[00:22:52] What does "unser weg ist noch nicht zu ende" mean? [08:30:31] hmmmm.... now the gerrit bot pings me for all patches that i "own" [08:50:29] (PS1) Addshore: More infomative errors in wbeditentity [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75078 [08:58:34] (CR) Tobias Gritschacher: [C: 2] More infomative errors in wbeditentity [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75078 (owner: Addshore) [08:58:58] :) [09:00:31] (Merged) jenkins-bot: More infomative errors in wbeditentity [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75078 (owner: Addshore) [09:04:40] DanielK_WMDE: please have a look at my daily and give some enlightenment [09:10:19] Danwe_WMDE: doing that now [09:10:27] great [09:14:44] liangent: hey, you there? [09:18:10] Denny_WMDE: hi, if you've got time I've got few conceptual questions regarding badges [09:18:18] Danwe_WMDE: you got mail [09:22:22] lazowik: go ahead [09:22:36] DanielK_WMDE: indeed.plenty. [09:22:55] so, I already asked whether there should be a set of allowed badges [09:22:59] (on bugzilla) [09:23:42] another question is how to organize badges in api [09:24:16] i.e. let clients do stuff on their own with the whole array [09:24:28] or provide functions to add/delete? [09:24:46] i thought its conceptual :) [09:24:55] well... [09:25:16] ok, so let's stay with the first [09:25:35] (question) [09:26:07] (CR) Daniel Werner: "(1 comment)" [extensions/DataValues] - https://gerrit.wikimedia.org/r/74599 (owner: Henning Snater) [09:27:03] so the set of badges should be user extensible [09:27:10] the question is how to do this the easiest way [09:27:22] one suggestion from the team is, to simply take wikidata items to represent badges [09:27:22] user? [09:27:31] editor, community [09:27:46] not us deciding what badges there are [09:27:50] yep [09:27:54] that's up to the communities [09:28:04] so it could be an set of items [09:28:07] and that's it [09:28:17] this also allows you to get labels for them [09:28:22] hmm, but one set per wikidata instance or client instance? [09:28:33] per wikidata instance [09:28:52] "featured article" should be the same item for all clients [09:28:58] no matter if english or arabic [09:28:59] then the biggest problem is what to do if a badge is "deprecated" [09:29:13] the question is ambiguous [09:29:24] do you mean if there is a badge saying "this is deprecated" [09:29:29] nope [09:29:34] or if the usage of badge X is considered deprecated? [09:29:41] yep [09:29:48] same as for anything else [09:29:50] nothing [09:29:52] simply when badge is deleted from set of allowed [09:30:03] there would be no set of allowed badges [09:30:11] ah [09:30:12] any item makes an allowed badge [09:30:17] ok [09:30:25] and now that i say it, i am not so sure about it myself [09:30:32] i am channeling some other team members here :) [09:30:50] but i think this goes well with the wiki anything goes idea [09:30:50] great [09:32:53] Denny_WMDE: https://gerrit.wikimedia.org/r/74635 https://gerrit.wikimedia.org/r/74642 [09:41:26] (CR) Aude: "(3 comments)" [extensions/Wikibase] - https://gerrit.wikimedia.org/r/72909 (owner: Pragunbhutani) [09:51:10] lazowik: was that all? [09:51:34] regarding the api: i am personally a fan of add / remove / set as the verbs [09:51:40] how is it done for labels? [09:53:55] (CR) Denny Vrandecic: [C: 2] Fix instantiation issue in DataValueFactory and add deprecation notice [extensions/DataValues] - https://gerrit.wikimedia.org/r/74635 (owner: Jeroen De Dauw) [09:54:07] (Merged) jenkins-bot: Fix instantiation issue in DataValueFactory and add deprecation notice [extensions/DataValues] - https://gerrit.wikimedia.org/r/74635 (owner: Jeroen De Dauw) [09:54:36] (PS7) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [09:54:47] (CR) jenkins-bot: [V: -1] Have DescriptionMatchFinder return instances of EntityId rather then internal ids [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 (owner: Jeroen De Dauw) [09:56:57] (PS1) Aude: add render function to Template class [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75082 [09:57:31] (PS8) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [09:57:43] (CR) jenkins-bot: [V: -1] Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 (owner: Jeroen De Dauw) [09:58:33] Denny_WMDE: I'll take a look [09:58:45] maybe when you come up with something write on buzgzilla [09:58:52] ah, there's another one [09:58:54] order [09:59:31] should it be taken into account? maybe some weighing? [10:00:22] no weighing [10:00:44] do not care about order [10:00:55] ok [10:01:10] and I think you meant aliases? [10:01:28] yes, i did [10:04:07] looks good [10:04:12] will do as aliases [10:05:49] Denny_WMDE: but nevertheless whether there should be a set of allowed of not, badges should be pointing to items? [10:05:55] that's decided? [10:08:29] let us think about it for a second [10:08:39] what would the badges be used for? [10:08:47] FA, GA, maybe stub [10:08:50] anything else? [10:09:17] The advantage of a small and short set that we *do* decide on is that we can make a good simple UI [10:09:40] mhm [10:11:11] (PS2) Daniel Werner: Implemented ValueFormatter for Time DataValue [extensions/DataValues] - https://gerrit.wikimedia.org/r/74599 (owner: Henning Snater) [10:11:13] and it's easier to track spam [10:12:33] yes [10:12:48] until when do you need a decision? [10:12:54] can it go for wednesday? [10:12:57] sure [10:13:17] I've only expanded SimpleSiteLink (merged) and Item (in review) [10:13:44] so it's still easy to change from strings to items [10:13:47] (PS9) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [10:13:57] (CR) jenkins-bot: [V: -1] Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 (owner: Jeroen De Dauw) [10:15:12] (PS1) Daniel Werner: Make use of Time data value formatter in EntityView [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75084 [10:15:27] (PS10) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [10:16:25] (PS3) Daniel Werner: (bug 48937) ValueFormatter for Time DataValue [extensions/DataValues] - https://gerrit.wikimedia.org/r/74599 (owner: Henning Snater) [10:20:50] (PS11) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [10:22:28] (PS4) Michał Łazowik: (bug 40810) Add badges support to Item [extensions/WikibaseDataModel] - https://gerrit.wikimedia.org/r/74676 [10:23:27] (CR) Michał Łazowik: [C: -1] "Do not merge until decision is made whether badges should point to items." [extensions/WikibaseDataModel] - https://gerrit.wikimedia.org/r/74676 (owner: Michał Łazowik) [10:27:29] (PS12) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [10:28:46] (CR) Daniel Kinzler: [C: -1] "(2 comments)" [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75082 (owner: Aude) [10:30:49] addshore: maybe this helps? https://gerrit.wikimedia.org/r/#/c/75054/ [10:31:55] *checks* [10:32:08] jenkins down? [10:32:18] *waits for the merge first* :> [10:34:31] hashar: is jenkins down? [10:34:39] no https://integration.wikimedia.org/zuul/ [10:34:55] hashar: https://integration.wikimedia.org/ci/ is not loading for me [10:35:04] it is for me :> [10:35:14] yeah that page is slow from time to time [10:35:17] haven't figured out why [10:35:21] beside Jenkins needing to die :-] [10:35:28] hashar: and stuff has not run here for 20 minutes: https://gerrit.wikimedia.org/r/#/c/74645/ [10:36:04] hashar: yeah, lets create our own CI system and call is Hashar :D [10:36:28] the root cause is that sometime Jenkins would load the whole build history [10:36:33] out of flat xml files [10:36:50] and the server is quite buggy when it comes to read a ton of files (its very slow) [10:37:12] so I will get the build histories out of Jenkins and have the HTML console log stored as flat file somewhere [10:37:26] huh [10:37:32] hashar: have you see the coveralls stuff? [10:37:44] for now Zuul is