[00:01:21] https://www.wikidata.org/wiki/Special:Contributions/Dexbot [00:10:36] something is wrong here: https://eo.wikibooks.org/wiki/Kategorio:%C4%88io [00:10:49] it is added here, but does not show iw's: https://www.wikidata.org/wiki/Q1281#sitelinks-wikibooks [00:13:24] same issue: https://fr.wikibooks.org/wiki/Cat%C3%A9gorie:Principale [00:14:42] ow, someone changed [00:15:59] crap [01:15:38] [13WikidataBrowserTests] 15JeroenDeDauw comment on pull request #51 1464c9938: why is there `enwiki` above and `en` below? 02http://git.io/AQVv [01:22:44] dennyvrandecic: and it's definitly Q1337 [05:54:15] 10Wikidata, 10MediaWiki-extensions-ContentTranslation, 5ContentTranslation-Release4, 3LE-Sprint-83: Link translated articles with other languages through Wikidata automatically - https://phabricator.wikimedia.org/T87410#1064879 (10Arrbee) [08:28:52] (03CR) 10Adrian Lang: Check result of getEntityRevision (031 comment) [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192519 (https://phabricator.wikimedia.org/T90268) (owner: 10Adrian Lang) [08:42:08] good morning [08:42:19] sjoerddebruin: heb je even ? [08:47:56] GerardM-: altijd. [09:20:44] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 7I18n, 5Patch-For-Review: The messages Wikibase-validator-sitelink-conflict and Wikibase-undo-title show HTML when used as an error after trying to reset an item page - https://phabricator.wikimedia.org/T85340#1065013 (10Aklapper) [09:32:11] 10Wikidata, 10Wikidata-Quality: evaluate mapping when writing dump information to database - https://phabricator.wikimedia.org/T90688#1065032 (10dominic.sauer) 3NEW [09:32:31] DanielK_WMDE: aude: https://www.wikidata.org/wiki/Special:ItemDisambiguation?language=en&label=Chinatown do you really think the plain item id for the last item is useful? [09:32:42] I'd rather write something like 'no description given yet' [09:32:48] 10Wikidata, 10Wikidata-Quality: evaluate mapping when writing dump information to database - https://phabricator.wikimedia.org/T90688#1065047 (10dominic.sauer) [09:33:08] Adrian_WMDE: since the item is already linked, i say not [09:40:09] Adrian_WMDE: http://tools.wmflabs.org/wikidata-terminator/?lang=en&term=Church+Of+St+Mary&doit=1 :) [10:00:05] [13WikidataBrowserTests] 15WMDE-Fisch comment on pull request #51 1464c9938: The first one calls the API, where the whole key is checked. The second one uses the frontend where the usage of the complete key is kind of undesired, and the "en" will be auto completed internally. ... 02http://git.io/A5CN [11:38:50] [13WikidataBrowserTests] 15zeljkofilipin comment on pull request #51 1464c9938: Is this clear/send_keys dance necessary? Just setting the element does not do the job? Something like:... 02http://git.io/A55I [12:07:38] hoo: if you have something for scrum of scrums tonight please let me and Tobi_WMDE_SW know [12:08:18] Lydia_WMDE: https://phabricator.wikimedia.org/T89169 [12:08:33] thx [12:32:20] wow https://en.wikipedia.org/wiki/Template:Link_FA :) [12:33:51] hoo: I would say Wikidata has just "unlocked" another achievement :) [12:37:27] Yeah :) [12:39:46] FYI Magnus updated the Wikidata stats by hand [12:39:48] \o/ well done hoo [12:40:50] GerardM-: sadly he didn't make the page wider. [12:42:57] I am happy it fits on my screen [12:43:52] Well, the bottom of the chart doesn't look good.. [13:02:10] Hoi ... one way of looking at it is that it is the best we have and it is free [13:02:19] if you want to improve it, it is open source [13:02:30] it is all the statistics we have [13:36:15] Wow, what a nice e-mail on Wikidata-i [13:37:39] manybubbles: please add me as admin to https://wikitech.wikimedia.org/wiki/Nova_Resource:Wikidata-query when you have time [13:39:04] jzerebecki: done [13:41:24] thx [13:53:39] (03CR) 10Hoo man: [C: 04-1] "documentation nit pick" (031 comment) [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192776 (owner: 10Adrian Lang) [13:55:42] (03CR) 10Hoo man: [C: 032] Fix broken undo error message [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192364 (https://phabricator.wikimedia.org/T85340) (owner: 10Thiemo Mättig (WMDE)) [13:58:29] (03Merged) 10jenkins-bot: Fix broken undo error message [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192364 (https://phabricator.wikimedia.org/T85340) (owner: 10Thiemo Mättig (WMDE)) [13:58:51] (03CR) 10Hoo man: "Per master, forgot to +2 this yesterday" [extensions/Wikibase] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192598 (https://phabricator.wikimedia.org/T85340) (owner: 10Hoo man) [13:59:53] (03CR) 10Hoo man: [C: 032] "Per master, forgot to +2 this yesterday" [extensions/Wikibase] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192600 (owner: 10Hoo man) [14:02:24] (03Merged) 10jenkins-bot: Add our additional user rights to $wgAvailableRights [extensions/Wikibase] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192600 (owner: 10Hoo man) [14:10:09] (03CR) 10Hoo man: [C: 032] Don't implicitly parse messages passed to OutputPage::addWikiText [extensions/Wikibase] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192598 (https://phabricator.wikimedia.org/T85340) (owner: 10Hoo man) [14:10:40] (03PS4) 10Adrian Lang: Pass LanguageNameLookup to ItemDisambiguation [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192776 [14:10:49] (03CR) 10Adrian Lang: Pass LanguageNameLookup to ItemDisambiguation (031 comment) [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192776 (owner: 10Adrian Lang) [14:13:31] (03Merged) 10jenkins-bot: Don't implicitly parse messages passed to OutputPage::addWikiText [extensions/Wikibase] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192598 (https://phabricator.wikimedia.org/T85340) (owner: 10Hoo man) [14:14:55] (03CR) 10Hoo man: [C: 032] Pass LanguageNameLookup to ItemDisambiguation [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192776 (owner: 10Adrian Lang) [14:16:48] (03PS1) 10Hoo man: Update Wikibase [extensions/Wikidata] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192800 [14:17:19] (03Merged) 10jenkins-bot: Pass LanguageNameLookup to ItemDisambiguation [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192776 (owner: 10Adrian Lang) [14:17:21] (03CR) 10Hoo man: [C: 032] Update Wikibase [extensions/Wikidata] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192800 (owner: 10Hoo man) [14:18:14] 10Wikidata, 10MediaWiki-extensions-ContentTranslation, 5ContentTranslation-Release4, 7Design, 3LE-Sprint-83: When creating a new page, suggest doing a translation instead - https://phabricator.wikimedia.org/T88600#1065657 (10Amire80) [14:19:09] (03Merged) 10jenkins-bot: Update Wikibase [extensions/Wikidata] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192800 (owner: 10Hoo man) [14:20:14] 10Wikidata, 10§ Wikidata-Sprint-2015-02-25, 7Performance: Migrate all relevant Lua modules from mw.wikibase.getEntity to mw.wikibase.getEntityObject - https://phabricator.wikimedia.org/T88950#1065661 (10Lucie) [14:34:25] !merge 