 <bijan> Then add a list match to the isomorphism 18:35:35 <alanr> right 18:35:41 <sandro> I'd think translators should preserve list ordering, even though the OWL semantics don't care. 18:35:43 <bijan> I.e., "compare rdf lists as bags" 18:36:06 <alanr> sandro: would be nice, but not what the spec says 18:36:10 <sandro> but that's a perfect test case (whether list ordering can be changed) 18:36:11 <bijan> But it's that sort of thing that's needed, not logical equivalence testing (wich is far too weak) 18:36:46 <alanr> my point about logical equivalence wasn't that it was great, but that it was something. We don't have anything yet, some something is better than nothing. 18:36:51 <sandro> That's a shame (that the spec says list order can be changed). Ah well. 18:37:35 <alanr> What appears worse is that it doesn't appear that you can count on the number of annotations in the structure being preserved in the fs->rdf->f2 18:37:37 <msmith> it says set, not list 18:37:43 <bijan> I disagree that it's something 18:37:50 <alanr> s/f2/fs/ 18:37:57 <alanr> you would be wrong then :) 18:38:17 <alanr> that's a hard argument to make :) 18:38:26 <sandro> rofl 18:38:32 <bijan> You could test for same number of bytes 18:38:34 <bijan> That's a test 18:38:38 <bijan> That's "something" 18:38:46 <msmith> msmith has left #owl 18:38:54 <alanr> not something that if it fails you know the mapping has failed 18:39:30 <alanr> ok got to go - cheers (all in good spirit) 18:39:37 <bijan> Ok, test for byte equality, that succeeds if your conversion succeeded 18:39:45 <bijan> Similarly weak and pointless 18:39:53 <bijan> And he's gone 19:50:09 <alanr> alanr has joined #owl 20:54:02 <Zakim> Zakim has left #owl 21:00:54 <sandro> sandro has joined #owl 21:46:06 <IanH> IanH has joined #owl 22:09:25 <sandro> sandro has joined #owl # SPECIAL MARKER FOR CHATSYNC. DO NOT EDIT THIS LINE OR BELOW. SRCLINESUSED=00000465