waiting for Jenkins to trigger the build and Jenkins is currently busy reading a few thousand XML files :( [10:37:50] coveralls? [10:50:44] aude: is test2.wikipedia a client for test.wikidata ? [10:50:53] addshore: not yet [10:51:09] ahh! :) what is it a client for? xD [10:51:17] wikidata :) [10:51:27] it's a fake english wikipedia [10:51:34] I see :) >> http://grab.by/oFmq [10:51:58] im guessing that error is only ocuring because im trying to use the widget on test2 then? :> [10:51:58] huh [10:52:02] stupid jenkins [10:52:13] as wikidata shouldnt add a sitelink to test2 ? :) [10:52:35] addshore: probably [10:53:03] yup, not in the list of site :) all makes sense now ;p [10:54:02] test2 has a mixed personality [10:54:16] Denny_WMDE: adding it into Wikibase.git creates a dependency cycle on the repo level, that is not nice [10:54:32] Denny_WMDE: we'll also have hassle getting CI to work nicely with this [10:54:36] Both Jenkins and Travis [10:54:56] And I'll add policing costs to make sure people do not mess up the dependencies [10:54:59] JeroenDeDauw: OK, and data model? [10:55:34] aude, what is http://wikidata-test-client.wikimedia.de/ ? :P [10:55:57] test wikipedia [10:56:15] * aude waiting on a subdomain for test wikivoyage [10:56:23] Denny_WMDE: DM only depends on DVs right now. If DTs is added in these, the whole paclage becomes dependent on everything DTs are dependent on (ie ValueValidatros) [10:56:33] aude: should I see this? :P http://wikidata-test-client.wikimedia.de/wiki/Tennis_Equipment [10:56:56] oh gah [10:57:00] xD [10:57:01] * aude resetting the wiki! [10:57:02] Denny_WMDE: I can live with putting it in DM as temporary solution though [10:57:14] it's a problem on labs [10:57:36] what happens? :O [10:57:38] Denny_WMDE: because this dependency situation is not worse then what we have now - atm it might just need DVs, but DVs is bundled together with DTs to begin with [10:58:10] Denny_WMDE: oh, there is one other type of reason not to put it there though [10:58:20] Denny_WMDE: will lose git history for DTs then [10:58:25] JeroenDeDauw: I would appreciate it to not increase the number of components for now, and I promise to keep on top of this discussion and to get a resolution. [10:58:29] Denny_WMDE: Danwe_WMDE is going to scream about that [10:58:29] drop database! [10:59:46] Denny_WMDE: you are saying to delay the repo split? [10:59:57] Denny_WMDE: the number of components actually decreases - it goes from 5 to 4 [11:00:10] JeroenDeDauw: good, then ok :) [11:00:34] Denny_WMDE: ok what? [11:00:42] Diff is a pure PHP library, not a MW extension, right? JeroenDeDauw [11:00:51] JeroenDeDauw: Denny_WMDE: why would I scream about that? [11:00:52] I am going to restart jenkins [11:00:54] JeroenDeDauw: to the split that reduces the number of component [11:01:32] JeroenDeDauw: Denny_WMDE: about not having more components or loosing git history (which would suck big time)? [11:02:37] Danwe_WMDE: you would scream about the git history [11:02:49] Danwe_WMDE: that is what jeroen implied [11:02:57] yes, we should not loose it [11:03:08] (i don't get why we would loose it though, it is still there, just in another place, no?) [11:03:26] (uh, i actually do not want to have that discussion right now though...) [11:03:34] that doesn't help me when I am working on a file and I have to look up who changed a certain line [11:03:36] (CR) jenkins-bot: [V: -1] Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 (owner: Jeroen De Dauw) [11:04:19] (CR) jenkins-bot: [V: -1] Only resetEntities once in GetEntitiesTest [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75054 (owner: Hoo man) [11:04:19] Danwe_WMDE: so what is your verdict on single vs multiple var keywoeds? :D [11:04:41] JeroenDeDauw: So you're talking about the DTs history only? Putting DTs into DataModel? [11:04:47] JeroenDeDauw: I mix it [11:04:50] aude: im guesing the above also has something to do with why I get a 'Not Found' for the api? :> [11:04:57] yay, Jenkins down again -.- [11:05:09] Danwe_WMDE: yeah, just DT history [11:05:20] Danwe_WMDE: geuss the history is not all important here? This component only contains very little stuff anyway [11:05:33] addshore: yeah [11:05:35] JeroenDeDauw: DT should stay in DT. It still is reusable outside of Wikibase, especially with ValueView which should still stay outside of WB but VV depends on DTs! [11:05:40] (CR) Addshore: [C: 2] Only resetEntities once in GetEntitiesTest [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75054 (owner: Hoo man) [11:06:00] * aude setting up our test users and otherwise not open for everyone creating accounts [11:06:02] jekinds dead now or alive again? :) [11:06:16] addshore: Restarting [11:09:48] Danwe_WMDE: you mix it? How inconsistent [11:09:50] ;p [11:10:22] Danwe_WMDE: yeah I know, now convicne Denny_WMDE ;p [11:11:19] Danwe_WMDE: are you fine with ending up with the names data-values/data-values data-values/data-types data-values/value-view and data-values/common? [11:12:13] (PS13) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE]. [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [11:13:16] JeroenDeDauw: I guess that's fine [11:13:29] better as Wikibase\ValueThingies [11:13:39] (CR) jenkins-bot: [V: -1] Only resetEntities once in GetEntitiesTest [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75054 (owner: Hoo man) [11:14:52] (PS1) Jeroen De Dauw: Add PHP coveralls to TravisCI build [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75091 [11:15:06] Tobi_WMDE: https://gerrit.wikimedia.org/r/#/c/75091/ [11:15:30] Danwe_WMDE: are you talking about data-values/common? [11:15:45] Danwe_WMDE: this one will not hold WB specific stuff - that will be in WB [11:15:58] Danwe_WMDE: just things such as geo, time, quanity [11:16:15] Danwe_WMDE: those are also useful outside of WB, for instance in SMW [11:16:27] (CR) jenkins-bot: [V: -1] Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE]. [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 (owner: Jeroen De Dauw) [11:16:39] no, I am talking about thaving a data-values/data-types and VView instead of making them a WB thing. [11:17:11] (CR) jenkins-bot: [V: -1] Add PHP coveralls to TravisCI build [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75091 (owner: Jeroen De Dauw) [11:18:19] JeroenDeDauw: but what's with the dash "-" > [11:18:21] ? [11:18:36] (CR) Tobias Gritschacher: [C: 2 V: 2] Add PHP coveralls to TravisCI build [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75091 (owner: Jeroen De Dauw) [11:18:37] (Merged) Tobias Gritschacher: Add PHP coveralls to TravisCI build [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75091 (owner: Jeroen De Dauw) [11:18:39] DataValues/DataValues? [11:18:59] Danwe_WMDE: I'm using composer names here, those have dashes [11:19:37] Danwe_WMDE: https://packagist.org/explore/ [11:20:07] ok, whatever [11:20:16] Danwe_WMDE: that is what I suggested no? Having a data-values/data-types and a data-values/value-view [11:20:25] (CR) Tobias Gritschacher: [V: 2] "jenkins full of fail" [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75054 (owner: Hoo man) [11:20:31] yes. But you said Denny_WMDE won't like that? [11:21:01] (PS6) Daniel Werner: dv.util.inherit changed so __proto__ of related objects displays a proper name [extensions/DataValues] - https://gerrit.wikimedia.org/r/50753 [11:21:03] Danwe_WMDE: the concern here is that we end up with two more git repos, which Denny_WMDE wants to avoid given the discussion on wikitech [11:21:07] (CR) Tobias Gritschacher: "jenkins full of fail" [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75054 (owner: Hoo man) [11:21:08] (Merged) Tobias Gritschacher: Only resetEntities once in GetEntitiesTest [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75054 (owner: Hoo man) [11:21:50] Danwe_WMDE: so ValueView