192362 [14:34:25] merge merge merge MERGEEEEEEEEEEEEE https://gerrit.wikimedia.org/r/#/c/192362/ [14:34:25] Lydia_WMDE: DanielK_WMDE what about https://phabricator.wikimedia.org/T88441? can we close this? investigation has happened AFAICS [14:34:43] hah, the bot is still here :D [14:36:32] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-03, 10§ Wikidata-Sprint-2015-02-25: investigate how language fallbacks can be done in entity search - https://phabricator.wikimedia.org/T88441#1065682 (10Tobi_WMDE_SW) [14:37:25] 10Wikidata, 10§ Wikidata-Sprint-2015-02-03, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review: spec of how timevalue works and is supposed to work - https://phabricator.wikimedia.org/T88438#1065685 (10Tobi_WMDE_SW) [14:37:47] 10Wikidata, 10§ Wikidata-Sprint-2015-02-03, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review: figure out current state of time storage and display - https://phabricator.wikimedia.org/T88437#1065687 (10Tobi_WMDE_SW) [14:38:52] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10Wikidata-Sprint-2015-01-08§, 10§ Wikidata-Sprint-2015-01-21, and 4 others: batch entity revision lookup for performance improvements - https://phabricator.wikimedia.org/T87238#1065690 (10Tobi_WMDE_SW) [14:39:49] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-03, 10§ Wikidata-Sprint-2015-02-25: Allow simultaneous editing of mainsnak and references - https://phabricator.wikimedia.org/T87759#1065693 (10Tobi_WMDE_SW) [14:40:06] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-03, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review: Allow simultaneous editing of mainsnak and references - https://phabricator.wikimedia.org/T87759#998162 (10Tobi_WMDE_SW) [14:40:35] 10Wikidata, 10§ Wikidata-Sprint-2015-02-03, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review: Remove time conversion logic from JavaScript code - https://phabricator.wikimedia.org/T89243#1065698 (10Tobi_WMDE_SW) [14:41:11] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-03, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review: Special:GoToLinkedPage should support redirects - https://phabricator.wikimedia.org/T78526#1065704 (10Tobi_WMDE_SW) [14:41:35] !evil hoo [14:41:35] hoo is eeeeeevil >:( [14:42:06] !nyan sjoerddebruin [14:42:06] ~=[,,_,,]:3 [14:42:17] !nlwiki [14:42:22] Oh, we don't have that one. :P [14:42:29] 10Wikibase-JavaScript-Api, 10Wikidata, 10Wikidata-Sprint-2015-01-08§, 10§ Wikidata-Sprint-2015-01-21, and 3 others: Fix Job for running WBJSAPI QUnit tests on Jenkins - https://phabricator.wikimedia.org/T86176#1065713 (10Tobi_WMDE_SW) [14:46:08] 10Wikibase-DataModel, 10Wikidata, 10§ Wikidata-Sprint-2015-01-21, 10§ Wikidata-Sprint-2015-02-03: Remove Claim class - https://phabricator.wikimedia.org/T87388#1065745 (10Tobi_WMDE_SW) All PRs mentioned above are merged. Is there more to this or can we close this? @thiemowmde @jeroendedauw [14:47:17] 10Wikibase-DataModel, 10Wikidata, 10§ Wikidata-Sprint-2015-01-21, 10§ Wikidata-Sprint-2015-02-03, and 2 others: Remove Claim class - https://phabricator.wikimedia.org/T87388#1065747 (10Tobi_WMDE_SW) [14:48:02] 10Wikidata, 10Wikidata-Sprint-2014-12-09§, 10Wikidata-Sprint-2015-01-08§, 10§ Wikidata-Sprint-2015-01-21, and 3 others: File based caching: Split of site lookup for readonly access - https://phabricator.wikimedia.org/T77990#1065753 (10Tobi_WMDE_SW) [14:50:40] Lydia_WMDE: nobody has worked on this since over a month. what's the status of this and when did we commit on doing it? also: this was in the review column in the last sprint which is probably not a good idea since it need actual implementation to get this moving [14:54:06] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-01-21, 10§ Wikidata-Sprint-2015-02-03, and 2 others: Split entityViewInit into WikibaseRepo and WikibaseView part - https://phabricator.wikimedia.org/T87434#1065764 (10Tobi_WMDE_SW) [14:55:32] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review: Introduce wikibase.view.ViewFactory - https://phabricator.wikimedia.org/T90720#1065767 (10Tobi_WMDE_SW) 3NEW a:3adrianheine [14:56:42] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review: Introduce wikibase.view.ViewFactory - https://phabricator.wikimedia.org/T90720#1065767 (10Tobi_WMDE_SW) https://gerrit.wikimedia.org/r/#/c/188357/ [14:57:42] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-25: Browsertests for time value input - https://phabricator.wikimedia.org/T88542#1065789 (10Tobi_WMDE_SW) a:3WMDE-Fisch [14:58:40] 10Wikidata, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review: Edit button in heading overlaps alias text - https://phabricator.wikimedia.org/T89956#1065790 (10Tobi_WMDE_SW) a:3aude [14:58:49] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review: Item layout is a bit inconsistent when there are no statements and "other languages" is collapsed - https://phabricator.wikimedia.org/T90395#1065791 (10Tobi_WMDE_SW) a:3aude [14:58:57] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review: h1 underline does not extend full width of wikibase-entityview-main - https://phabricator.wikimedia.org/T90426#1065792 (10Tobi_WMDE_SW) a:3aude [14:59:11] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review, 7Performance: Frontend JS should not re-render qualifiers and references upon load. - https://phabricator.wikimedia.org/T88425#1065793 (10Tobi_WMDE_SW) a:3Snaterlicious [15:00:35] Lydia_WMDE: the task I was talking about is https://phabricator.wikimedia.org/T35704 [15:04:46] aude: DanielK_WMDE: Lydia_WMDE: apparently labels DO work on Special:Contributions. but only if I'm logged in. you were probably logged out when we were lokking at it during the meeting.. [15:04:58] Tobi_WMDE_SW: interesting [15:05:21] suppose a result of how caching works (e.g. user-based stuff only for logged in) [15:06:48] aude: I also see dutch labels if I switch to an exotic language [15:07:03] so it seems to me fallback still works as it did before.. [15:07:45] ok [15:10:16] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 7I18n, 5Patch-For-Review: The messages Wikibase-validator-sitelink-conflict and Wikibase-undo-title show HTML when used as an error after trying to reset an item page - https://phabricator.wikimedia.org/T85340#1065822 (10Lydia_Pintscher) 5Open>3Reso... [15:12:38] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata: Label no longer fallback to foreign language out of fallback chain in Special:Contributions and Special:RecentChanges - https://phabricator.wikimedia.org/T89212#1065845 (10Tobi_WMDE_SW) When I'm logged in and look at https://www.wikidata.org/wiki/Special:C... [15:13:11] Tobi_WMDE_SW: https://phabricator.wikimedia.org/T88441 <- yes i think we can close this [15:14:01] Tobi_WMDE_SW: https://phabricator.wikimedia.org/T35704 <- katie wanted to look into this further and give actionable feedback [15:14:02] Lydia_WMDE: you close it, i move it to done? [15:14:03] ;) [15:14:10] k [15:14:29] Tobi_WMDE_SW: special:contributions: i am pretty sure i was logged in -.