[00:05:51] (03PS1) 10Hoo man: Avoid "too much recursion" in jquery.wikibase.sitelinklistview [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163778 (https://bugzilla.wikimedia.org/71453) [00:21:16] [13WikibaseQuery] 15mariushoch created 06jQueryJson (+1 new commit): 02http://git.io/5lpEHg [00:21:16] 13WikibaseQuery/06jQueryJson 14f368916 15Marius Hoch: Remove dependency on jquery.json in favour of the JSON polyfill... [00:21:36] [13WikibaseQuery] 15mariushoch opened pull request #53: Remove dependency on jquery.json in favour of the JSON polyfill (06master...06jQueryJson) 02http://git.io/s-pPcQ [00:23:00] [travis-ci] wmde/WikibaseQuery/jQueryJson/f368916 : Marius Hoch The build has errored. http://travis-ci.org/wmde/WikibaseQuery/builds/36623736 [06:20:09] [13WikibaseDataModelJavascript] 15snaterlicious 04force-pushed 06fingerprint from 1424ff846 to 14d066ff8: 02http://git.io/AW1kmQ [06:20:09] 13WikibaseDataModelJavascript/06fingerprint 14f3efd7c 15snaterlicious: Prevent using init.mw.php as entry point [06:20:09] 13WikibaseDataModelJavascript/06fingerprint 141d54ac1 15snaterlicious: Renamed namespace.js to __namespace.js [06:20:09] 13WikibaseDataModelJavascript/06fingerprint 144d1c06f 15snaterlicious: Added Term [06:20:21] [13WikibaseSerializationJavaScript] 15snaterlicious pushed 3 new commits to 06multilingualunserializer: 02http://git.io/IuQiqw [06:20:21] 13WikibaseSerializationJavaScript/06multilingualunserializer 14c46794d 15snaterlicious: Test consolidations [06:20:21] 13WikibaseSerializationJavaScript/06multilingualunserializer 146c7bae6 15snaterlicious: Removed obsolete init.js [06:20:21] 13WikibaseSerializationJavaScript/06multilingualunserializer 144d33b50 15snaterlicious: Use version constant in init.mw.php [06:23:18] [13WikibaseSerializationJavaScript] 15snaterlicious 04force-pushed 06multilingualunserializer from 144d33b50 to 144e37d22: 02http://git.io/is_rIw [06:23:18] 13WikibaseSerializationJavaScript/06multilingualunserializer 14771d231 15snaterlicious: Untangled ResourceLoader modules [06:23:18] 13WikibaseSerializationJavaScript/06multilingualunserializer 141bf57e7 15snaterlicious: Reflect contributors [06:23:18] 13WikibaseSerializationJavaScript/06multilingualunserializer 148a9dd99 15snaterlicious: Extracted MultilingualUnserializer from EntityUnserializer [06:29:22] [13WikibaseSerializationJavaScript] 15snaterlicious pushed 1 new commit to 06multilingualunserializer: 02http://git.io/kC5GZQ [06:29:22] 13WikibaseSerializationJavaScript/06multilingualunserializer 14b1b6dac 15snaterlicious: Removed unused scope parameters [06:42:59] [13WikibaseDataModelJavascript] 15snaterlicious pushed 1 new commit to 06fingerprint: 02http://git.io/efmgtQ [06:42:59] 13WikibaseDataModelJavascript/06fingerprint 14a25dc23 15snaterlicious: Updated constructor names [07:02:13] (03CR) 10Henning Snater: "Unable to reproduce. Did you purge the page?" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/161453 (owner: 10Henning Snater) [07:21:57] (03CR) 10Henning Snater: "Although I definitely support scraping the API interaction out of the views, I am not sure this is heading into the right direction. It wo" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/162555 (owner: 10Adrian Lang) [07:36:54] Thiemo_WMDE: benestar|cloud: noooooooooo [07:36:56] Check this http://www.mediawiki.org/wiki/Wikibase/DataModel/Primer [07:37:02] point 1.6 [07:37:13] Claims, consisting of: Statements [07:37:16] *boom* [07:37:44] Oh, looks like it's just switched [07:46:01] (03PS1) 10Henning Snater: Badge selector: Reposition menu when initializing [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163805 [07:46:03] (03PS1) 10Henning Snater: badge selector: Adapt styles of existing selectors [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163806 [07:46:05] (03PS1) 10Henning Snater: badge selector: Display item id if no label exists [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163807 [08:01:49] (03CR) 10Thiemo Mättig (WMDE): "setTimeout? Feels like a not-so-nice hack. Isn't there an other way to avoid this? Henning?" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163778 (https://bugzilla.wikimedia.org/71453) (owner: 10Hoo man) [08:06:25] Lydia_WMDE: :( https://bugzilla.wikimedia.org/show_bug.cgi?id=71459 [08:06:46] Thiemo_WMDE: hey! welcome back! more fun with years ---^ [08:13:56] (03PS11) 10Daniel Kinzler: Introduce usage tracking service. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/161477 (https://bugzilla.wikimedia.org/64586) [08:23:50] (03CR) 10jenkins-bot: [V: 04-1] Introduce usage tracking service. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/161477 (https://bugzilla.wikimedia.org/64586) (owner: 10Daniel Kinzler) [08:28:11] [13WikibaseQuery] 15thiemowmde pushed 1 new commit to 06jQueryJson: 02http://git.io/T2gQ5A [08:28:11] 13WikibaseQuery/06jQueryJson 1494ecad6 15Thiemo Mättig: Update DataModel [08:28:30] *sigh* [08:29:00] (03CR) 10Henning Snater: "I guess there is no other short-term solution. wbsetentity does not allow setting multiple site links at once. I would like to move to usi" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163778 (https://bugzilla.wikimedia.org/71453) (owner: 10Hoo man) [08:32:50] [travis-ci] wmde/WikibaseQuery/jQueryJson/94ecad6 : Thiemo Mättig The build failed. http://travis-ci.org/wmde/WikibaseQuery/builds/36646440 [08:42:18] (03PS12) 10Daniel Kinzler: Introduce usage tracking service. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/161477 (https://bugzilla.wikimedia.org/64586) [08:44:57] DanielK_WMDE: :/ [08:45:09] Thiemo_WMDE: :) [08:45:17] [13WikibaseQuery] 15thiemowmde pushed 1 new commit to 06jQueryJson: 02http://git.io/vfFYHA [08:45:17] 13WikibaseQuery/06jQueryJson 14bdf25c7 15Thiemo Mättig: Fix namespace [08:48:59] [travis-ci] wmde/WikibaseQuery/jQueryJson/bdf25c7 : Thiemo Mättig The build is still failing. http://travis-ci.org/wmde/WikibaseQuery/builds/36647619 [08:56:42] (03PS12) 10Daniel Kinzler: Collect usage tracking info in ParserOutput. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/161487 (https://bugzilla.wikimedia.org/64588) [08:59:00] (03PS9) 10Daniel Kinzler: Record usage info from ParserOutput to the database. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/162650 (https://bugzilla.wikimedia.org/64589) [09:01:47] [13WikibaseDataModel] 15JeroenDeDauw comment on pull request #205 14961ece1: Well go ahead, click the edit button 02http://git.io/LFTWCA [09:02:37] [13WikibaseDataModel] 15JeroenDeDauw comment on pull request #205 14961ece1: It is not optional. Though an explicit check for it would indeed be good 02http://git.io/fIoRYg [09:08:18] DanielK_WMDE: https://bugzilla.wikimedia.org/show_bug.cgi?id=71458 [09:10:02] (03CR) 10jenkins-bot: [V: 04-1] Record usage info from ParserOutput to the database. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/162650 (https://bugzilla.wikimedia.org/64589) (owner: 10Daniel Kinzler) [09:13:11] JeroenDeDauw: do you have a minute? composer is blocking me again. [09:13:17] (03CR) 10Adrian Lang: "Using setTimeout for breaking the stack and executing code asynchronously is fine from my point of view." [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163778 (https://bugzilla.wikimedia.org/71453) (owner: 10Hoo man) [09:14:17] (03CR) 10Adrian Lang: "Moving to wbsetentity seems to be a lot easier with this step, since you would not have to touch the views at all." [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/162555 (owner: 10Adrian Lang) [09:19:24] Lydia_WMDE: have you seen the stats ? https://tools.wmflabs.org/wikidata-todo/stats.php?reverse they are crazy ... I do not trust them at all [09:20:23] it gives the impression something major has happened [09:20:46] GerardM-: it might be the change of the dump format. or the introduction of redirects [09:20:56] i'd bet on the former [09:21:06] leading to the parser not handling all items [09:21:14] need to leave for a few mins [09:21:40] the introduction of 1,761,373 items without a label ? [09:23:01] does that mean that things are truly broken ? [09:26:59] JeroenDeDauw: what changed in WBQ? it keep yelling at me. [09:27:32] Thiemo_WMDE: the thing is that nothing changes in it while things in its dependencies changed [09:28:01] Thiemo_WMDE: I can have a look at the errors if you want [09:28:08] The DM related changes should be easy to match [09:28:13] JeroenDeDauw: the errors make no sense to me. it misses a table called repowiki.unittest_wbq_mainsnak_boolean [09:28:23] there are no DM related errors. [09:28:31] Oh dear... [09:28:35] That evil stuff [09:28:52] Thiemo_WMDE: perhaps you don't have QueryEngine installed? [09:29:09] Thiemo_WMDE: WBQ now has a CLI that has an uninstall and an install command [09:29:13] So just reinstall QE [09:29:16] via the CLI [09:29:20] QE is a lib in WBQ. [09:29:27] arg... [09:29:29] Indeed [09:29:45] Thiemo_WMDE: not the installation of the lib [09:29:50] The tables the lib installs [09:30:16] Thiemo_WMDE: it is not