[00:04:30] hoo, my previous edit (January 14) probably helped a little =) [00:05:15] yeah, it dropped [00:05:25] but only by maybe 2million/ day [00:05:30] but not sure, the graph is hard to rad [00:05:31] * read [01:36:07] https://wikitech.wikimedia.org/wiki/Add_a_wiki#Wikidata lovely documentation… we might want to fix the sites stuff :P [02:07:46] Good evening [02:08:23] hi Halibutt [02:10:25] Is this the place to seek help with a template I'm creating, or is Wikidata:Project chat a better place? [02:11:24] You can get quick responses here, but at this time of the day you might now [02:11:29] I could help you [02:11:37] but it's past 3am here and I'm about to go to bed [02:11:58] Maybe someone else... otherwise just ask on Wikidata, you'll surely get a reply there [02:12:29] I need help with a meta/Wikipedia template I'm thinking of. I described the idea here https://www.wikidata.org/wiki/Wikidata:Project_chat#Import_WP_article_names_and_links. In short: is there a way to invoke interwiki links in a pre-defined language, and an empty field if there is none [02:13:19] With Lua, yes see https://www.mediawiki.org/wiki/Extension:Wikibase_Client/Lua#mw.wikibase.getEntity and https://www.mediawiki.org/wiki/Extension:Wikibase_Client/Lua#mw.wikibase.entity:getSitelink [02:13:36] so first you load the entity, like "local entity = mw.wikibase.getEntity()" [02:13:55] and then you can call :getSitelink( "abcwiki" ) on it [02:14:05] Anyway, I'm leaving now [02:14:49] ok, thanks [02:14:53] see you around [08:57:33] Adrian_WMDE: Hi [08:57:38] Hi [09:00:47] Adrian_WMDE: https://gerrit.wikimedia.org/r/#/c/274654/ [09:01:19] Adrian_WMDE: If i am on track, let me know :) [09:11:38] I'm afraid this is much more complicated than that [09:13:32] Adrian_WMDE: so can you brief me on what is needed? [11:45:36] https://test.wikidata.org/wiki/Property:P732 [12:11:08] PROBLEM - wikidata.org dispatch lag is higher than 300s on wikidata is CRITICAL: HTTP CRITICAL: HTTP/1.1 200 OK - pattern not found - 1138 bytes in 0.056 second response time [12:16:50] RECOVERY - wikidata.org dispatch lag is higher than 300s on wikidata is OK: HTTP OK: HTTP/1.1 200 OK - 1657 bytes in 0.152 second response time [14:18:33] hashar: you around? enkins sais "no space left on device" [14:22:40] DanielK_WMDE: oh man [14:22:58] https://integration.wikimedia.org/ci/computer/ [14:23:21] Mighty jenkins has disconnected the faulty slave already https://integration.wikimedia.org/ci/computer/integration-slave-trusty-1011/ [14:23:22] !! [14:24:33] hashar: awesome, thansk! [14:55:06] benestar: https://gerrit.wikimedia.org/r/#/c/274424/ [16:30:53] Lydia_WMDE: http://www.rottentomatoes.com/m/shaun_of_the_dead/ [16:31:56] Lydia_WMDE: and then http://www.rottentomatoes.com/m/hot_fuzz [16:32:23] Lydia_WMDE: that's the scene https://www.youtube.com/watch?v=hGX0mmmRccM [18:02:42] Thiemo_WMDE: merge?: https://gerrit.wikimedia.org/r/#/c/273933/ [18:09:35] Thiemo_WMDE: thx for the second pair of eyes [18:13:27] Thiemo_WMDE: You might want to join one of #wikimedia-editing or #wikimedia-dev as well. [18:53:29] okay, so i'm wondering how to report this: with the recent interface change that added the "Identifiers" grouping, there was a regression when adding additional statements to existing properties. [18:53:56] for example the property for "actors/starring" (whatever it's called in ENglish) has only one statement, so I go to add many more [18:54:12] so I click "+ add" below it [18:54:31] but when I do the text box that appears isn't focused, so i have to click it before starting to type [18:54:48] it's a tiny little thing, but as y'all know, it's the tiny little things that can be annoying in the long run [18:55:29] so yeah, no, i'm wondering how i should put that in a phabbug in a way