- [15:14:36] Lydia_WMDE: so we put the geo task in our current sprint? aude? [15:14:52] Lydia_WMDE: see my attached screenshot, for me it works [15:14:59] in the task [15:15:05] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata: add language fallback to entity search/suggester - https://phabricator.wikimedia.org/T76150#1065847 (10Lydia_Pintscher) [15:15:06] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-03, 10§ Wikidata-Sprint-2015-02-25: investigate how language fallbacks can be done in entity search - https://phabricator.wikimedia.org/T88441#1065846 (10Lydia_Pintscher) 5Open>3Resolved [15:15:09] i'll have a look yeah [15:15:10] strange [15:16:08] Tobi_WMDE_SW: i would like to give it more a way forward in the review for geodata [15:16:24] aude: we add it to the sprint then? [15:16:32] had looked at it on the airplane but couldn't comment on gerrit then [15:16:33] Tobi_WMDE_SW: ok [15:16:51] not dure if somebody will work on it if we are not [15:17:29] aude: so we need to actually have implementation-time for it in this sprint [15:17:42] [13WikidataBrowserTests] 15WMDE-Fisch comment on pull request #51 1464c9938: The WebDriver::Element we use here does not support the set method. Furthermore I think that send_keys is needed in some cases for the frontend JS to detect changes in the input field. 02http://git.io/AFmL [15:18:00] Tobi_WMDE_SW: that's not what i want in this sprint necessarily [15:18:38] just some specific super easy actions at this point to get it mergeable [15:19:14] then it might be a task for us, if nobody else does, to add some sort of smoke test maybe so we know when it breaks [15:19:22] [13WikidataBrowserTests] 15tobijat comment on pull request #51 1464c9938: @zeljkofilipin yes, main reason for this is that the native WD element does not have another way for this. at least I did not find one. 02http://git.io/AFYv [15:19:31] aude: ok [15:19:33] will add [15:19:41] which is a certainty, like our stuff with spam blacklist etc [15:19:47] k [15:19:58] 10MediaWiki-extensions-WikibaseClient, 10Wikidata, 10MediaWiki-extensions-GeoData, 10Wikipedia-App-Android-App, and 3 others: GeoData should pull coordinates from Wikidata on sites with the Wikibase Client - https://phabricator.wikimedia.org/T35704#1065864 (10Tobi_WMDE_SW) [15:20:44] 10Wikidata: http://test.wikidata.org/entity/Q837 is missing apache config/redirect - https://phabricator.wikimedia.org/T89964#1065865 (10Lydia_Pintscher) p:5High>3Low [15:21:21] 10Wikidata: http://test.wikidata.org/entity/Q837 is missing apache config/redirect - https://phabricator.wikimedia.org/T89964#1065867 (10Lydia_Pintscher) 5Open>3declined a:3Lydia_Pintscher As discussed in today's sprint start probably not worth doing. [15:25:00] 10Wikidata, 10MediaWiki-extensions-ContentTranslation, 5ContentTranslation-Release4, 3LE-Sprint-83: Link translated articles with other languages through Wikidata automatically - https://phabricator.wikimedia.org/T87410#1065886 (10Arrbee) p:5Normal>3Low [15:28:03] (03PS3) 10Aude: More consistent layout w/ collapsed "other languages" and no statements [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192545 (https://phabricator.wikimedia.org/T90395) [15:37:48] (03PS4) 10Aude: More consistent layout w/ collapsed "other languages" and no statements [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192545 (https://phabricator.wikimedia.org/T90395) [15:38:10] 10Wikidata, 10§ Wikidata-Sprint-2015-02-03, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review: figure out current state of time storage and display - https://phabricator.wikimedia.org/T88437#1065910 (10Tobi_WMDE_SW) There are quite some (under-statement!) patches up for review that fix different issues wit... [15:42:20] (03PS5) 10Aude: More consistent layout w/ collapsed "other languages" and no statements [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192545 (https://phabricator.wikimedia.org/T90395) [15:54:18] 10Wikidata, 10MediaWiki-extensions-ContentTranslation, 5ContentTranslation-Release4, 3LE-Sprint-83: Link translated articles with other languages through Wikidata automatically - https://phabricator.wikimedia.org/T87410#1066013 (10Amire80) a:3Amire80 [16:08:13] [13WikidataBrowserTests] 15WMDE-Fisch created 06sitelink_fixmes from 06sitelinks_T52362 (+0 new commits): 02http://git.io/AFia [16:10:55] [13WikidataBrowserTests] 15WMDE-Fisch 04deleted 06sitelink_fixmes at 1464c9938: 02http://git.io/AFXG [16:11:04] [13WikidataBrowserTests] 15WMDE-Fisch created 06sitelink_fixmes (+1 new commit): 02http://git.io/AFXl [16:11:04] 13WikidataBrowserTests/06sitelink_fixmes 141aff795 15WMDE-Fisch: get rid of fixmes and refactored sitelink creation [16:12:37] [13WikidataBrowserTests] 15WMDE-Fisch opened pull request #52: fixed fixmes and slightly refactored sitelink creation (06master...06sitelink_fixmes) 02http://git.io/AF1U [16:20:02] 10Wikidata, 10§ Wikidata-Sprint-2015-02-25, 5Patch-For-Review: update and consolidate json dump documentation - https://phabricator.wikimedia.org/T87329#1066051 (10Tobi_WMDE_SW) [16:32:44] (03CR) 10Hoo man: Consolidate duplicate PropertyParserFunction/Lua code (034 comments) [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189415 (https://phabricator.wikimedia.org/T48160) (owner: 10Hoo man) [16:37:10] (03PS6) 10Hoo man: Consolidate duplicate PropertyParserFunction/Lua code [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189415 (https://phabricator.wikimedia.org/T48160) [16:40:42] (03CR) 10Hoo man: "Renamed EntityStatementsRenderer to StatementTransclusionInteractor and resolved several other things pointed out by Daniel." [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189415 (https://phabricator.wikimedia.org/T48160) (owner: 10Hoo man) [16:41:20] DanielK_WMDE: https://gerrit.wikimedia.org/r/189415 Resolved your comments and added integration tests for the new behavior [16:41:23] Merge? :P [16:49:25] aude: perhaps also have a look at this ---^ [16:49:56] (03CR) 10Hoo man: [C: 032] Remove redundant build settings [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192603 (https://phabricator.wikimedia.org/T85799) (owner: 10JanZerebecki) [16:50:28] aaaah have to rebase my patch on that [16:51:16] aude: Sorry :S [16:51:25] it's ok :) [16:52:37] (03Merged) 10jenkins-bot: Remove redundant build settings [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192603 (https://phabricator.wikimedia.org/T85799) (owner: 10JanZerebecki) [16:53:33] aude, hoo: I think PropertyClaimsRenderer could perhaps be folded into StatementTransclusionInteractor. That thing has to support language fallback anyway, right? [16:53:44] DanielK_WMDE: my patch is to get rid of it [16:53:48] bad name anyway [16:53:56] yea... [16:54:14] Yeah, aude's follow up does that [16:54:20] but then you are both working on the same code. you should coordinate on agree... [16:54:33] sure [16:54:42] aude: if you work on getting rid of that