actually a library, it's an infrastructure component [09:31:23] i have it "installed" (this word keeps confusing me, creating a schema is not an "installation"). [09:31:37] note the "unittest_" in the missing table name. [09:34:22] JeroenDeDauw: https://github.com/wmde/WikibaseQuery/pull/48/files is still waiting. but bot related to my current problem. [09:36:40] [13WikibaseQuery] 15JeroenDeDauw closed pull request #48: Rearrange CLI command names (06master...06commands) 02http://git.io/SIBcWw [09:38:52] Thiemo_WMDE: no its not! :) [09:39:32] [travis-ci] wmde/WikibaseQuery/master/3cfd62b : Jeroen De Dauw The build has errored. http://travis-ci.org/wmde/WikibaseQuery/builds/36651621 [09:40:30] DanielK_WMDE: Tobi_WMDE_SWE: are you calling in for the daily? [09:41:01] GerardM-: yeah that does indeed seem weird. but as i said it might just be that the parsing broke because of the new dump format. magnus would need to have a look [09:42:43] [13WikibaseQuery] 15thiemowmde 04force-pushed 06jQueryJson from 14bdf25c7 to 14ca2c957: 02http://git.io/avAKRg [09:42:43] 13WikibaseQuery/06jQueryJson 149225f7f 15Marius Hoch: Remove dependency on jquery.json in favour of the JSON polyfill... [09:42:43] 13WikibaseQuery/06jQueryJson 14ca2c957 15Thiemo Mättig: Update DataModel [09:42:56] Tobi_WMDE_SWE: did you get around to creating tasks for the remaining stuff for simple query move away from dbal? [09:43:30] (03PS10) 10Daniel Kinzler: Record usage info from ParserOutput to the database. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/162650 (https://bugzilla.wikimedia.org/64589) [09:43:51] yay, mystery failures :/ [09:45:10] Lydia_WMDE: i'll join the hangout in a minute [09:45:26] k [09:46:03] [travis-ci] wmde/WikibaseQuery/jQueryJson/ca2c957 : Thiemo Mättig The build is still failing. http://travis-ci.org/wmde/WikibaseQuery/builds/36651975 [09:51:46] (03CR) 10jenkins-bot: [V: 04-1] Record usage info from ParserOutput to the database. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/162650 (https://bugzilla.wikimedia.org/64589) (owner: 10Daniel Kinzler) [10:02:40] [13WikibaseDataModel] 15thiemowmde pushed 1 new commit to 06statement_: 02http://git.io/pNeBww [10:02:40] 13WikibaseDataModel/06statement_ 148ce9df0 15thiemowmde: Update Statement.php [10:04:26] (03PS13) 10Daniel Kinzler: Introduce usage tracking service. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/161477 (https://bugzilla.wikimedia.org/64586) [10:04:35] (03PS13) 10Daniel Kinzler: Collect usage tracking info in ParserOutput. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/161487 (https://bugzilla.wikimedia.org/64588) [10:04:44] (03PS11) 10Daniel Kinzler: Record usage info from ParserOutput to the database. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/162650 (https://bugzilla.wikimedia.org/64589) [10:09:22] DanielK_WMDE: https://bugzilla.wikimedia.org/show_bug.cgi?id=71461 if you have any clue [10:11:58] JeroenDeDauw: so there are no tests running? https://github.com/wmde/WikibaseEntityStore/pull/4 they also do not run on my machine. still ok to merge? needed for WBQ. [10:12:36] ... uh, tests just started. ok. [10:14:39] Henning_WMDE: Adrian_WMDE: https://gerrit.wikimedia.org/r/163778 merge? [10:15:26] I also thought that we might want to break after 10 iterations or so each, that would prevent a lot of browser redrawing (but is less nice) [10:15:36] merge works for me [10:15:47] Henning_WMDE: https://bugzilla.wikimedia.org/show_bug.cgi?id=58990 <- can be closed? [10:17:28] aude: ^ If we plan to go for Wikidata today that *needs* to be backported [10:18:25] hoo: ok [10:19:23] I realized we don't even have any example item on testwikidata for that [10:19:45] (03CR) 10jenkins-bot: [V: 04-1] Record usage info from ParserOutput to the database. [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/162650 (https://bugzilla.wikimedia.org/64589) (owner: 10Daniel Kinzler) [10:19:46] I only noticed it while testing another change locally tonight :P [10:23:12] hoo: Your change is fine like it is for now. We should switch to editentity at some point anyhow. [10:23:44] Sure thing [10:24:02] Lydia_WMDE: I guess it can be closed. [10:24:15] ok [10:24:42] (03CR) 10Henning Snater: [C: 032] Avoid "too much recursion" in jquery.wikibase.sitelinklistview [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163778 (https://bugzilla.wikimedia.org/71453) (owner: 10Hoo man) [10:25:12] (03PS1) 10Aude: Avoid "too much recursion" in jquery.wikibase.sitelinklistview [extensions/Wikibase] (wmf/1.25wmf1) - 10https://gerrit.wikimedia.org/r/163822 (https://bugzilla.wikimedia.org/71453) [10:35:32] (03Merged) 10jenkins-bot: Avoid "too much recursion" in jquery.wikibase.sitelinklistview [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163778 (https://bugzilla.wikimedia.org/71453) (owner: 10Hoo man) [10:38:09] (03PS1) 10Adrian Lang: Don't try to get entity null or undefined from store [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163824 [10:42:59] jzerebecki: JeroenDeDauw: https://bugzilla.wikimedia.org/show_bug.cgi?id=68379#c1 <- this has now been done right? [10:45:00] reproduced the bug in vagrant [10:45:22] wow, that was annoying though to switch it between hhvm and php5 [10:47:54] Lydia_WMDE: yes, updated the bug [10:48:46] aude: Nice, I guess [10:49:55] needs my memcached role [11:01:56] [13WikibaseDataModelJavascript] 15thiemowmde comment on pull request #16 14a25dc23: Typo (2 times). 02http://git.io/0EXAGQ [11:19:55] [13WikibaseDataModelJavascript] 15thiemowmde comment on pull request #16 14a25dc23: Uh, this is a nested loop. What's wrong with `if( !this._items[i].equals( list._items[i] ) )`? 02http://git.io/TJzaxA [11:24:43] [13WikibaseDataModelJavascript] 15thiemowmde comment on pull request #16 14a25dc23: Indention. 02http://git.io/hDPUvQ [11:24:49] aude: can you log in to wikidata-jenkins2.eqiad.wmflabs ? [11:24:58] [13WikibaseDataModelJavascript] 15thiemowmde comment on pull request #16 14a25dc23: Indention. 02http://git.io/dr2ipg [11:26:07] jzerebecki: I can't [11:27:21] damn how hard is it to replace those with new instances? besides setting them up, they need to be registered somehow with jenkins? [11:27:29] i can via bastion [11:27:37] aude: Really? [11:27:43] ssh wikidata-jenkins2 [11:28:29] and not via bastion [11:28:47] omnomnom [11:29:14] Lydia_WMDE: that item is about the UI? So how should I know? [11:29:16] [13WikibaseDataModelJavascript] 15thiemowmde comment on pull request #16 14a25dc23: Indention of the constructor function is a bit strange, compare to `List.js`. 02http://git.io/TFatMQ [11:29:31] Lydia_WMDE, aude: what bugs should I look at?... [11:30:54] aude: doh https://wikitech.wikimedia.org/wiki/Nova_Resource:Wikidata-build [11:30:57] of course I can't ssh [11:31:01] I'm not in the project even [11:31:03] m( [11:31:05] * hoo hides [11:33:59] [13WikibaseDataModelJavascript] 15thiemowmde comment on pull request #16 14a25dc23: Not relevant at the moment, just a notice: This will fail if a `Set` is used to store boolean or null values. Same is true for some other methods like `hasItemForKey` above. 02http://git.io/hvLn_Q [11:36:39] aude: the serialization issue does not look like it's easy to fix... [11:37:44] DanielK_WMDE: Can't we "simply" implement __serialize and put a stable serialization there? [11:37:53] (03CR) 10Adrian Lang: [C: 032] "I broke this. I still think two calls are unnecessary, but it works and I don't think investigating this makes sense right now." [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163805 (owner: 10Henning Snater) [11:39:12] (03CR) 10Adrian Lang: [C: 032] badge selector: Adapt styles of existing selectors [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163806 (owner: 10Henning Snater) [11:40:41] (03CR) 10Adrian Lang: [C: 032] badge selector: Display item id if no label exists [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163807 (owner: 10Henning Snater) [11:42:11] hoo: that would require the Entity to know the serialization service. Which it definitly shouldn't. [11:42:38] hoo: we can just serializer/unserialkize explicitly. easy enough, though may need some refactoring to accommodate EntityRevision. [11:42:46] but that's not a quick fix for a backport. [11:42:56] yes :S [11:43:29] i'd really love to see the offending blob, though [11:43:34] But how is explicit serialization different here? Does that use a more robust format? [11:43:46] yes. [11:43:47] json. [11:43:55] Ah, right [11:43:58] the canonical json, that is. [11:44:13] (with some tweaks, maybe) [11:44:17] if you give me the memcached key, I can pull you the blob [11:44:20] easy enough [11:44:27] aks katie [11:45:18] [13WikibaseDataModelJavascript] 15snaterlicious pushed 2 new commits to 06fingerprint: 02http://git.io/lk5siA [11:45:18] 13WikibaseDataModelJavascript/06fingerprint 14962ef93 15snaterlicious: Refactored List/Set [11:45:18] 13WikibaseDataModelJavascript/06fingerprint 14d5eb4ea 15snaterlicious: Consolidated Fingerprint [11:46:13] [13WikibaseDataModelJavascript] 15snaterlicious 04force-pushed 06fingerprint from 14d5eb4ea to 14dd78e69: 02http://git.io/AW1kmQ [11:46:13] 13WikibaseDataModelJavascript/06fingerprint 14dd78e69 15snaterlicious: Consolidated Fingerprint [11:49:30] (03CR) 10Daniel Kinzler: Conditionally add the "capiunto.infobox.main" RL module (031 comment) [extensions/Capiunto] - 10https://gerrit.wikimedia.org/r/163065 (https://bugzilla.wikimedia.org/69540) (owner: 10Hoo man) [11:50:08] DanielK_WMDE: hoo it's saved in binary for some odd reason [11:51:07] i can think of some work arounds but want to know exactly waht the bug is [11:51:34] [13WikibaseDataModelJavascript] 15snaterlicious pushed 1 new commit to 06fingerprint: 02http://git.io/DgY9TA [11:51:34] 13WikibaseDataModelJavascript/06fingerprint 1463f6e29 15snaterlicious: Updated Readme [11:52:08] (03Merged) 10jenkins-bot: Badge selector: Reposition menu when initializing [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163805 (owner: 10Henning Snater) [11:52:10] (03CR) 10Hoo man: "Luckily MediaWiki isn't that dumb" (031 comment) [extensions/Capiunto] - 10https://gerrit.wikimedia.org/r/163065 (https://bugzilla.wikimedia.org/69540) (owner: 10Hoo man) [11:52:11] aude: it has to be saved in binary, php serializations of objects contain null bytes. [11:52:17] as in \0 [11:52:38] (03Merged) 10jenkins-bot: badge selector: Adapt styles of existing selectors [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163806 (owner: 10Henning Snater) [11:52:39] those are funny to work with :D [11:52:40] (03Merged) 10jenkins-bot: badge selector: Display item id if no label exists [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163807 (owner: 10Henning Snater) [11:52:57] hoo: would you have peek at https://gerrit.wikimedia.org/r/#/c/162650/ ? I must be missing something silly. [11:53:35] Sure [11:54:21] hoo: the fun bit is: i was able to reproduce this locally this morning. not any more. and i didn't change a thing... [11:54:56] the unreported error was saying that newFromText() is an unknown method in the Mock for TitleFactory. which isn't true. [11:54:58] *shrug* [11:55:21] Thiemo_WMDE: guess you want a rel of EntityStore? [11:56:25] lolwat https://github.com/wmde/Diff/pull/12 [11:56:30] So now the PR is void [11:56:35] [13WikibaseDataModelJavascript] 15snaterlicious pushed 1 new commit to 06fingerprint: 02http://git.io/sv-lgQ [11:56:35] 13WikibaseDataModelJavascript/06fingerprint 149abd1af 15snaterlicious: Fixed indentation [11:56:45] JeroenDeDauw: it was about the specific comment not related to the UI ;-) (but jan answered so no worries) [11:57:09] Lydia_WMDE: where are you hiding? (you evil person!) [11:57:48] i was having lunch with paul [11:57:52] at my desk now ;-) [11:57:55] [13WikibaseDataModelJavascript] 15thiemowmde comment on pull request #16 149abd1af: Just wondering. The constructor in WikibaseDataModel is missing the third rank parameter. Should we add it there? 02http://git.io/5PK5ng [11:58:26] aude: we can make CachingEntityRevisionLookup more resillient to serialization errors I guess... but that would just hide the isse :( [11:58:44] thought of that, but agree [11:59:03] aude: one thing I want to know is: does this happen even if the value was written by hhvm, or only if vanilla php writes to the cache, and hhvm tries to read the data? [11:59:18] seems only hhvm writes it then php5 reads it [11:59:29] not tried other way around [12:00:16] aude: a quick fix would be a try{} catch (RuntimeException) {} around any call to $this->cache->get() in CachingEntityRevisionLookup. [12:00:25] The catch clause should call wfLogWarning [12:00:26] yep [12:00:38] already had a patch for that [12:00:56] would eb ok for the deployment branch i guess [12:00:57] but not the best solution [12:01:06] indeed [12:01:10] PHP Fatal error: Call to a member function getOption() on a non-object in /var/www/html/mwt/wiki/extensions/BetaFeatures/includes/BetaFeaturesUtil.php on line 42 [12:01:13] *sigh* [12:01:24] hoo: made a bug for that [12:01:29] oh? [12:01:38] oh, i thought that was fixed [12:01:42] didn't get around to fixing [12:01:44] i have to go in like 10 minutes :/ [12:02:02] https://bugzilla.wikimedia.org/show_bug.cgi?id=71350 [12:02:45] what is the fix for that? or rather, what's the underlying problem? [12:03:06] i tried making a mock user that had that method [12:03:11] assign it to context [12:03:17] didn't work [12:03:43] best thing would be not to have the static funciton called [12:03:53] have a wrapper maybe as interface with beta features [12:04:19] DatabaseBase::rollback: Explicit rollback of implicit transaction. Something may be out of sync! [Called from DatabaseBase::rollback in /var/www/html/mwt/wiki/includes/db/Database.php at line 3645] [12:04:22] what the hell [12:04:26] (many of them) [12:06:50] hoo: oh, you too now? got that locally at some point. quite odd. [12:07:25] hoo: but these are not the actual errors. they are triggered when rollback() is called due to an earlier errors. [12:07:31] they are just hiding the actual problem [12:07:36] i have seen those as well [12:07:41] mh [12:08:15] (part of the problem is that startAtomic/endAtomic doesn't seem to have a corresponding rollbackAtomic) [12:08:27] anyway, off for now. bbl [12:16:09] (03PS1) 10Hoo man: Make SidebarHookHandlersTest::testNewFromGlobalState more robust [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163832 [12:16:14] aude: ^ [12:17:37] [13WikibaseDataModelJavascript] 15thiemowmde comment on commit 14962ef93: This is a bit odd, isn't it? Is it really necessary that this is in `Groupable`? Can't this be made specific to `Set`? 02http://git.io/cnzqaw [12:27:21] [13WikibaseDataModelJavascript] 15thiemowmde comment on pull request #16 149abd1af: This test here actually means that a `Group` can only contain `Set` objects, nothing else. I suggest to change the type hint here from `Groupable` to `Set`. 02http://git.io/a-EZVw [12:45:55] Instantiator\Exception\InvalidArgumentException; [12:45:57] mh? [12:46:03] DanielK_WMDE: ^ [12:46:18] haha [12:46:33] JeroenDeDauw: ? What component is that? [12:46:47] Trolololo Ocramius is trololo [12:46:54] Daniel uses it in Wikibase and I'm confused now [12:46:57] hoo: one that is getting pulled in by phpunit [12:47:11] So its not normally included in our code [12:47:16] * hoo kicks phpstorm [12:47:26] Yeah, annoying it does that [12:47:33] Thanks, btw :) [12:49:41] (03CR) 10Hoo man: [C: 04-1] ""Root problem" is the missing loop in SqlUsageTracker" (035 comments) [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/162650 (https://bugzilla.wikimedia.org/64589) (owner: 10Daniel Kinzler) [12:50:33] JeroenDeDauw: Want to merge the evilness https://gerrit.wikimedia.org/r/163832 ? :P [12:51:26] (03CR) 10Hoo man: [C: 032] Avoid "too much recursion" in jquery.wikibase.sitelinklistview [extensions/Wikibase] (wmf/1.25wmf1) - 10https://gerrit.wikimedia.org/r/163822 (https://bugzilla.wikimedia.org/71453) (owner: 10Aude) [12:51:35] aude: Lydia_WMDE: Anything else to backport? [12:54:27] Wait, story time is ongoing :P [12:54:43] * hoo keeps forgetting that although he gets notifications m( [13:00:12] hoo: probably some hacks around the memcached problem [13:00:38] aude: How? Can we use different cache keys for hhvm/ zend for now? [13:01:28] hoo: that's one thing [13:01:28] (03Merged) 10jenkins-bot: Avoid "too much recursion" in jquery.wikibase.sitelinklistview [extensions/Wikibase] (wmf/1.25wmf1) - 10https://gerrit.wikimedia.org/r/163822 (https://bugzilla.wikimedia.org/71453) (owner: 10Aude) [13:01:43] and maybe catch the exception and do fallback [13:01:59] (03PS8) 10Hoo man: Refactored toolbars [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/161453 (owner: 10Henning Snater) [13:02:52] (03CR) 10Hoo man: "Noted that this requires purging the parser cache in the commit message. After purging the page this worked for me (only tested briefly)." [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/161453 (owner: 10Henning Snater) [13:03:09] wfIsHHVM or defined( 'HHVM_VERSION' ) [13:03:11] * hoo dislikes having to purge the parser cache this often [13:03:21] aude: Such a hack is fine with me [13:03:28] * aude too [13:03:42] You will do that in configuration, I guess? [13:03:47] yes [13:03:48] If so, I'll make a new build [13:03:50] ok [13:03:50] and then the fallback [13:04:01] Also for backport today? [13:07:21] (03PS1) 10Aude: Workaround for hhvm memcached issue in CachingEntityRevisionLookup [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163838 [13:07:43] hoo: DanielK_WMDE ^ [13:07:49] (03CR) 10Hoo man: [C: 032] Don't try to get entity null or undefined from store [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163824 (owner: 10Adrian Lang) [13:09:20] aude: Wont this mean that hhvm/ zend (may) overwrite the same content over and over? [13:09:51] I guess hhvm traffic is low enough for that to be ok performance wise? [13:09:52] you have better idea? (we'll also vary the cache