depends on DTs right? [11:22:59] (CR) Daniel Werner: [C: 1] Implemented MessageProvider for valueView experts [extensions/DataValues] - https://gerrit.wikimedia.org/r/70611 (owner: Henning Snater) [11:23:13] JeroenDeDauw: yes [11:24:16] JeroenDeDauw: If Denny_WMDE wants to avoid more repos, leave DTs and VViews in the current repo, together? [11:25:55] JeroenDeDauw: can you point me to the discussion on wikitech? [11:28:09] Tobi_WMDE: thx! [11:28:36] Lydia_WMDE: there wasn't really much in datavalues.. [11:28:52] ok [11:29:15] [travis-ci] wikimedia/mediawiki-extensions-Wikibase#437 (master - 6c9c499 : jeroendedauw): The build was broken. [11:29:15] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-Wikibase/compare/ee0a60ef01b7...6c9c499a04b1 [11:29:15] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-Wikibase/builds/9347773 [11:30:20] hehe? http://www.wikidata.org/w/index.php?title=Q888&diff=prev&oldid=60685581 [11:39:09] addshore: !hehe [11:47:38] (CR) Tobias Gritschacher: [C: 2] dv.util.inherit changed so __proto__ of related objects displays a proper name [extensions/DataValues] - https://gerrit.wikimedia.org/r/50753 (owner: Daniel Werner) [11:47:49] (Merged) jenkins-bot: dv.util.inherit changed so __proto__ of related objects displays a proper name [extensions/DataValues] - https://gerrit.wikimedia.org/r/50753 (owner: Daniel Werner) [12:15:59] (PS1) Jeroen De Dauw: Fix path to vendor/bin [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75095 [12:16:38] Tobi_WMDE: https://gerrit.wikimedia.org/r/#/c/75095/ [12:16:54] Not had this problem anywhere else... How lovely our Wikibase.git setup is ;p [12:17:25] (CR) Tobias Gritschacher: [C: 2] Fix path to vendor/bin [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75095 (owner: Jeroen De Dauw) [12:19:10] (Merged) Jeroen De Dauw: Fix path to vendor/bin [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75095 (owner: Jeroen De Dauw) [12:24:06] Denny_WMDE: this is nice - https://github.com/twitter/bootstrap/releases [12:28:44] (CR) Tobias Gritschacher: [C: 2] Implemented MessageProvider for valueView experts [extensions/DataValues] - https://gerrit.wikimedia.org/r/70611 (owner: Henning Snater) [12:28:53] (Merged) jenkins-bot: Implemented MessageProvider for valueView experts [extensions/DataValues] - https://gerrit.wikimedia.org/r/70611 (owner: Henning Snater) [12:33:08] (PS1) Jeroen De Dauw: Attempt to fix coveralls build [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75097 [12:33:33] Tobi_WMDE: https://gerrit.wikimedia.org/r/#/c/75097/ [12:34:57] (CR) Tobias Gritschacher: [C: 2 V: 2] Attempt to fix coveralls build [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75097 (owner: Jeroen De Dauw) [12:34:58] (Merged) Tobias Gritschacher: Attempt to fix coveralls build [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75097 (owner: Jeroen De Dauw) [12:43:06] (CR) Tobias Gritschacher: [C: 2] Removed obsolete tests/phpunit.php test runner [extensions/Diff] - https://gerrit.wikimedia.org/r/74845 (owner: Jeroen De Dauw) [12:43:19] (Merged) jenkins-bot: Removed obsolete tests/phpunit.php test runner [extensions/Diff] - https://gerrit.wikimedia.org/r/74845 (owner: Jeroen De Dauw) [12:44:11] [travis-ci] wikimedia/mediawiki-extensions-Wikibase#440 (master - 982c52f : Tobias Gritschacher): The build was broken. [12:44:11] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-Wikibase/compare/7f437e2704bd...982c52f2cce1 [12:44:12] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-Wikibase/builds/9350137 [12:50:42] https://travis-ci.org/wikimedia/mediawiki-extensions-Wikibase/jobs/9350139 funny... the failure I saw locally yesterday [12:52:04] (PS1) Jeroen De Dauw: Attempt to fix coveralls build [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75098 [12:52:06] Tobi_WMDE: ^^ [12:54:02] (CR) Tobias Gritschacher: [C: 2] Attempt to fix coveralls build [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75098 (owner: Jeroen De Dauw) [12:55:25] (Merged) jenkins-bot: Attempt to fix coveralls build [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75098 (owner: Jeroen De Dauw) [13:00:48] (CR) Jeroen De Dauw: [C: -1] (bug 48937) ValueFormatter for Time DataValue [extensions/DataValues] - https://gerrit.wikimedia.org/r/74599 (owner: Henning Snater) [13:00:51] wmse: is going to have a workshop with Wikidata as subject next week ! :O [13:01:43] (CR) Jeroen De Dauw: "(2 comments)" [extensions/DataValues] - https://gerrit.wikimedia.org/r/74599 (owner: Henning Snater) [13:05:17] (PS2) Aude: add render function to Template class [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75082 [13:05:25] [travis-ci] wikimedia/mediawiki-extensions-Wikibase#441 (master - d83aa63 : jenkins-bot): The build is still failing. [13:05:25] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-Wikibase/compare/982c52f2cce1...d83aa6354f7f [13:05:25] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-Wikibase/builds/9350813 [13:05:26] (PS1) Jeroen De Dauw: Add debug into to Travis build to find out why coveralls is not getting invoked [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75099 [13:05:40] Tobi_WMDE: ^ [13:08:29] (CR) Jeroen De Dauw: [C: 2] Add debug into to Travis build to find out why coveralls is not getting invoked [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75099 (owner: Jeroen De Dauw) [13:08:30] (Merged) Jeroen De Dauw: Add debug into to Travis build to find out why coveralls is not getting invoked [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75099 (owner: Jeroen De Dauw) [13:08:52] (PS14) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [13:09:06] (CR) jenkins-bot: [V: -1] Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 (owner: Jeroen De Dauw) [13:10:32] hoo|away: please see my mail on the build faling on the wd tech list [13:11:22] This is bad - most people are running 5.4+mysql [13:11:32] And having randomly faling tests... [13:11:42] Martin Folwer gave a good talk/rage on that [13:19:44] [travis-ci] wikimedia/mediawiki-extensions-Wikibase#442 (master - f5749b0 : jeroendedauw): The build was fixed. [13:19:44] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-Wikibase/compare/d83aa6354f7f...f5749b08a46f [13:19:44] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-Wikibase/builds/9351298 [13:22:44] (PS1) Jeroen De Dauw: Add debug into to Travis build to find out why coveralls is not getting invoked [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75100 [13:22:57] (CR) Jeroen De Dauw: [C: 2 V: 2] Add debug into to Travis build to find out why coveralls is not getting invoked [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75100 (owner: Jeroen De Dauw) [13:22:58] (Merged) Jeroen De Dauw: Add debug into to Travis build to find out why coveralls is not getting invoked [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75100 (owner: Jeroen De Dauw) [13:24:56] JeroenDeDauw: Ok, seen... going to investigate later (as told, I luckily was able to reproduce this) [13:26:21] (PS15) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [13:26:56] Hi. I don't know how old this is: why are there so many items of women with (along with sex:female) sex: male attributed to "Virtual International Authority File"? [13:27:05] https://www.wikidata.org/wiki/Wikidata:Database_reports/Constraint_violations/P21#.22Single_value.22_violations [13:28:30] And men with sex: female attributed to VIAF: https://www.wikidata.org/wiki/Q3354983 [13:28:52] A bug in the bot? [13:29:20] (CR) jenkins-bot: [V: -1] Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 (owner: Jeroen De Dauw) [13:30:34] It's mentioned in [[User_talk:VIAFbot#Wrong_information_in_VIAF]], a blog post: http://hangingtogether.org/?p=2986 [13:30:35] [1] 10https://www.wikidata.org/wiki/User_talk:VIAFbot%23Wrong_information_in_VIAF [13:31:35] (PS16) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [DO NOT MERGE] [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [13:34:47] (PS1) Jeroen De Dauw: Attempt to fix coveralls invocation [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75101 [13:35:18] (CR) Jeroen De Dauw: [C: 2 V: 2] Attempt to fix coveralls invocation [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75101 (owner: Jeroen De Dauw) [13:35:19] (Merged) Jeroen De Dauw: Attempt to fix coveralls invocation [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75101 (owner: Jeroen De Dauw) [13:35:35] Should I boldly remove the wrong sex properties? [13:39:22] hashar: jenkins down? [13:39:28] again [13:39:44] have you looked at the Zuul status page ? [13:40:27] (PS1) Daniel Kinzler: Stop depending on global state for toString. [extensions/WikibaseDataModel] - https://gerrit.wikimedia.org/r/75102 [13:40:31] hashar: any chance this PHP causes problems when run on Jenkins: debug_print_backtrace();exit; [13:41:02] JeroenDeDauw: ---^ [13:41:05] you are goping to like that :) [13:41:12] going even [13:42:15] I have no idea :) [13:42:45] hashar: I submitted this before it went fail earlier and again now [13:43:13] (PS2) Daniel Kinzler: Introducing ReferencedUrlFinder [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74394 [13:43:14] (PS5) Daniel Kinzler: Make ReferencedEntityFinder use Entity::getAllSnaks [extensions/Wikibase] - https://gerrit.wikimedia.org/r/72962 [13:43:14] (PS1) Daniel Kinzler: Improve tests for EntityView. [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75103 [13:43:37] hashar: looks like the build keeps on running [13:43:45] hashar: and blocks the other builds [13:44:07] been running for 15 min, definitly does not take that long normally [13:45:29] (PS3) Daniel Kinzler: Introducing ReferencedUrlFinder [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74394 [13:46:46] aude: does this appear protected or unprotected to you when logged out? http://test.wikidata.org/wiki/Q58 [13:47:07] (CR) Tobias Gritschacher: [C: 2] jQuery.ui.suggester: Proper event handler management [extensions/DataValues] - https://gerrit.wikimedia.org/r/73937 (owner: Henning Snater) [13:47:08] [travis-ci] wikimedia/mediawiki-extensions-Wikibase#444 (master - 4eb9cff : jeroendedauw): The build was broken. [13:47:08] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-Wikibase/compare/f5cff839e642...4eb9cff0e6e5 [13:47:08] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-Wikibase/builds/9352099 [13:47:13] aude: ignore that :D [13:47:16] k [13:47:39] (CR) Tobias Gritschacher: [V: 2] jQuery.ui.suggester: Proper event handler management [extensions/DataValues] - https://gerrit.wikimedia.org/r/73937 (owner: Henning Snater) [13:47:40] (Merged) Tobias Gritschacher: jQuery.ui.suggester: Proper event handler management [extensions/DataValues] - https://gerrit.wikimedia.org/r/73937 (owner: Henning Snater) [13:50:48] DanielK_WMDE_: https://gerrit.wikimedia.org/r/#/c/74277/ [13:51:06] thanks :) [13:53:16] (PS1) Jeroen De Dauw: Remove coveralls build for now as more effort is needed to make it work [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75104 [13:55:38] (CR) Jeroen De Dauw: [C: 2] Stop depending on global state for toString. [extensions/WikibaseDataModel] - https://gerrit.wikimedia.org/r/75102 (owner: Daniel Kinzler) [13:55:45] DanielK_WMDE_: oh, good! [13:55:57] DanielK_WMDE_: then at least ill no longer be blocked by jenkins failing ;p [13:56:16] At least now I know how to kill Jenkins when I feel like it :D [13:57:03] hashar: can you kill the job running for WikibaseQueryEngine? [13:58:06] JeroenDeDauw: done :) [13:59:46] (PS17) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [13:59:57] (PS18) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [14:00:04] (CR) Jeroen De Dauw: [V: 2] Stop depending on global state for toString. [extensions/WikibaseDataModel] - https://gerrit.wikimedia.org/r/75102 (owner: Daniel Kinzler) [14:00:05] (Merged) Jeroen De Dauw: Stop depending on global state for toString. [extensions/WikibaseDataModel] - https://gerrit.wikimedia.org/r/75102 (owner: Daniel Kinzler) [14:02:12] (CR) jenkins-bot: [V: -1] Have DescriptionMatchFinder return instances of EntityId rather then internal ids [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 (owner: Jeroen De Dauw) [14:02:36] I start to fear that the performance per core of this quad core notebook is rather poor #unittests #slooow [14:03:20] (PS7) Jeroen De Dauw: Add getEntityIdFormatter to top level factory. [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74642 [14:03:45] TobiasG_WMDE: can you have a look at https://gerrit.wikimedia.org/r/#/c/74642/ ? [14:04:20] hashar: ty [14:04:44] (PS19) Jeroen De Dauw: Have DescriptionMatchFinder return instances of EntityId rather then internal ids [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 [14:05:05] hashar: hehe "https://integration.wikimedia.org/ci/job/mwext-WikibaseQueryEngine-testextensions-master/None/console : LOST in 30m 32s" [14:05:56] (CR) jenkins-bot: [V: -1] Have DescriptionMatchFinder return instances of EntityId rather then internal ids. [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 (owner: Jeroen De Dauw) [14:08:30] Denny_WMDE1: https://gerrit.wikimedia.org/r/#/c/74642/ https://gerrit.wikimedia.org/r/#/c/74645/ [14:09:37] (CR) Tobias Gritschacher: [C: 2] Preventing template strings from getting parsed in Diff view headings [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74593 (owner: Henning Snater) [14:09:38] (Merged) Tobias Gritschacher: Preventing template strings from getting parsed in Diff view headings [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74593 (owner: Henning Snater) [14:14:22] aude: there? [14:14:30] ? [14:14:52] (CR) Tobias Gritschacher: [C: 2] Add getEntityIdFormatter to top level factory. [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74642 (owner: Jeroen De Dauw) [14:15:03] aude: Remember that unit tests that only ran on SQLite because the duplicate label checher in unit tests is borken on MySQL? [14:15:08] yeah [14:15:16] is it broken, really? [14:15:29] or because of temporary tables does not support something [14:15:56] aude: Well, temporary tables can't be joined, that's why it's awry on MySQL [14:16:46] ok [14:16:51] aude: Anyway... that caused another bug... [14:16:58] hmm? [14:17:00] https://bugzilla.wikimedia.org/show_bug.cgi?id=51809 [14:17:19] (Merged) jenkins-bot: Add getEntityIdFormatter to top level factory. [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74642 (owner: Jeroen De Dauw) [14:17:34] is that another test that needs to be sqlite only? [14:17:53] * aude still needs to fix my sqlite test wiki [14:18:03] Danwe_WMDE: ping [14:18:14] aude: No, I'll work around it [14:18:28] ok [14:20:20] later... food first [14:20:44] ok [14:20:58] (PS2) Tobias Gritschacher: Split off adaptLetterCase and autocompleteString from suggester [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74117 (owner: Henning Snater) [14:21:01] (PS2) Tobias Gritschacher: Split off adaptLetterCase and autocompleteString from suggester [extensions/DataValues] - https://gerrit.wikimedia.org/r/74116 (owner: Henning Snater) [14:22:05] [travis-ci] wikimedia/mediawiki-extensions-Wikibase#445 (master - 3c2f82e : Tobias Gritschacher): The build is still failing. [14:22:05] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-Wikibase/compare/4eb9cff0e6e5...3c2f82e59f72 [14:22:05] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-Wikibase/builds/9353543 [14:31:17] [travis-ci] wikimedia/mediawiki-extensions-Wikibase#446 (master - 10d29a6 : jenkins-bot): The build was fixed. [14:31:17] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-Wikibase/compare/3c2f82e59f72...10d29a6a19a9 [14:31:17] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-Wikibase/builds/9353835 [14:35:53] (PS1) Jeroen De Dauw: Add error handling to EntitiesByPropertyValue [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75108 [14:37:15] (CR) jenkins-bot: [V: -1] Add error handling to EntitiesByPropertyValue [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75108 (owner: Jeroen De Dauw) [14:41:57] (CR) Tobias Gritschacher: [C: 2] Split off adaptLetterCase and autocompleteString from suggester [extensions/DataValues] - https://gerrit.wikimedia.org/r/74116 (owner: Henning Snater) [14:42:03] (CR) Tobias Gritschacher: [C: 2] Split off adaptLetterCase and autocompleteString from suggester [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74117 (owner: Henning Snater) [14:42:06] (Merged) jenkins-bot: Split off adaptLetterCase and autocompleteString from suggester [extensions/DataValues] - https://gerrit.wikimedia.org/r/74116 (owner: Henning Snater) [14:42:44] addshore: if you could review https://gerrit.wikimedia.org/r/#/c/74645/ that would be awesome [14:43:56] (Merged) jenkins-bot: Split off adaptLetterCase and autocompleteString from suggester [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74117 (owner: Henning Snater) [14:46:14] JeroenDeDauw: I think the DataValues review queue is clear now from a JS perspective.. [14:46:36] (PS1) Jeroen De Dauw: Make WikibaseRepo non-final so it can be mocked [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75111 [14:53:27] (CR) Addshore: [C: -1] "(4 comments)" [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75108 (owner: Jeroen De Dauw) [14:55:21] [travis-ci] wikimedia/mediawiki-extensions-Wikibase#447 (master - 59ca555 : Henning Snater): The build was broken. [14:55:21] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-Wikibase/compare/10d29a6a19a9...59ca555d8d39 [14:55:21] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-Wikibase/builds/9354781 [15:09:21] (PS2) Jeroen De Dauw: Add error handling to EntitiesByPropertyValue [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75108 [15:09:33] (CR) Jeroen De Dauw: "(4 comments)" [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75108 (owner: Jeroen De Dauw) [15:10:29] Denny_WMDE1: this commit is sort of hinding work on the query stuff https://gerrit.wikimedia.org/r/#/c/74645/ [15:11:02] (CR) jenkins-bot: [V: -1] Add error handling to EntitiesByPropertyValue [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75108 (owner: Jeroen De Dauw) [15:15:24] (PS1) Tobias Gritschacher: (bug 51813) apply length constraints to properties too [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75115 [15:15:31] (CR) Addshore: "Issue?" [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75108 (owner: Jeroen De Dauw) [15:17:25] (PS2) Tobias Gritschacher: (bug 51813) apply length constraints to properties too [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75115 [15:18:37] (PS3) Tobias Gritschacher: (bug 51813) apply length constraints to properties too [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75115 [15:20:41] (CR) Denny Vrandecic: [C: 2] Have DescriptionMatchFinder return instances of EntityId rather then internal ids. [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 (owner: Jeroen De Dauw) [15:20:51] JeroenDeDauw: confused. I had +2'ed that before [15:20:52] (Merged) jenkins-bot: Have DescriptionMatchFinder return instances of EntityId rather then internal ids. [extensions/WikibaseQueryEngine] - https://gerrit.wikimedia.org/r/74645 (owner: Jeroen De Dauw) [15:21:11] and it had no changes compared to patchset 5,that i reviewed. uh, whatever. [15:21:15] +2'ing again [15:21:20] TobiasG_WMDE: \o/ [15:22:30] Denny_WMDE1: so any objections against me making the further repo split introducing ValueView.git and DataValuesCommon.git now? [15:22:56] (PS3) Jeroen De Dauw: Wired up ByPropertyValue API module [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/74639 [15:23:03] (PS3) Jeroen De Dauw: Add error handling to EntitiesByPropertyValue [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75108 [15:24:49] JeroenDeDauw: not from my side. but talk with aude first, and keep an eye on deployment, also on beta (as with greg-g last time) [15:25:37] (CR) Jeroen De Dauw: [C: 2] "CR'd by Denny" [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/74639 (owner: Jeroen De Dauw) [15:25:37] (Merged) Jeroen De Dauw: Wired up ByPropertyValue API module. [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/74639 (owner: Jeroen De Dauw) [15:25:50] (CR) jenkins-bot: [V: -1] Add error handling to EntitiesByPropertyValue [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75108 (owner: Jeroen De Dauw) [15:26:24] JeroenDeDauw: was this because of the jenkins hiccup? [15:26:46] Denny_WMDE1: is this more important than poking at the wikivoyage stuff, to make sure everything works fine tomorrow? [15:26:57] we'll have a few settings changes etc [15:27:13] then i want to make sure datatypes is okay as a new extension [15:27:27] then if data types is okay, the others should be okay [15:27:39] aude: no, wikivoyage is priority [15:27:48] beta uses same config as deployment, in many/most places [15:27:59] so it means putting this stuff into deployment, for localisation update [15:28:04] k [15:28:20] or we could take wikibase off of beta [15:28:23] ^^JeroenDeDauw so let's keep the split on hold until wikivoyage is done and the datatypes split is done first [15:28:31] ok, thanks [15:28:34] should be soon [15:28:43] would prefer to stay on beta [15:28:49] just if this explodes, then people will blame the wikivoyage stuff [15:28:54] or whatnot [15:29:13] aude: Did you yet test the linkItem widget with wy? I didn't yet find the time to set this up [15:29:18] hoo: doing [15:29:42] not as much luck setting up the test systems with puppet, though almost done [15:29:44] Thanks :) You'll get raised arm, if it works :D [15:29:47] i'm testing in my wikifarm [15:29:51] :) [15:30:04] Denny_WMDE1: aude: ok, lets wait till DTs is handled first [15:30:12] JeroenDeDauw: ok [15:30:16] JeroenDeDauw: thx [15:31:55] (CR) Tobias Gritschacher: [C: 2] Make WikibaseRepo non-final so it can be mocked [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75111 (owner: Jeroen De Dauw) [15:32:10] (CR) Tobias Gritschacher: [C: 2] Remove coveralls build for now as more effort is needed to make it work [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75104 (owner: Jeroen De Dauw) [15:32:11] (PS4) Jeroen De Dauw: Add error handling to EntitiesByPropertyValue [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75108 [15:33:28] (Merged) jenkins-bot: Make WikibaseRepo non-final so it can be mocked [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75111 (owner: Jeroen De Dauw) [15:34:01] (CR) Addshore: [C: 2] Add error handling to EntitiesByPropertyValue [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75108 (owner: Jeroen De Dauw) [15:34:53] (Merged) jenkins-bot: Remove coveralls build for now as more effort is needed to make it work [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75104 (owner: Jeroen De Dauw) [15:35:23] (Merged) jenkins-bot: Add error handling to EntitiesByPropertyValue [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75108 (owner: Jeroen De Dauw) [15:36:08] [travis-ci] wikimedia/mediawiki-extensions-WikibaseQuery#18 (master - b76f872 : jeroendedauw): The build was broken. [15:36:08] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-WikibaseQuery/compare/492f74e63775...b76f8720081f [15:36:08] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-WikibaseQuery/builds/9356591 [15:42:25] addshore: pair programming time? [15:42:40] aude: is this obsolete now? https://gerrit.wikimedia.org/r/#/c/74291/ it has a mmerge conflict [15:43:07] JeroenDeDauw: yes [15:43:33] adding it there can mess with deployments, so needs testing [15:45:11] (CR) Denny Vrandecic: [C: 2] Change return