already, great - that makes me more confident in approving hoo's patch. [16:54:51] excellent [16:54:54] hoo's patch looks good at a glance [16:55:16] +1 then? [16:55:31] if i manual check it then, +2? [16:55:42] it's been stuck for a while.... [16:56:09] Would be nice to have that merged. I hope I covered all edge cases in the tests at least for Lua [16:56:11] aude: sounds good - I give CR+1 and go into the RDF call, and you verify and +2 :) [16:56:16] ok [16:56:21] or are you ging to be in the call too? [16:56:29] nah.... already too many people [16:56:40] heh [16:56:46] just be sane with geo stuff :) [16:56:55] (03CR) 10Daniel Kinzler: [C: 031] "Looks good, didn't verify." (031 comment) [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189415 (https://phabricator.wikimedia.org/T48160) (owner: 10Hoo man) [16:56:57] using what markus has done, simple is ok for now [16:57:39] aude: well, markus only supplies the "bloated" version, we need the "lean" version of the graph [16:57:46] hm [16:58:40] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 7Performance: Implement deferred deserialization of Entity - https://phabricator.wikimedia.org/T90707#1066139 (10daniel) [16:59:03] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 7Performance: Benchmark deferred entity deserialization - https://phabricator.wikimedia.org/T90706#1066140 (10daniel) [16:59:43] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 7Performance: Benchmark deferred entity deserialization - https://phabricator.wikimedia.org/T90706#1065473 (10daniel) @aklapper yea, happens when i file tickets during a meeting :) [16:59:59] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata: RdfBuilder error on trying to getLanguageCode() - https://phabricator.wikimedia.org/T90484#1066144 (10despens) After some checking, indeed that must be the case. However, I don't know what configuration mistake I made. Is there a specification of what exa... [17:02:25] [13Time] 15Jc3s5h comment on commit 142f758f5: The further in the past an event occurred, the less likely it is the time of day will be known. It is unlikely the time of day will be known for events that occurred in AD 59 so there should be no requirement that the time of day be stated.... 02http://git.io/Abmt [17:02:52] hmmm actually it seems to me that the parser fucntion has an issue on master [17:04:29] aude: On master? What issue? [17:04:49] stuff is not rendered, but is in your patch [17:04:57] [13Time] 15Jc3s5h comment on commit 141fb5929: The comments refer to ISO but it is now agreed that the time string is not ISO, it merely resembles ISO 8601. 02http://git.io/AbYu [17:05:01] * aude thinks your patch does the right thing [17:05:26] 10Wikidata: Deploy changes_subscription table - https://phabricator.wikimedia.org/T90733#1066147 (10Tobi_WMDE_SW) 3NEW [17:05:42] i patched wikibase for https://phabricator.wikimedia.org/T90492 but need some guidance to make it less awful [17:05:45] i will try to figure out when/why it went wrong, but then think your patch is ok [17:06:20] and make sure the issue is not in anything we plan to deploy [17:07:16] actually i'm not sure if enough information about foreign file repos is available to me to make it totally flexible [17:07:32] for example, language name-spaces [17:07:36] aude: Ok [17:07:59] however, atm it works for me [17:08:17] despens: So you have a patch? [17:08:58] aude: oO I can reproduce... on master the parser function is broken for me as well [17:09:05] I only tested my patch [17:09:12] hoo: :( [17:09:13] hoo: theoretically, just don't know how to submit to gerrit [17:09:58] hoo: since i'm not familiar with the mediawiki software the code is probably awful [17:10:23] despens: It might be... but that's why we have code review :) [17:10:24] https://www.mediawiki.org/wiki/Gerrit/Tutorial [17:10:29] thnx [17:10:34] That should cover everything [17:11:34] gerrit is slow here... [17:13:39] [13Time] 15brightbyte comment on commit 142f758f5: @Jc3s5h As far as I understant, this restriction stems from the ambiguity of the input string. All the parser has is a bunch of numbers, and it needs to somehow guess what each number means.... 02http://git.io/AbGP [17:14:08] (03PS1) 10Tobias Gritschacher: disable save button until statement is changed [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192820 [17:14:22] (03CR) 10Tobias Gritschacher: [C: 032 V: 032] New Wikidata Build - 2015/02/24 [extensions/Wikidata] - 10https://gerrit.wikimedia.org/r/192546 (owner: 10Tobias Gritschacher) [17:19:22] (03CR) 10Hoo man: [C: 032] disable save button until statement is changed [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192820 (owner: 10Tobias Gritschacher) [17:19:39] 10Wikidata, 10Continuous-Integration, 10§ Wikidata-Sprint-2015-02-03, 10§ Wikidata-Sprint-2015-02-25, and 2 others: mw-debug.log missing in Jenkins jobs (Failed to be created "Permission denied") - https://phabricator.wikimedia.org/T85799#1066166 (10JanZerebecki) 5Open>3Resolved [17:22:33] (03Merged) 10jenkins-bot: disable save button until statement is changed [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192820 (owner: 10Tobias Gritschacher) [17:25:10] hoo: on wmf19, purging one of my pages with bazillion lua and parser function calls seems extra slow [17:25:47] could be just me [17:25:58] now is ok [17:26:45] aude: Compared to what? master or older Wikibase? [17:27:09] both [17:27:22] but think it's ok now, just refreshing apc, localisation etc when i change [17:27:37] and don't see the bug in the branch [17:27:49] Yeah, bug is not on the branch [17:27:57] at a glance the branch doesn't seem slower than master [17:28:42] Do we have integration tests for the parser function [17:28:43] s [17:28:50] i hope so [17:28:59] I mean such that do actual $wgParser->parse( '{{...}}' ) [17:29:00] Ok [17:29:11] don't know [17:30:07] Seems like we only have unit tests, ok [17:32:49] i am bisecting [17:33:02] incidentally c995b5b76534078ed30f6a43a8e17fd6694c9f0d revision produces a fatal error :( [17:33:36] oh [17:33:40] that is on master, yes [17:34:09] Argument 2 passed to Wikibase\DataAccess\SnaksFinder::getMainSnaksByRanks() must be of the type array, string given [17:34:19] which must have been fixed, since i don't get that on master [17:35:50] aude: Oh, right [17:35:56] we invoke SnaksFinder wrong on master [17:36:04] LanguageAwareRenderer L185 [17:36:40] third parameter is acceptable ranks, not language code [17:37:00] yep [17:37:28] an integration test (like we have for Lua) would have caught this) [17:37:40] :( [17:38:03] alright, merging your patch fixes it [17:46:07] (03CR) 10Aude: [C: 032] "looks good now (and fixes a bug in the parser function)" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189415 (https://phabricator.wikimedia.org/T48160) (owner: 10Hoo man) [17:46:31] \o/ [17:46:44] * aude going home... [17:46:52] Good Lydia imitation, hoo. [17:46:54] shall rebase my patch later or tomorrow [17:46:59] :) [17:47:06] !lydia