key) [13:10:40] Varying the cache key is a good idea... I'm not a huge fan of the try catch, though [13:10:57] suppose we want log warning also [13:11:02] Yeah [13:11:21] We had enough problems with not using cache and stuff there that hammered down es* and stuff [13:11:32] So we should be very careful [13:12:26] (03CR) 10Hoo man: [C: 04-1] "Needs to warn/ log" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163838 (owner: 10Aude) [13:13:59] (03CR) 10jenkins-bot: [V: 04-1] Workaround for hhvm memcached issue in CachingEntityRevisionLookup [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163838 (owner: 10Aude) [13:14:15] what [13:14:39] oh damn [13:16:15] github seems awry [13:16:21] (03PS2) 10Hoo man: Workaround for hhvm memcached issue in CachingEntityRevisionLookup [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163838 (owner: 10Aude) [13:16:55] aude: Will you do the configuration path? [13:16:58] * patch [13:19:08] hoo: i can [13:19:21] * aude free from meeting i:) [13:19:23] :) [13:19:42] I can as well, just want to avoid doing the work twice [13:19:51] hoo: go ahead [13:20:05] also the fact we don't have hhvm yet, although think we want it back asap [13:20:13] s/yet/now* [13:20:41] aude: are the global functions available there or do we need to use an extension function closure? [13:20:41] i don't know if there is an issue though with non-hhvm items and hhvm accessing them [13:21:05] should be there but we should check [13:21:46] aaaa [13:21:57] looks like the localisation cache varies based on hhvm [13:22:18] (03CR) 10Hoo man: [C: 04-1] "Needs changes to SummaryFormatterTest (I will probably do that sometime soonish)" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/159327 (owner: 10Anomie) [13:22:37] https://github.com/wikimedia/operations-mediawiki-config/blob/master/wmf-config/CommonSettings.php#L179-L182 [13:22:42] What happened to this jenkins job? https://integration.wikimedia.org/ci/job/mwext-Wikibase-repo-api-tests/886/console [13:22:55] 2% of tests in 6 Minutes [13:23:27] Adrian_WMDE: Jan is/ was updating the labs instances, might be related [13:23:29] jzerebecki: ^ [13:23:34] k [13:23:35] aude: I see [13:23:38] so should be fine [13:23:54] they might have had similar issues [13:24:00] maybe [13:25:21] (03CR) 10jenkins-bot: [V: 04-1] Don't try to get entity null or undefined from store [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163824 (owner: 10Adrian Lang) [13:25:56] (03PS2) 10Adrian Lang: Don't try to get entity null or undefined from store [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163824 [13:26:37] (03PS3) 10Aude: Workaround for hhvm memcached issue in CachingEntityRevisionLookup [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163838 [13:26:59] hoo: i would put that in the branch only and if it doesn't end up being logged / code hit, then we don't need it next time [13:27:19] aude: Sounds ok to me [13:27:33] * aude not happy about it and continues investigating the actual bug [13:27:47] but shared memcached seems not good idea generally [13:29:19] WMF uses memcached for things beyond caching, so splitting it up totally is not possible [13:29:22] eg. logins :S [13:29:25] jzerebecki: did Thiemo run away? [13:29:47] But the concept of having to worry about non-scalar PHP data in there is awry [13:30:09] JeroenDeDauw: no idea [13:30:27] hoo: +1 [13:30:35] i would like json serialization for entity revision [13:30:51] we can handle it more robustly, imho [13:32:24] Adrian_WMDE: i was finished with that instance hours before that job got started on it. [13:33:01] (03CR) 10Hoo man: [C: 04-2] "We don't want this on master" [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163838 (owner: 10Aude) [13:33:13] (03PS1) 10Hoo man: Workaround for hhvm memcached issue in CachingEntityRevisionLookup [extensions/Wikibase] (wmf/1.25wmf1) - 10https://gerrit.wikimedia.org/r/163843 [13:33:29] (03CR) 10Hoo man: [C: 032] Workaround for hhvm memcached issue in CachingEntityRevisionLookup [extensions/Wikibase] (wmf/1.25wmf1) - 10https://gerrit.wikimedia.org/r/163843 (owner: 10Hoo man) [13:33:49] I doubt it's even needed, but who knows [14:18:37] (03CR) 10WikidataJenkins: "Build Successful" [extensions/Wikidata] (wmf/1.25wmf1) - 10https://gerrit.wikimedia.org/r/163851 (owner: 10Aude) [14:20:29] hmmm, are badges not stored in page links? [14:20:38] not recorded* [14:21:15] (03CR) 10JanZerebecki: [C: 032] "Based on Hoo his review, just to trigger submission