value of LanguageFallbackChain::extractPreferredValue() [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74648 (owner: Liangent) [15:45:55] I *think* the tests from WikibaseDataModel have side effects on the Wikibase repo tests [15:46:02] * hoo cries [15:46:43] (CR) jenkins-bot: [V: -1] Change return value of LanguageFallbackChain::extractPreferredValue() [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74648 (owner: Liangent) [15:51:43] hoo: that's quite possible. they are old, and poorly isolated [15:52:19] I killed jenkins by mistake, sorry [15:54:17] [travis-ci] wikimedia/mediawiki-extensions-Wikibase#448 (master - 89f5509 : jenkins-bot): The build was fixed. [15:54:17] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-Wikibase/compare/59ca555d8d39...89f5509a4a98 [15:54:17] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-Wikibase/builds/9356929 [15:54:20] (CR) Denny Vrandecic: [C: 2] LanguageWithConversion has a new reverseTranslate() function. [extensions/Wikibase] - https://gerrit.wikimedia.org/r/73754 (owner: Liangent) [15:55:19] (CR) jenkins-bot: [V: -1] LanguageWithConversion has a new reverseTranslate() function. [extensions/Wikibase] - https://gerrit.wikimedia.org/r/73754 (owner: Liangent) [15:56:43] (CR) Tobias Gritschacher: [C: 2] Implemented wikibase toolbarlabel and toolbarbutton widget [extensions/Wikibase] - https://gerrit.wikimedia.org/r/71633 (owner: Henning Snater) [15:57:18] JeroenDeDauw: have you thought about https://gerrit.wikimedia.org/r/#/c/72939/ ?? [15:57:37] (PS1) Jeroen De Dauw: Fixed incorrect doc [extensions/WikibaseDatabase] - https://gerrit.wikimedia.org/r/75124 [15:58:13] (CR) jenkins-bot: [V: -1] Implemented wikibase toolbarlabel and toolbarbutton widget [extensions/Wikibase] - https://gerrit.wikimedia.org/r/71633 (owner: Henning Snater) [15:58:45] (CR) Tobias Gritschacher: [V: 2] Implemented wikibase toolbarlabel and toolbarbutton widget [extensions/Wikibase] - https://gerrit.wikimedia.org/r/71633 (owner: Henning Snater) [15:58:46] (Merged) Tobias Gritschacher: Implemented wikibase toolbarlabel and toolbarbutton widget [extensions/Wikibase] - https://gerrit.wikimedia.org/r/71633 (owner: Henning Snater) [15:58:51] (CR) jenkins-bot: [V: -1] Fixed incorrect doc [extensions/WikibaseDatabase] - https://gerrit.wikimedia.org/r/75124 (owner: Jeroen De Dauw) [15:58:56] (CR) Denny Vrandecic: [C: 2] "(1 comment)" [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74653 (owner: Liangent) [16:02:44] (Merged) jenkins-bot: New LanguageFallbackChainFactory functions [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74653 (owner: Liangent) [16:03:08] (CR) Jeroen De Dauw: [C: 2] (bug 51813) apply length constraints to properties too [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75115 (owner: Tobias Gritschacher) [16:03:09] (Merged) Jeroen De Dauw: (bug 51813) apply length constraints to properties too [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75115 (owner: Tobias Gritschacher) [16:10:35] [travis-ci] wikimedia/mediawiki-extensions-Wikibase#449 (master - 24dd4d8 : jenkins-bot): The build is still failing. [16:10:35] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-Wikibase/compare/89f5509a4a98...24dd4d8ffc7b [16:10:35] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-Wikibase/builds/9356963 [16:14:08] [travis-ci] wikimedia/mediawiki-extensions-WikibaseQuery#19 (master - c446608 : jeroendedauw): The build was broken. [16:14:08] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-WikibaseQuery/compare/b76f8720081f...c446608b8ac6 [16:14:08] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-WikibaseQuery/builds/9356988 [16:18:15] DanielK_WMDE: Hey there :) [16:18:25] DanielK_WMDE: Read my last email? [16:23:59] nileshc: i have skimmed it, but decided that it's too hot today to read it closely, and left that for tomorrow [16:24:14] i'm a but worried though about introducing a dependency on another technology [16:24:23] it' [16:24:30] it's not easy to get that deployed [16:24:35] (any maintained) [16:25:51] Yes. I haven't proposed anything concrete though. I just talked about the way I was thinking. We should of course try not having another dependency. [16:26:00] nileshc: i'll think about your query performance issue; if i can't come up with a solution, i can probably come up with some people to ask :) [16:26:34] DanielK_WMDE: Great, thanks. :) BTW, wikibase already has Lucene and SOLR right? [16:26:56] not directly, no [16:27:07] Lucene is an optional dependency inherited from MediaWiki core. [16:27:15] We have experiemnted with solar, but are not using it. [16:27:17] oh..so using either of that to implement what I was talking about would be an additional dependency? [16:28:14] at the moment, yes. but it's likely that mediawiki core will soon move to Solar (or, as it now seems, ElasticSearch) [16:28:41] we also have Redis deployed, and Hodoop is used for analysis, so it's not completely new. [16:29:18] oh, right: if you want to learn about the fuiture of mediawiki search, and discuss performance issues, talk to "manybubble" in #mediawiki [16:29:26] err, "manybubbles", i think [16:29:31] Hmmm...I see...I'll give it a shot in thinking in terms of Redis or Solr too...Redis is as far as I know just another KV store...so the same thing still will apply for it. [16:29:34] he'S the new full time search guy [16:29:59] Alrighty..thanks. [16:30:05] yes, redis probably doesn't help, it's best for communication queues [16:30:38] nileshc: you can also ask Chad (^demon) about the search plans [16:31:43] okay...I'll ping him when he comes online.. [16:32:14] Knowing about the dependencies will help me base my plans in the right path. [16:51:49] In which order is the table that returns mw.wikibase.getEntity() given? [16:57:18] Hi Mi [16:57:35] Paintman hi [17:02:11] There is some logic in the order of property values with the function mw.wikibase.getEntity() ? [18:04:17] Anyone have experience with using MySQL tables having ~700 columns? [18:09:28] nileshc: no, but i heard it is a bad idea, especially for sparse data [18:09:32] is the data sparse? [18:11:05] nope..did you read the email I sent to you, Daniel, Aude? If not I'll explain my use-case.. [18:12:27] Denny_WMDE1: I have this wikidata dataset. If I represent it as a table, it has 12 million rows and 700 columns (if I build a table with items as rows and properties as columns) [18:12:49] I need to run queries like this on it : http://stackoverflow.com/questions/2681408/how-would-i-order-a-table-by-the-number-of-matching-params-in-the-where-clause-o [18:13:53] Denny_WMDE1: I guess I'll need to bring in some NoSQL solution like MongoDB. But then we have a new problem - how do I make SQL queries like that with MongoDB. I can make OR queries in MongoDB easily, but how do I achieve that ORDER BY effect? [18:14:29] I'll be back in 10 mins. [18:15:46] 8,400,000,000 [18:15:48] That's a lot of cells [18:16:01] Denny_WMDE1: Actually the data is indeed sparse, since on average 10-20 (even less I think) cells per row will be filled. [18:16:35] nileshc: as Reedy say. that would be 8.4 B cells [18:16:50] Denny_WMDE1: Actually even less than that! http://www.wikidata.org/wiki/User:Byrial/Statement_statistics [18:16:54] I mean the sparsity. [18:16:54] since it is sparse, a document based db might me better [18:16:55] So effectively KVP [18:17:04] it's 15 M statements [18:17:22] putting them into 8.4 B cells is not optimal, but certainly possible [18:17:35] Reedy: yes [18:17:45] Reedy, KVP? [18:17:50] key value pairs [18:17:54] Ah. yes. [18:18:09] what you would do with a document based db like mongo [18:18:13] nileshc: Anything is useable if you can make a decent/valid case for doing so [18:18:15] als solr is really good at that [18:18:19] even for WMF usage [18:18:19] At the beginning I was thinking about a