is \o/ [17:47:07] Key was added [17:47:22] !lydia :) [17:47:22] \o/ [17:47:35] :D ok :) [17:49:11] (03Merged) 10jenkins-bot: Consolidate duplicate PropertyParserFunction/Lua code [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189415 (https://phabricator.wikimedia.org/T48160) (owner: 10Hoo man) [17:49:23] 10MediaWiki-extensions-WikibaseClient, 10Wikidata, 7Tracking: Reimplement Wikibase data structures in Lua - https://phabricator.wikimedia.org/T57920#1066368 (10hoo) [17:49:25] 10MediaWiki-extensions-WikibaseClient, 10Wikidata, 7Need-volunteer: Lua: Allow using labels in mw.wikibase.entity.formatPropertyValues - https://phabricator.wikimedia.org/T48160#1066366 (10hoo) 5Open>3Resolved [17:50:46] hoo closed this task as "Resolved" by committing rEWBAdd4e34d38fab: Consolidate duplicate PropertyParserFunction/Lua code.Via Daemons · Wed, Feb 25, 6:49 PM [17:50:47] wow [17:50:56] Since when does that work? [18:14:12] (03PS5) 10Hoo man: Allow getting the user's language and splitting the ParserCache from Lua [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189324 [18:16:05] m( [18:19:35] (03PS6) 10Hoo man: Allow getting the user's language and splitting the ParserCache from Lua [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189324 [18:20:47] (03CR) 10Hoo man: "Manually rebased PS4 (in PS5 I accidentally rebased PS3)" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189324 (owner: 10Hoo man) [18:26:27] (03CR) 10Daniel Kinzler: [C: 04-1] "I'm confused - how is the purpose and contract of EntityStatementsRenderer different from StatementTransclusionInteractor? Both take an e" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189652 (owner: 10Aude) [18:26:38] hoo: --^ [18:27:32] DanielK_WMDE_: That's based upon an old patch set of my change [18:27:43] before I renamed EntityStatementsRenderer [18:30:12] (03CR) 10Hoo man: "This is based upon an old version of my change... I guess it will make more sense after being rebased onto master." [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189652 (owner: 10Aude) [18:37:39] 10Wikidata: Deploy changes_subscription table - https://phabricator.wikimedia.org/T90733#1066581 (10daniel) [18:41:52] 10MediaWiki-extensions-WikibaseClient, 10Wikidata, 10§ Wikidata-Sprint-2015-02-25: Implement hybrid mode for change dispatcher, which uses both wb_items_per_site and wb_changes_subscription. - https://phabricator.wikimedia.org/T90755#1066614 (10daniel) 3NEW [18:46:18] 10Wikidata: Maintenance script to update a client wiki's subscriptions based on the usage tracking table. - https://phabricator.wikimedia.org/T90756#1066652 (10daniel) 3NEW [18:46:46] (03PS1) 10Daniel Kinzler: Fix documentation errir in populateChangesSubscription.php [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192844 [19:16:29] (03PS1) 10Hoo man: Introduce PropertyParserFunctionIntegrationTest [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192859 [19:25:06] (03PS2) 10Hoo man: Introduce PropertyParserFunctionIntegrationTest [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192859 [19:49:41] aude: I wrote some simple integration tests for the parser function so that this wont happen again: https://gerrit.wikimedia.org/r/192859 [19:49:51] thanks [19:54:49] (03PS2) 10Aude: Fix documentation error in populateChangesSubscription.php [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192844 (owner: 10Daniel Kinzler) [19:55:00] (03CR) 10Aude: [C: 032] Fix documentation error in populateChangesSubscription.php [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192844 (owner: 10Daniel Kinzler) [19:56:39] (03CR) 10Aude: "@daniel didn't someone ask for EntityStatementsRenderer to be renamed to StatementTransclusionInteractor (?) thingy?" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189652 (owner: 10Aude) [19:57:12] (03CR) 10Hoo man: "@Aude: Daniel himself asked for that" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/189652 (owner: 10Aude) [19:57:22] ah, yeah :) [19:57:44] not rebasing today, but probably will tomorrow [20:00:05] (03Merged) 10jenkins-bot: Fix documentation error in populateChangesSubscription.php [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192844 (owner: 10Daniel Kinzler) [20:41:31] about to submit my patch, though apparently it spans different git repos [20:41:58] how would i make them be seen as one? [20:44:12] i edited a javascript in extensions/ValueView/lib/jquery.ui and this seems to be a submodule? [20:49:53] despens: you can't make changes to several repos as one. you have to propose patches to them separately, cross-linking them [20:50:00] it's quite annoying, i know [20:50:23] despens: You probably edited this repo? https://github.com/wmde/ValueView [20:52:10] DanielK_WMDE_: how would i cross-link them? [20:52:22] (03PS1) 10Smalyshev: Convert values to named nodes [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192879 [20:54:07] benestar: indeed, what is the right repo on gerrit for that? [20:55:13] despens: we work on github only for this small extensions [20:55:15] afaik [20:57:11] benestar: what is this thing? https://gerrit.wikimedia.org/r/#/admin/projects/mediawiki/extensions/ValueView [20:57:33] ok, so that changed ^^ [20:57:46] was quite a while not active so I might have missed some things [20:58:00] but is it the actual wikibase value view? because it seems to be put under 'mediawiki' [20:59:31] it cannot be the right one, it was last updated in 2013 [21:05:59] despens: i think that was before the move to github... and we are thinking about moving it back to gerrit. oh, well. [21:06:23] despens: you cross-link them simply by posting a link as a comment (or better, the commit message) [21:06:50] if you go into the ValueView directory locally, git remote -vv should tell you where it came from [21:07:36] DanielK_WMDE_: yes, strange it only shows the wikibase repo on gerrit [21:07:46] DanielK_WMDE_: i will put a pull-request on github [21:16:41] [13ValueView] 15despens opened pull request #159: patch for https://phabricator.wikimedia.org/T90492 (06master...06despens-T90492) 02http://git.io/AAne [21:18:01] despens: thanks [21:21:00] (03PS1) 10Dragan Espenschied: https://phabricator.wikimedia.org/T90492 [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192885 [21:24:56] Lydia_WMDE: will work on fixing guided tours first the next days [21:25:10] benestar: awesome [21:25:13] hopefully I can adapt the changes in the api without changing to much of my old code [21:25:21] they broke some backwards compatability :/ [21:25:22] that'd be awesome [21:25:29] hi benestar :) [21:25:31] Welcome back! [21:25:33] hey hoo :) [21:25:35] ty [21:25:43] DanielK_WMDE_: ok, managed. i'm sure there are better ways to do it, but it would be a small step making it useful for rhizome [21:25:45] benestar: there are probably also a few issues on our side [21:26:05] changed loading and removed dependencies [21:26:12] Lydia_WMDE: well, it's all a bit hackish because they don't