again." [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163824 (owner: 10Adrian Lang) [14:22:11] nah, they seem to be https://www.wikidata.org/wiki/Special:WhatLinksHere/Q17437796 [14:23:12] [13WikibaseDataModelSerialization] 15JeroenDeDauw pushed 1 new commit to 06master: 02http://git.io/GDPnew [14:23:12] 13WikibaseDataModelSerialization/06master 14c6d7bb3 15Jeroen De Dauw: Update .travis.yml [14:24:36] [travis-ci] wmde/WikibaseDataModelSerialization/master/c6d7bb3 : Jeroen De Dauw The build passed. http://travis-ci.org/wmde/WikibaseDataModelSerialization/builds/36675106 [14:29:42] (03CR) 10Aude: [C: 032] Update Wikidata, fix for jquery.wikibase.sitelinklistview.js and hhvm [extensions/Wikidata] (wmf/1.25wmf1) - 10https://gerrit.wikimedia.org/r/163851 (owner: 10Aude) [14:29:55] (03Merged) 10jenkins-bot: Don't try to get entity null or undefined from store [extensions/Wikibase] - 10https://gerrit.wikimedia.org/r/163824 (owner: 10Adrian Lang) [14:36:38] [travis-ci] wmde/WikibaseDataModelSerialization/master/c6d7bb3 : Jeroen De Dauw The build passed. http://travis-ci.org/wmde/WikibaseDataModelSerialization/builds/36675106 [14:37:41] [13Ask] 15JeroenDeDauw pushed 1 new commit to 06master: 02http://git.io/qHsogw [14:37:41] 13Ask/06master 1403c008b 15Jeroen De Dauw: Update phpunit.xml.dist [14:37:58] (03Merged) 10jenkins-bot: Update Wikidata, fix for jquery.wikibase.sitelinklistview.js and hhvm [extensions/Wikidata] (wmf/1.25wmf1) - 10https://gerrit.wikimedia.org/r/163851 (owner: 10Aude) [14:40:46] [travis-ci] wmde/Ask#35 (master - 03c008b : Jeroen De Dauw): The build was fixed. [14:40:46] [travis-ci] Change view : https://github.com/wmde/Ask/compare/3fffb4e73a3e...03c008b3fe07 [14:40:46] [travis-ci] Build details : http://travis-ci.org/wmde/Ask/builds/36676325 [14:52:00] I'm not sure if I'm understanding this correctly, but I've made a JSON database out of a table on Wikipedia. [14:52:13] I'd like to add it to the Wikidata for the article. [14:52:40] Is there a way for me to do that? [14:53:13] I don't know a tool for it. [14:53:24] What do you mean tool? [14:53:29] You should probably make a bot with the api. [14:53:29] I can't just upload the .json file? [14:53:33] no [14:54:56] But Wikidata items should have tabular data in a machine-readable format, right? [14:55:02] Or am I misunderstanding what Wikidata is about. [14:55:38] You can export Wikidata in json, but you can't just import a jsonfile I think. [14:56:16] OK, well either way, how is tabular data usually handled? [14:56:59] People write bots for it... [14:57:09] Maybe multichill can tell you some more. [14:57:20] (or tell me that I'm telling bull shit) [14:57:35] If you can point me at an article that has Wikidata for a table. [14:58:14] Generate tables with Wikidata? [14:58:36] Not possible within Wikipedia yet... Not sure. [14:59:03] OK, well like look at this item: https://en.wikipedia.org/wiki/Demographics_of_the_United_States [14:59:18] It has tables in it, see? [14:59:21] Yes. [14:59:31] But they're not machine-readable. [14:59:57] So whether you update the data by hand or create a tool to go back and forth, you'd think that that tabular array of data should be stored in WikiData eventually, right? [15:00:13] Yeah, that would be great. [15:00:25] But there is still a lot to do. [15:01:05] Is that actually within the current scope of the project? Or is it just primarily about organizing the relationships between items? [15:01:21] Everything is in scope. ;) [15:01:25] I mean there's not currently a way to store tabular data in any format on Wikidata? [15:01:52] If we have the data types for them... [15:02:04] And relevant properties... [15:10:28] aude: Lydia_WMDE Thiemo_WMDE https://dl.dropboxusercontent.com/u/7313450/entropy/gif/fiersome.gif [15:11:01] rofl [15:11:06] :D [15:11:22] mascot = done! [15:20:17] x0011BF: ji, still there? [15:20:23] *hi [15:20:24] Yes [15:21:05] Wikidata is not really meant for tabular data like demographic timeseries w/ several dimensions [15:21:15] it is best for entity-centric graph-like data [15:21:30] OK [15:21:33] Lydia_WMDE: https://bugzilla.wikimedia.org/show_bug.cgi?id=71469 (in case daniel needs something to do) [15:21:36] or anyone [15:21:40] There is a suggestion for tabular data, just a second [15:22:02] https://meta.wikimedia.org/wiki/DataNamespace [15:22:02] oh, it's a dennyvrandecic [15:22:20] hey JeroenDeDauw! the fiersome lion is awesome [15:23:13] O see/ [15:23:15] I see. [15:23:19] Yeah, I'll keep an eye on that. [15:23:47]