column oriented store like HBase, or even Mongo is great. [18:18:55] Reedy: Really? I was apprehensive of the extra dependency. [18:19:01] ok [18:19:06] Mongo would be great for this, but I can't figure out this - [18:19:14] http://stackoverflow.com/questions/2681408/how-would-i-order-a-table-by-the-number-of-matching-params-in-the-where-clause-o [18:19:15] I think the Analytic guys use Mongo for something or other [18:19:18] try elasticsearch instead of mongo [18:19:42] it's basically a document based db as well [18:19:50] with each property a vector [18:20:19] I need to compare against many fields - OR like parameters in the query. And I need to sort the results by the number of matching parameters. [18:20:38] can you describe a concrete example query? [18:20:48] That can be done in SQL as explained in the stackoverflow page...one second, I'm writing it here. [18:21:22] SELECT * FROM users ORDER BY (username = 'eitan') + (email = 'eitan@eitan.com') + (company = 'eitan') DESC [18:21:41] Please check out the link if you have any confusion. [18:21:54] (PS3) Denny Vrandecic: Change return value of LanguageFallbackChain::extractPreferredValue() [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74648 (owner: Liangent) [18:21:57] full table scans :D [18:22:09] (CR) Denny Vrandecic: [C: 2] Change return value of LanguageFallbackChain::extractPreferredValue() [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74648 (owner: Liangent) [18:22:47] hoo: Exactly. [18:23:15] but these won't be strings...all the cells would be integers. [18:24:47] (Merged) jenkins-bot: Change return value of LanguageFallbackChain::extractPreferredValue() [extensions/Wikibase] - https://gerrit.wikimedia.org/r/74648 (owner: Liangent) [18:25:50] Doesn't make it any better.. [18:26:20] * much [18:26:55] Doesn't much it any better? [18:27:16] Doesn't make it much better :P [18:27:23] :D [18:28:24] I wonder how the performance will be :P The other option is my initial plan - mongo/hbase + mapreduce [18:33:11] (PS3) Aude: add render function to Template class [extensions/Wikibase] - https://gerrit.wikimedia.org/r/75082 [18:36:13] the unit tests drive me crazy... I've already tried to disable all tests cases which could add the string "spam" into the DB and it still appears there ggrrr [18:38:28] Maybe we should just remove the unique index from the table or disable the test which lets the error show up [18:42:17] Denny_WMDE1, Reedy: Does WMF already use ElasticSearch in any of the projects? [18:42:22] AFAIK it's not used in wikibase? [18:42:58] solr or elasticsearch will be used for wikimedia sooner than later [18:43:08] the two of them are basically interchangeable [18:44:19] nileshc: just checked the query on stackoverflow [18:45:30] Oh. I see. [18:45:30] i am not sure this would scale... [18:46:54] what you are doing is basically what happens for document search [18:46:56] you have a search term [18:47:10] like "weather new york" [18:47:21] and you look for the best hit in your documents [18:47:33] which one of your document has most of the words in your search inside [18:47:44] it doesn't really matter that these are just words [18:47:51] or property-value pairs [18:48:06] i.e. "topic:weather location:new+york" is the same thing [18:48:13] and that's where lucene excels [18:48:23] (which powers both solr and elasticsearch) [18:48:29] It doesn't matter even if they are integers and are indexed right? [18:48:47] nope, if you try to make indexes on 700 columns [18:48:51] over 12 million rows [18:48:54] you will be unhappy [18:49:04] oh. yeah. 700 columns index is unheard of. [18:50:05] The problem is I can't seem to find how to replicate that kind of query in mongodb or elasticsearch, except writing a mapreduce job to do the full table scan on mongodb. [18:50:35] Denny_WMDE1: Any ideas on this? Perhaps I should shoot a question on stackoverflow. [18:51:18] i would say, don't do that query [18:51:25] but rely on lucene doing it for you [18:51:48] No, I meant how to have that kind of ordering. [18:52:08] lucene does ranking fitting to your query [18:52:34] which means the more columns match the higher the result is ranked? [18:52:36] using some smart information retrieval magic [18:52:40] aye [18:52:46] and you can boost certain columns [18:52:51] normalize them [18:52:54] etc., the whole thing [18:53:10] whoa. I haven't explored lucene yet. This sounds like godsend. [18:54:00] I've heard somewhere that I can add my own similarity functions. It would be great if I could do that, since I need to assign weights to certain columns. [18:54:05] well, you're doing search :) [18:54:13] you should check out that lucene thing ;) [18:54:53] if it just weights, you don't need a full blown sim-function [18:54:54] Haha. Yeah. Thanks Denny. This was helpful. :) [18:55:12] you can do it just by boosting some columns, i.e. giving them weights [18:55:21] yw :) [18:55:37] I'm going to smile all night after hearing this boosting thing. :P [18:55:57] (PS1) Jeroen De Dauw: Added test for Entity and fixes inconsistency in id field handling [extensions/WikibaseDataModel] - https://gerrit.wikimedia.org/r/75149 [18:56:20] (PS2) Jeroen De Dauw: Fixed incorrect doc [extensions/WikibaseDatabase] - https://gerrit.wikimedia.org/r/75124 [18:56:54] I was putting Lucene away thinking that it's mainly for text-based search, full-text search etc. Once back I did those with Sphinx. Now you got me going for it. [18:57:01] (PS1) Jeroen De Dauw: DO NOT MERGE [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75150 [18:57:44] ^demon: I did not mean to imply getting tid of these global scope asusmptions is easy ;p [18:57:55] I'm quite well aware it is very hard [18:58:21] (CR) jenkins-bot: [V: -1] DO NOT MERGE [extensions/WikibaseQuery] - https://gerrit.wikimedia.org/r/75150 (owner: Jeroen De Dauw) [18:58:22] Or rather, a lot of work :) [18:58:37] <^demon> I know you know, but I thought someone else might read it as "Oh well let's just do that and we're done!" :) [20:18:34] Day 9348: Still trying to figure out why unit tests seem to randomly fail with MySQL. We should fix our group attribution badly. My water reserves will be exhausted soon. [20:18:35] :D [20:38:09] [travis-ci] wikimedia/mediawiki-extensions-Wikibase#454 (master - a29cd53 : Translation updater bot): The build was broken. [20:38:09] [travis-ci] Change view : https://github.com/wikimedia/mediawiki-extensions-Wikibase/compare/685ed24a8069...a29cd53be321 [20:38:09] [travis-ci] Build details : http://travis-ci.org/wikimedia/mediawiki-extensions-Wikibase/builds/9367374 [20:42:16] hoo|away: do they ever fail for you locally? [20:42:20] or only in ci? [20:53:49] aude: can haz osm question? [20:58:07] Denny_WMDE1: both, but they only fail on travis sometimes [20:58:52] hoo|away: if you can repeat it locally, did you already try the divide and conquer method? [21:00:04] Denny_WMDE1: huh? I guess not [21:00:32] i mean, simply removing tests [21:00:39] Sure [21:00:43] until you have a set where it consistantly fails [21:00:44] Several times [21:01:09] and do you get such a test? [21:01:12] set? [21:01:24] not yet [21:01:26] Denny_WMDE1: ? [21:02:05] aude: i try to figure out the OSM relation for Brac [21:02:10] this is the node http://www.openstreetmap.org/?nodeid=52028676 [21:02:15] but how do i get the relation? [21:02:28] hoo|away: because of the randomness? [21:03:48] Denny_WMDE1: No, it fails iwth a message mentioning the string "spam" and I removed all tests with the word "spam" in the source [21:03:51] but still it failed [21:04:19] well, remove more [21:04:28] remove half of them [21:04:33] see if it still fails [21:04:36] remove the other half [21:04:37] etc. [21:04:48] keep doing that [21:04:54] sadly my system's very slow [21:04:59] :( sorry [21:05:09] Denny_WMDE1: hmmm.... the coastline is defined [21:05:23] i don't see a relation defined there [21:06:06] some will exist like http://www.openstreetmap.org/browse/relation/2146410 [21:07:31] and can I figure it out somehow? [21:11:09] Denny_WMDE1: there is http://ra.osmsurround.org/ [21:11:13] not sure how well it works [21:11:34] or i just looked at the coastline [21:13:16] :( I do not understand this search [21:13:23] Brač is the place I am looking for [21:19:14] http://nominatim.openstreetmap.org/search.php?q=Bra%C4%8D is what the nominatim search finds [21:19:28] it's a point :/ [21:19:45] if there was a relation, surely it would be found there [21:21:33] :( [21:21:37] so nothing to link to [21:21:53] :´( [21:22:13] * Sky2042_afk links Denny's head. [21:22:37] not that i see [21:23:16] * hoo goes to assassinate our unit tests [21:23:20] the island is basically a hole in the sea, though someone can add an additional relation for the island [21:23:34] If I run phpunit.php --group WikibaseClient,WikibaseRepo,WikibaseLib I get even more failures [21:23:44] as there still are tests not in the Wikibase group -.