support all features I wanna have natively [21:26:19] that might be part of the issue as well if i understood lucie correnctly when she investigated [21:26:57] I will see, it's not that much code so I can also write it sanely new :) [21:27:28] \o/ [21:29:53] Lydia_WMDE: btw they also dont work that nicely and use synchronous ajax requests as well :( [21:30:06] is the guided tours extension still in development? [21:30:15] i don't think so right now [21:32:20] well cya [21:33:08] despens: left soem feedback, but one of our frontend devs would be much better able to advise [21:33:16] on the details [21:34:12] maybe there is some exception, but generally ValueView is supposed to not know about wikibase and be reusable in other things [21:34:23] but wikibase knows about value view [21:35:16] whoops [21:35:24] I just repeated what Katie is saying [21:35:32] to slow [21:35:35] ooo [21:37:02] despens: Btw, there already is a source option in _create... it just defaults to a function that works on commons [21:37:23] I guess you would just inject a different anonymous function from Wikidata to use another repo [21:37:27] and not the url [21:37:59] s/Wikidata/Wikibase/ [21:42:33] aude, hoo: makes sense, since they are separate repos! [21:47:04] aude, hoo: i will wait for further feedback. i think injecting a function from wb into the valueview widget is even more dangerous than what i made [21:48:18] despens: Why dangerous? [21:50:57] hoo: maybe i misunderstood what you meant, but wouldn't you want to replace _initDefaultSource from another javascript? [21:51:31] hoo: anyway, i don't know enough about the architecture [21:52:40] despens: It can be injected [21:52:43] look at _create [21:52:49] it's supposed to be injected in fact [21:52:53] we do taht in various places [21:54:25] thnx [21:56:29] hoo: honestly don't understand what _create is doing or how would i even call it [21:57:00] despens: You don't directly call it, it's part of the widget [21:57:01] after the widget has been created? when is that? how to get the widget's reference, etc [21:58:42] No, you pass that in when creating [21:59:36] do you know where it is created? [22:00:34] in the expert [22:01:01] ha i always wondered what this 'expert' thing is :) [22:01:26] http://api.jqueryui.com/jQuery.widget/ ;) [22:01:35] that also has documentations on the options [22:01:39] how to apss them [22:02:14] src/experts/CommonsMediaType.js that's the exper [22:04:36] ok, but: that is also part of valueview, so how is it supposed to know about the configured commons site? [22:05:19] i think somehow via the valueViewBuilders in wikibase repo [22:05:29] ./repo/resources/experts/getStore.js in wikibase [22:05:49] https://github.com/wikimedia/mediawiki-extensions-Wikibase/blob/master/repo/resources/wikibase.ValueViewBuilder.js [22:05:52] well, you probably also don't want the configuration there directly [22:05:52] or what hoo says [22:06:23] * aude not sure how it's all wired together without examining the code more closely [22:07:01] can try to consult Daniel Werner on the best way of doing this [22:07:26] despens: either adrian or henning when they are around [22:07:31] Daniel Werner is no longer part of the team [22:07:37] unlikely in the evenings [22:07:38] so you probably wont catch him [22:09:17] I guess I'd inject it in the entityviewinit via wikibase.experts.getStore -> the expert -> the widget [22:11:05] ok, so i would leave valueview alone completely [22:11:15] the more I think about it... the more I believe that this probably shouldn't be part of ValueView... like the wikibase-item/property experts [22:11:34] You would probably need to alter the expert to pass the parameter on [22:11:37] maybe [22:12:30] just for clearance: an expert is an edit widget for a data type? [22:12:56] It's more of a wrapper around that [22:14:33] Nice, we even have a high level documentation for thathttps://github.com/wmde/ValueView#jqueryvalueviewexpert [22:14:35] * https://github.com/wmde/ValueView#jqueryvalueviewexpert [22:26:56] help needed [22:27:08] hi Trijnstel :) [22:27:32] I got a database error when creating an item [22:27:34] hi hoo ! [22:27:39] we know [22:27:42] should be fixed now [22:27:46] try again [22:27:52] yeah, just try again [22:28:00] but thanks for poking us [22:28:00] oh [22:28:05] didn't know that :) [22:28:06] * Trijnstel checks [22:28:14] I thought I had done something wrong [22:28:17] feel free to poke any time [22:28:32] no, still there... [22:28:36] Exception encountered, of type "BadMethodCallException" [22:28:41] Trijnstel: Rule of thumb: No matter what you do, a database error is always our fault [22:28:42] (when clicking on the newly created item) [22:28:45] ouh [22:28:54] and the database error came when I created it [22:29:01] A database query error has occurred. This may indicate a bug in the software. [22:29:01] Function: WikiGrok\QuestionStore::store [22:29:01] Error: 1146 Table 'wikidatawiki.wikigrok_questions' doesn't exist (10.64.32.28) [22:29:16] https://www.wikidata.org/wiki/Special:Contributions/Trijnstel <- see the latest item [22:29:18] I can't open it [22:29:18] I see them in the logs [22:29:35] works for me now [22:29:45] not for me... [22:29:46] oh wait [22:29:52] i am on test.wikidata [22:29:53] * Trijnstel waits [22:29:57] ah.. [22:30:23] 2015-02-25 22:30:00 mw1034 wikidatawiki: [8a686c15] /wiki/Q19360262 BadMethodCallException from line 244 of /srv/mediawiki/php-1.25wmf18/extensions/Wikidata/extensions/Wikibase/repo/includes/View/EntityViewPlaceholderExpander.php: Call to a member function getEntity() on a non-object (NULL) [22:30:23] #0 /srv/mediawiki/php-1.25wmf18/extensions/Wikidata/extensions/Wikibase/repo/includes/View/EntityViewPlaceholderExpander.php(193): Wikibase\Repo\View\EntityViewPlaceholderExpander->renderTermBox() [22:30:34] what's that? [22:30:50] Probably the exception you encountered [22:30:51] we fixed that in master [22:30:59] Did we? [22:31:05] or maybe not merged yet [22:31:11] I thought I backported everything important [22:31:12] or it is merged [22:31:16] mh [22:31:17] but I created multiple items today... [22:31:19] and no problems [22:31:20] * aude merged it [22:31:22] so why now? [22:31:29] Trijnstel: Mediawiki was just updated [22:31:35] oh.. [22:31:40] probably because of the database error adn things catchingi up [22:33:44] hoo: https://gerrit.wikimedia.org/r/#/c/192519/ [22:34:02] yeah, found it [22:34:07] backport? [22:34:14] sure [22:34:35] doesn't merge cleanly [22:34:39] :( [22:34:53] will do locally [22:34:59] yeah [22:37:44] (03PS1) 10Hoo man: Check result of getEntityRevision [extensions/Wikibase] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192967 (https://phabricator.wikimedia.org/T90268) [22:38:34] aude: ^ +2? [22:38:52] sure [22:39:02] (03CR) 10Aude: [C: 032] Check result of getEntityRevision [extensions/Wikibase] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192967 (https://phabricator.wikimedia.org/T90268) (owner: 