that's easy to explain [18:56:46] Jhs: the explanation you just gave here would work fine on phab. As a title, I suggest "Input box not focused when adding statements". [18:57:04] Jhs: you should give all details, e.g. the data type of the proeprty of the statement [18:59:59] Are there plans to fix "Cannot find site jenkins_u0_mw in sites table" ? [19:01:48] DanielK_WMDE, okay, thanks. i always suck at making up titles for phab, so that's precisely what i needed ;) [19:02:04] Krinkle: https://phabricator.wikimedia.org/T126596 [19:02:27] Krinkle: is that what you are looking for? [19:02:46] Yes, it sometimes breaks Jenkins because it corrupts javascript respones with php errors [19:03:01] in some of the branches we reaches 0 errors so decided to fail on new errors [19:03:06] but then wikibase happened [19:03:11] Krinkle: please mention that in phab, we may bump the prio [19:03:16] which somehow triggers it in older branches [19:03:32] I'd rather revert all commits since it was introduced ;-) [19:03:48] since what was introduced? [19:04:12] The code that triggers the warning [19:04:37] Krinkle: i don't think that's new at all [19:04:46] It wasn't there on every commit 3-4 weeks ago [19:04:56] Otherwise I'd have been there to revert it much sooner [19:05:20] Krinkle: did you test every commit against wikibase client before that? [19:05:27] but since we can't fail no error because of these errors being unfixed for so long, it also causes other code to get worse with errors not being caught [19:05:41] DanielK_WMDE: No, but I know that our test wasn't reporting this error before 3-4 weeks ago. [19:06:02] Krinkle: the code that triggers the warnign is from 2013 [19:06:24] So either the code is new, or it wasn't tested before. Either way, not my problem. [19:06:25] i don't know why it surfaces now. [19:06:36] maybe the wiki names used by jenkis have changed [19:06:51] maybe it used to run as enwiki or something we have in the test data for the sites table [19:06:52] who knows [19:06:56] Once this error is fixed, Jenkins will fail on new errors. So this won't happen again (TM) [19:06:57] something obviouusly changed [19:08:01] Either the error doesn't matter and it shouldn't be emitted. Or it's tolerating something and likely not really testing something (which makes the test useless). [19:08:36] If this isn't fixed before next week I'm considering removing Wikibase from the all-group extensions test job. That way it'll only fail on your own commits and we can improve other prod code. [19:08:47] (it won't fail hard, it'll still remain tolerated like now) [19:11:36] Krinkle: interestingly, I do not see this failure in the wikibase repo. i wonder why. [19:13:10] DanielK_WMDE: https://integration.wikimedia.org/ci/job/mwext-mw-selenium-composer/2033/ [19:13:11] https://integration.wikimedia.org/ci/job/mwext-mw-selenium-composer/2033/artifact/log/mw-error.log/*view*/ [19:13:16] From https://gerrit.wikimedia.org/r/#/c/274766/ [19:13:20] 2016-03-03 18:31:07 integration-slave-trusty-1006 jenkins_u1_mw: [e812b46f] /jenkins-mwext-mw-selenium-composer-2033/index.php/Special:BlankPage ErrorException from line 300 of /mnt/jenkins-workspace/workspace/mwext-mw-selenium-composer/src/includes/debug/MWDebug.php: PHP Notice: Cannot find site jenkins_u1_mw in sites table [Called from [19:13:20] Wikibase\Client\WikibaseClient::newSiteGroup in /mnt/jenkins-workspace/workspace/mwext-mw-selenium-composer/src/extensions/Wikibase/client/includes/WikibaseClient.php at line 611] [19:17:55] Krinkle: oh, we are talking about selenium runs? [19:18:03] DanielK_WMDE: No [19:18:19] DanielK_WMDE: It's triggered on every change that has a job that involves Wikibase and an http request to MediaWii [19:18:29] This is just an example [19:18:34] first one I found [19:18:45] right [19:18:52] ok, so this is fixible with a config change [19:19:59] Krinkle: so this isn't triggered during phpunit runs - which would fail. [19:20:08] let me see if i can hack up a quick fix [19:20:27] DanielK_WMDE: Not during an actual test, since this code is probably not tested (perhaps should be) [19:20:34] but it is triggered before phpunit starts sometimes [19:21:04] It's primarily bugging -qunit jobs [19:21:11] Which go over http to the special page [19:21:15] and also to load.php [19:22:57] Krinkle: i'm trying to figure out where the wikibase config lives that is being used during these test runs. [19:24:02] DanielK_WMDE: Not 100% sure, btu might be the one loaded via isCI here https://github.com/wikimedia/mediawiki-extensions-Wikibase/blob/master/Wikibase.php [19:24:08] Which will be set even for non-phpunit runs [19:24:57] Krinkle: that's what I thought, but it's not consistent with what i see in the logs. [19:25:13] but anyway - i can give you a patch that *should* fix it, and you can just try it. [19:34:17] Krinkle: my current believe is that isCI is false in Wikibase.php, so the exampel settings don't get loaded at all, which causes the warning [19:34:40] Krinkle: i can make patches of stuff to try, but i have no idea how or where to try them [19:34:54] https://integration.wikimedia.org/ci/job/mwext-mw-selenium-composer/2033/artifact/log/LocalSettings.php/*view*/ [19:34:58] $wgWikimediaJenkinsCI = true; [19:35:13] It's only ever set to true [19:35:16] https://github.com/search?q=wgWikimediaJenkinsCI+@wikimedia&type=Code&ref=searchresults [19:37:21] Krinkle: yea, that's why i find this very strange. but jan now told me how to test. i'll try to come up with something [19:37:42] thanks [19:53:31] Krinkle: https://phabricator.wikimedia.org/T128779 [20:29:11] Thiemo_WMDE1: https://phabricator.wikimedia.org/T128779 [20:32:01] Krinkle: still not sure how config is loaded for the different jenkins jobs. I'm about to go home. I'll be working half day tomorrow, but i'll probably be online in the late (EU) evening. Pink jzerebecki about this, he knows all teh jenkins stuff much better than me. [20:32:40] it's still unclear to me which change triggered this problem. it seems like an entire chain of slightly incompatible chances is at the root of this. [20:32:59] Yeah, no worries [20:33:00] Thanks! [20:33:03] I'l check it out tomorrow [20:33:27] Hm, you should probably ping jzerebecki, not pink him... [20:34:35] DanielK_WMDE: What does it need from example settings? [20:34:52] DanielK_WMDE: With plain mediawiki installed on a plain server with Wikibase and no configuration changes, it shouldn't emit this warning, right? [20:35:07] Krinkle: I just added siteGroup there, that should fix the issue. [20:35:12] but my change had no effect [20:35:21] also, logs indicate that other exampel settings don't take effect either [20:35:28] Any needs should either be provided by an sql update in the installer or at run time via an extension hook to provide any default settings based on other settings (e.g. put something in an array that point sto local or somehing) [20:35:43] for example, the siteGlobalID from exampel si mywiki, not jenkins_bla_foo. [20:35:50] that'S the default (fallback to db name) [20:36:42] Krinkle: yea, that's how it works. we can add it to the database, and we probably should. but we should still control the config, right? [20:52:33] hm, jenkins being super slow doesn't help with figuring this out [21:47:41] https://www.wikidata.org/w/index.php?title=Q3816775&type=revision&diff=309979540&oldid=305169980 <- yah! Bot reverts vandalism and properly references the statement in the process ;-) [21:48:44] :) [23:59:12] I got stuck on creating a query for people born on a date (as in 3rd of March). Any ideas?