- [21:28:11] hoo: :( [21:33:02] Denny_WMDE1: P131 correctly has the English connotation. [21:34:06] Sky2042_afk: you mean to connect an island to the surrounding sea? [21:34:25] no no, that your interpretation of it being the incorrect property is the correct interpretation. :) [21:34:46] ah, ok - so is there something like "located in" ? [21:35:19] also most of the other labels of P131 do not carry that administrative restriction [21:35:22] I'm looking. [21:35:38] i am not sure why the restriction is useful, but that's another topic [21:37:17] It's strictly for use for administrative units... [21:37:19] * Sky2042_afk shrugs. [21:37:22] Er [21:37:26] That probably is wrong to say. [21:39:12] There isn't a property for saying an island is in x body of water. [21:39:23] The closest you'll get is coordinate location + instance of island. [21:39:33] But that of course is only indirect. [21:39:41] Odd [21:41:14] Denny_WMDE1: Check out http://www.wikidata.org/wiki/Wikidata:Property_proposal/Place#Borders_ocean [21:42:28] hmmm… somehow I have the feeling turning P131 into a general "located in" is more appropriate than "borders ocean" [21:42:41] You would be wrong. [21:42:49] oh, care to explain? [21:42:54] I know we've had the discussion on why 131 is the way it is. [21:44:09] Ah yes, in the property discussion. [21:44:12] :P [21:44:16] hmm. so what arguments in particular were convincing? :) [21:44:19] http://www.wikidata.org/wiki/Property_talk:P131 [21:45:21] "is in" is too general. [21:45:25] Which is what you're suggesting. [21:45:35] hmmm, so i can't use it to say x is in y census area? [21:45:44] or is census area administrative? [21:45:57] why is it too general? [21:45:59] aude: You could probably make the claim that a census area is administrative, though I'm not sure I would. [21:46:11] Denny_WMDE1: reading the talk page, perhaps [21:46:30] Denny_WMDE1: Is anyone assigned to work on the unit tests? There's a LOT of work to do [21:46:52] The more I look at them, the more I think is needed :P [21:47:05] hoo: you mean in general? [21:47:07] hoo: everyone is assigned [21:47:24] hoo: or something specific? "improve unit test" doesn't make for a great item :) [21:47:37] we assigned improving the api tests [21:47:59] Denny_WMDE1: Well, there are several tasks: Regroup the tests, make the whole suit pass, API test refactor, ... [21:48:55] hmm, if P131 is not the right one, maybe there should be something to say "is located in" in a more generic way… [21:49:05] meh, i dislike the property proposal process [21:49:20] It's hardly onerous. [21:49:42] It's also largely in place to stop us from having the endless discussion like P107 has brewed. [21:50:39] hoo: yes, these are different things. the "make the whole suit pass" should be high. I'll comment on that tomorrow [21:51:16] Denny_WMDE1: You probably have to regroup them first only to see the failures :P [21:51:33] Sky2042_afk: meh, for now I just need something to say in which sea my island is :) [21:52:01] Denny_WMDE1: Well, someone owns the island, too. Go add that for P131. :P [21:52:24] s/someone/some country/ [21:52:33] sure, already did [21:52:43] but the island as a whole is actually not an administrative unit [21:52:57] * nikki has also wondered why there isn't something for saying something is located in something else [21:53:22] it is subdivided into six municipalities, which all belong to one county, which also has other islands (and some mainland area) [21:53:40] Then it seems island P131 country works. [21:53:52] * Sky2042_afk shrugs. [21:53:57] well, i agree [21:54:08] but it doesnt connect it the sea :) [21:54:16] Sure. [21:55:21] You could probably use http://www.wikidata.org/wiki/Property:P47 even though that's not quite the correct connotation. [21:55:39] yeah, was thinking about this too [21:55:56] Quote from the talk page, in fact: When there aren't land borders, like in Iceland, may I use no value with this property? - Sarilho1 (talk) 15:21, 24 March 2013 (UTC) [21:55:57] Yes, I think so. --Nightwish62 (talk) 19:00, 24 March 2013 (UTC) [21:55:59] Thanks! - Sarilho1 (talk) 19:04, 24 March 2013 (UTC) [21:56:23] hmm, that would mean not to use it for that [21:56:55] Well, it doesn't define that it's in the ocean, but it also does define that the island is an island. Which I suppose instance of does that. [21:56:57] Hmm. [21:57:41] yeah, instance of island is already stated [21:57:51] and as said, it is not an administrative unit as a whole [21:58:07] and P47 is supposed to be used only between administrative units of the same level [21:58:26] if i read this correctl [21:58:27] y [21:58:28] since there are people here, maybe you know the answer... is there a way to get a list of things which have a certain property set to a certain value? [21:58:29] Hi. Are category redirects supposed to have data items on wikidata? [21:59:05] Sid-G: I think we've been leaving those alone? I'm not sure. [21:59:27] nikki: not yet directly, we are working on this [21:59:43] nikki: you can download the knowledge base and do the query locally [21:59:45] nikki: Not currently. Queries as part of the UI are part of phase 3. You can go bug someone at [[WD:Database reports]] for a query on the non-UI side, however. [21:59:46] [2] 10https://www.wikidata.org/wiki/WD:Database_reports [22:00:03] Sky2042_afk: so I'm basically merging a bunch of categories over at hi-wp. Should I request deletion of the data items for the categories which are now category redirects? [22:00:17] ahh. good to know I'm not missing something obvious then :) [22:00:28] Sid-G: Do you have an example? [22:00:56] Sky2042_afk: Its a bunch of categories about Afganistan with minor spelling variations [22:00:59] Sid-G: Or rather, better question: Are those categories at hi-wp going to be deleted ever? [22:01:04] by download the knowledge base, do you mean the data dumps? I thought about that but couldn't find anything that said how to import it into anything [22:01:20] Sky2042_afk: uh, no. I don't think they ever will be deleted [22:01:46] maybe one of the tools works http://www.wikidata.org/wiki/Wikidata:Tools#External_tools ? [22:01:57] they tend to be proof of concept [22:02:16] Sid-G: You might post that question on [[WT:N]] or [[WD:PC]] then. I would probably delete them myself if they showed up at WD:RFD, but discussion might be better. [22:02:25] [3] 10https://www.wikidata.org/wiki/WT:N13 => [22:02:27] nikki: i have made here a nicer download: https://dl.dropboxusercontent.com/u/172199972/index.html [22:02:27] [4] 10https://www.wikidata.org/wiki/WD:PC [22:02:40] it's about 800 MB though [22:02:46] but the format should be readable [22:02:52] needs some description, though, i admit [22:03:00] Sky2042_afk: k, thnx :) [22:03:50] thanks, I'll have a look :)