10Hoo man) [22:39:13] Trijnstel: Thanks for letting us know... it's appreciated [22:42:58] hoo: thank you :) [22:43:04] I hope I can edit the item soon :P [22:43:11] (I thought I made a mistake) [22:43:29] (03Merged) 10jenkins-bot: Check result of getEntityRevision [extensions/Wikibase] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192967 (https://phabricator.wikimedia.org/T90268) (owner: 10Hoo man) [22:43:51] Trijnstel: I guess ~20 further minutes max. [22:43:58] If everything works right [22:43:59] ok :) [22:44:22] Trijnstel: i think you should re-create the item [22:44:33] maybe [22:44:36] Can I haz link? [22:44:39] i can't open the item either [22:44:40] https://www.wikidata.org/wiki/Special:Contributions/Trijnstel [22:44:47] aude: will the other one deleted then? [22:44:56] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-03, 10§ Wikidata-Sprint-2015-02-25: Edit buttons are getting cached, although their visibility depends on user rights - https://phabricator.wikimedia.org/T85252#1067952 (10daniel) Investigating this took me deep into the won... [22:44:59] hoo: unless you backport to wmf16/18 [22:45:07] https://www.wikidata.org/wiki/Special:EntityData/Q19360262.json [22:45:10] yikes [22:45:29] aude: Ah, right... we should do that as well, probably [22:45:55] we can investigate the other one but don't want to stop Trijnstel from editing [22:46:18] somehow it was incompletely "saved" [22:46:28] mysql:wikiadmin@db1058 [wikidatawiki]> SELECT * FROM wb_entity_per_page WHERE epp_entity_type = 'item' AND epp_entity_id = 19360262; [22:46:29] hmh, well, otherwise it has to wait until Friday :P (will be away tomorrow) [22:46:29] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10§ Wikidata-Sprint-2015-02-03, 10§ Wikidata-Sprint-2015-02-25: Edit buttons are getting cached, although their visibility depends on user rights - https://phabricator.wikimedia.org/T85252#1067962 (10daniel) a:5aude>3daniel [22:46:41] I'll run the epp rebuild script [22:46:42] I assume you don't need the error messages anymore? [22:46:54] hoo: ok [22:47:02] hopefully that fixes [22:47:10] Trijnstel: we got them [22:48:16] :D [22:50:29] DanielK_WMDE_: You just went through all the fun I went through when debugging aude's parser cache change... fun [22:51:12] DanielK_WMDE_: i'm not totally sure it's that simple but worth a try [22:51:36] It's not really [22:51:42] edit section is actually a different thing [22:51:45] but we can abuse it [22:51:51] maybe [22:52:03] it usually means that people can edit individual sections (which is a different user right) [22:52:04] not sure it works same way exactly as splitting by language [22:52:12] ah [22:52:17] But doesn't have anything to do with the editability of the whole page [22:52:19] aude: i just spent two hours verifying manually [22:52:21] well, not necessary [22:52:26] it's totally insane [22:52:28] and it works [22:52:48] Trijnstel: Here you https://www.wikidata.org/wiki/Q19360262 :) [22:53:05] weeeeeh :D [22:53:05] https://github.com/wikimedia/mediawiki/blob/master/includes/page/Article.php#L524 is where we might want to deviate from core [22:53:07] hoo: core uses it for exactly the same thing we want it for: to decide whether edit links hsould be shown [22:53:09] thnx!! [22:53:12] and show edit buttons but disabled [22:53:22] hoo: we leave it to core to decide when that should be [22:53:23] with the hover notice explaining why [22:53:37] that's not abusing the flag, that's *exactly* what it's ther efor [22:53:37] DanielK_WMDE_: Mh... core has different mechanisms for that (as explained above) [22:53:53] no edit buttons and no padlock and no clue is not nice [22:54:06] of course, we could have padlock [22:54:14] aude: how we render non-editable items is yet another issue [22:54:37] hoo: it may have different criteria. [22:54:41] would need to look at your patch... [22:54:45] Ok, if it still works [22:54:45] is henning snater still on the team? [22:54:48] but i think the one i saw is actually userCan( 'edit' ) [22:55:18] despens: yes, but we are sharing him with another team now. [22:55:21] maybe padlock is indeed best... [22:55:31] now that there are page indicators [22:55:52] aude: careful though - editsection will also be false when viewing old revisions and diffs [22:55:59] we don't want a padlock there [22:56:00] DanielK_WMDE_: yes [22:56:08] that's what makes this tricky [22:56:20] padlock is just for userCan [22:56:31] and not the iscurrent part [22:56:34] but i think we turn off cachign for these anyway [22:56:37] (also isPrintable) [22:56:40] pretty sure we do. or core does [22:56:51] not so sure [22:56:56] printable is a separate flag, another cache split [22:57:10] yeah, but has the same caching problem afaik [22:57:27] based on manual poking at this [22:58:02] Fun fact: editsection links are not made in Parser. They're made in [22:58:02] Linker, in Skin *and* in ParserOutput. [22:58:06] :P [22:58:21] omg, i know :o [22:58:52] look, https://test.wikidata.org/wiki/Q22 [22:58:55] no edit links [22:59:00] it's not protected [22:59:09] what did you do? [22:59:13] print vieW? [22:59:22] printable=yes&action=purge [22:59:27] m( [22:59:36] same issue as protected items [22:59:43] needs same solution [22:59:55] Yeah... add that to the bug? [22:59:55] people complain about edit links on printable [23:01:19] people print wikidata pages? [23:01:26] dennyvrandecic: apparenly [23:01:33] that's... awesome :? [23:01:48] not too many people complain but i've heard of that issue before [23:01:54] (03PS1) 10Hoo man: Update Wikibase to fix EntityViewPlaceholderExpander [extensions/Wikidata] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192974 [23:02:08] * hoo wants to Q1-Q1000 paperback :D [23:02:11] * the [23:02:20] :) [23:02:35] hoo: how much will you pay for it? [23:02:43] (03PS1) 10Chad: Remove obvious function-level profiling [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192975 [23:03:15] mh... depends on how much that actually is [23:03:23] but given it's funny: Probably a bit [23:03:37] i guess it is about 900-950 items :) [23:03:40] (03CR) 10jenkins-bot: [V: 04-1] Remove obvious function-level profiling [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192975 (owner: 10Chad) [23:04:11] mh [23:04:20] Wikidata, Vol. 18: Q1701-Q1800 [23:04:25] :D [23:04:37] I guess they would cover more items later on [23:04:51] yeah, probably [23:04:59] A few weeks ago I made a graph of 100k items/ size [23:05:14] It almost steadily declined, except for some peaks [23:05:30] no surprise [23:06:09] (03PS1) 10Daniel Kinzler: Split parser cache by editsection flag [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192976 (https://phabricator.wikimedia.org/T89962) [23:06:26] hoo, aude: --^ [23:07:21] (03CR) 10jenkins-bot: [V: 04-1] Split parser cache by editsection flag [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192976 (https://phabricator.wikimedia.org/T89962) (owner: 10Daniel Kinzler) [23:07:34] jenkins broken? [23:07:59] 23:07:03 chmod: cannot access `/mnt/home/jenkins-deploy/tmpfs/mwext-Wikibase-repo-tests/wikicache.*': No such file or directory [23:08:16] bah [23:08:23] screw that [23:10:22] DanielK_WMDE_: I think you're doing the same mistake Katie did [23:10:31] aude: ^ true? [23:10:59] EntityParserOutputGenerator is not the place to set these [23:11:04] i'm not convinced daniel's patch alone is sufficient [23:11:17] It probably just breaks caching completely [23:11:23] like your first PS did back then [23:11:28] but i'm too tired to manually try and poke [23:11:31] hoo: i know [23:12:31] * aude needs to be at the office earlier than usual tomorrow :o [23:12:45] oO why? [23:12:53] some meeting or something i think [23:13:07] was early today also [23:13:25] anyway, have more things to review for tomorrow :) [23:14:38] (03CR) 10Hoo man: [C: 032] Update Wikibase to fix EntityViewPlaceholderExpander [extensions/Wikidata] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192974 (owner: 10Hoo man) [23:15:32] DanielK_WMDE_: https://gerrit.wikimedia.org/r/188277 see my first two comments here [23:16:43] (03Merged) 10jenkins-bot: Update Wikibase to fix EntityViewPlaceholderExpander [extensions/Wikidata] (wmf/1.25wmf19) - 10https://gerrit.wikimedia.org/r/192974 (owner: 10Hoo man) [23:22:03] (03CR) 10Hoo man: "At a glance I think this suffers the same problems the first PS of https://gerrit.wikimedia.org/r/188277 did (see my comments there): We a" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192976 (https://phabricator.wikimedia.org/T89962) (owner: 10Daniel Kinzler) [23:26:22] (03PS1) 10Daniel Kinzler: Automatically split parser cache. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192986 [23:27:14] (03CR) 10jenkins-bot: [V: 04-1] Automatically split parser cache. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/192986 (owner: 10Daniel Kinzler) [23:28:19] DanielK_WMDE_: Does that fix the issues I suspect? [23:29:21] hoo: i don't understand what issue you suspect... and i spent quite a while tonight looking at parser cache keys :) [23:29:57] DanielK_WMDE_: Mh, ok... I'll leave it to Katie to review that, probably [23:29:58] hoo: what do you mean by "different"? yes, we use several different keys. that's how we split the cache [23:30:27] the problem I suspect is taht we save with different keys than what we use to retrieve [23:30:33] that way we have cache misses all the time [23:31:24] there is something in EntityHandler for that afaik [23:31:31] hoo: note that there is one more level of indirection: the list of *options* used to generate the cache keys is *cached* [23:32:06] DanielK_WMDE_: Yeah, I do know that [23:32:57] hoo, aude: oh! now i know what that strange wb1 bit was i kept seeing in the key! [23:33:13] i didn't pay attanetion to whether it was always there though. should verify. [23:37:12] hoo: i can verify cache hits as expected [23:37:23] i set a breakpoint and just look when it hits the cache. [23:37:46] yeah, guess that works [23:37:55] I usually use curl ... | grep 'Saved in' [23:38:14] nice :) [23:38:31] hoo: but it seems to work as desired. no misses [23:39:33] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10wikidata-query-service, 7Tracking: Query by one property and one value ("simple queries") (tracking) - https://phabricator.wikimedia.org/T54385#1068234 (10kaldari) [23:39:33] * aude hope it's that simple though suspect there are more pieces of the puzzle for this [23:39:39] at least on the ui part [23:40:11] There's a JS and a php part to showing edit links, yes [23:40:19] yeah [23:40:39] i think the ui does something with the edit links of the page isn't editable [23:40:49] i think that's dead code, because that should never happen [23:40:52] if the edit links are there [23:40:55] it was only exposed through the bug [23:41:00] exactly [23:41:15] i kind of like edit links but disabled with a hint [23:41:24] if we can't do that, then we need padlocks [23:41:26] we can start emitting them [23:41:34] it doesn't mappter *what* we do based on the flag [23:41:37] or both maybe [23:41:47] the cache now works correctly, we can go crazy on the links :) [23:41:52] ok [23:42:21] also for printable? [23:46:26] aude: should. getIsPrintable has $this->optionUsed( 'printable' ); [23:47:03] hm [23:47:05] aude: after the "magic" split patch is in, calling that getter should split the cache [23:47:14] no idea if we ever do call that getter [23:47:26] don't think we do [23:47:28] but if we don't, the output would be the same anyway, so no reason to split the cache [23:47:40] if we don't, why split the cache? [23:47:55] people don't want edit buttons in printable view [23:48:04] not supposed to be there [23:48:07] aude: oh, the getter must be called *after* the parserOutput was created [23:49:01] core somethow handles edit buttons in the skin or something strange [23:49:52] aude: in EntityParserOutputGenerator, do $options->setEditSections( $options->getEditSections() || $options->getIsPrintable() ). [23:50:08] you can do that before the ParserOutput is created and registered to prevent the cache split [23:50:12] that should work [23:50:21] (we don't need the extra split on printable, if all it does is set editsections=false) [23:50:31] yeah [23:50:36] 10MediaWiki-extensions-WikibaseRepository, 10Wikidata, 10wikidata-query-service, 7Tracking: Query by one property and one value ("simple queries") (tracking) - https://phabricator.wikimedia.org/T54385#1068267 (10kaldari) It doesn't look like there has been any activity on any of the blockers for this bug i... [23:50:48] ...and if we do start emitting special html for print mode, we would be calling the getter again [23:51:25] aude: wil lyou make a patch or file it? i'll probably forgot, and I'm about to go to bed... [23:51:55] * aude too tired [23:52:04] will think about and not forget when reviewing [23:52:10] :) [23:52:21] gn8 [23:52:26] pretty sure some additional tickets are needed for the ui parts at least [23:52:49] I'm not that sure... I mean it works pretty consistent with the PHP state now (when broken) [23:53:08] So it probably just uses taht state in soem way... I think that changed a lot [23:53:09] i checked otu the patch [23:53:13] no edit buttons [23:53:30] meaning no clue why no edit buttons [23:54:12] minimum, we need a padlock if we do this way [23:54:29] I guess proper fix is not that hard [23:54:50] daniel's patch is certainly an improvement [23:54:58] as long as no issues with cache misses etc [23:55:18] padlock or disabled edit buttons (with hover clue) or both [23:56:50] anyway, will stick around few minutes for swat and then done for today [23:57:33] Ok, I'll also be around a bit longer... and then two weeks leave :P [23:57:48] what??? [23:58:01] * aude panics [23:58:13] It's my last day until March 13 [23:58:17] aaah [23:58:40] well, enjoy (holidays i hope) [23:58:49] No, not at all... :( [23:58:55] University all the way [23:58:57] grrr [23:59:16] but ok