[01:08:08] apergos: Wikidata XML exports: [01:08:09] {"label":{"en":"Barack Obama","fr":"Barack Obama","ar":"\u0628\u0627\u0631\u0627\u0643 \u0623\u0648\u0628\u0627\u0645\u0627","ru":"\u0411\u0430\u0440\u0430\u043a \u041e\u0431\u0430\u043c\u0430","nb":"Barack Obama","it":"Barack Obama","de":"Barack [01:08:10] Obama","be-tarask":"\u0411\u0430\u0440\u0430\u043a \u0410\u0431\u0430\u043c\u0430","nan":"Barack Obama","ca":"Barack Obama"},"description":{"en":"President of the United States of America [01:14:55] Full history for Q1-Q100.... 76.1MB [01:15:10] 7z turns that into 887KB [05:35:11] Reedy: that's pretty gross. I guess you could pass that on to the wikidata folks and see if they want to do anything about that [13:31:06] hashar: hey there [13:46:55] New patchset: Stefan.petrea; "removed debianize.sh (using the one from wmf-analytics/debianize repo" [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/31827 [13:46:55] New patchset: Stefan.petrea; "Added some code update to provoke a jenkins rebuild" [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/31828 [14:17:39] Change abandoned: Stefan.petrea; "please abandon, we used this only for debugging jenkins" [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/31827 [14:54:09] andre__: ping [14:54:16] sumanah, pong [15:00:04] effective communication 101 [15:31:18] :) [17:16:05] New patchset: Stefan.petrea; "Test review, please abandon this. Only used for debug" [analytics/libanon] (master) - https://gerrit.wikimedia.org/r/31854 [17:39:53] New patchset: Stefan.petrea; "Fixed configure.in" [analytics/libanon] (master) - https://gerrit.wikimedia.org/r/31858 [19:27:35] New patchset: Stefan.petrea; "Added script to geocode all squid .gz files" [analytics/wikistats] (master) - https://gerrit.wikimedia.org/r/31877 [19:29:27] New review: Diederik; "Ok." [analytics/wikistats] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/31877 [19:29:27] Change merged: Diederik; [analytics/wikistats] (master) - https://gerrit.wikimedia.org/r/31877 [19:35:39] New patchset: Stefan.petrea; "Fixed exclude, removed debianize.sh" [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/31880 [19:36:21] New review: Diederik; "Ok." [analytics/udp-filters] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/31880 [19:57:47] New review: Stefan.petrea; "Ok." [analytics/udp-filters] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/31880 [19:57:47] Change merged: Stefan.petrea; [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/31880 [20:12:42] New patchset: Stefan.petrea; "Added git sumodule debianize" [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/31884 [20:16:28] New review: Diederik; "Ok." [analytics/udp-filters] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/31884 [20:16:28] Change merged: Diederik; [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/31884 [20:49:23] * werdna waves [20:49:28] Reedy: about? [20:49:29] tewwy: about? [20:49:32] Yeaah [20:49:36] how's it going? [20:50:14] Not brilliant. Nosebleeds, headaches, my desktop is being a PITA and my grandmother is rather ill [20:50:20] That sucks. [20:50:32] werdna: You've been out for a bit, haven't you? [20:50:44] marktraceur: I've been in the background [20:50:44] I'm always around [20:50:47] LURKING [20:50:47] just rarely on IRC [20:50:59] Waiting....for the perfect moment to pounce.... [20:50:59] but occasionally visible on bugzilla [20:51:14] heh [20:51:16] rarely [20:51:18] more often gerrit [20:58:27] Reedy: will I be seeing you on Saturday? [20:58:40] Yup [20:59:24] hashar: That's a point [20:59:24] You haven't added yourself to the hackathon attendee list [20:59:26] Reedy: good stuff. [20:59:53] Reedy: I will probably arrive around 12–13. I have a party the previous night that will probably go deep into the morning :/ [21:00:23] Are you using a dutch mobile number now? [21:00:23] yes [21:00:28] +31 6 41 37 28 70 [21:01:08] That's 3 different numbers for you now ;) [21:01:24] I have quite a few [21:01:37] Aussie, Dutch, UK, US, German (not sure it still works) [21:01:37] Yeah, I don't have your US one ;) [21:02:50] :p [21:02:50] I have his US and AU number, can give you those if you want [21:02:58] heh, I also have my US and AU numbers [21:02:59] * RoanKattouw puts the NL number in his phone [21:03:04] D'oh, of course you do :) [21:03:06] :p [21:03:13] Although really I should transfer it to my NL phone as well [21:03:33] My favourite Dutch phrase is a post-halloween poster I saw in Amsterdam [21:03:37] "Wie heeft mijn tomahawk gevonden" [21:03:49] lol [21:04:03] Who has found my tomahawk? [21:04:03] lol [21:05:03] And my favourite dutch word so far is "vlagophangers" – as in "bedankt de bestuur van uros (a student association in maastricht) en zijn vlagophangers. Wij heeft nu een mooie tijdelijk tafelkeeft." [21:05:18] only in dutch would somebody use a word that means "people who hung up a flag" [21:08:02] Yeah we construct words like taht [21:09:04] BTW are you quoting that or paraphrasing it? [21:09:58] Just wondering if people in Limburg really talk/write like that :) [21:11:05] werdna: German has some pretty amazing interfaces for squashing words together, too [21:11:25] RoanKattouw: paraphrasing, not really clear on the exact phrasing [21:11:31] I can look it up :p [21:11:31] I don't think I've ever asked a German speaker for a single word to describe something and been turned away. [21:11:46] Of course, they wind up being 30 or 40 characters. [21:12:00] werdna: Oh good. It's not bad at all for someone who's just been in the country for a few months, but if that was your learning /input/ you'd have me a bit worried [21:12:01] Bedankt Bestuur van Uros & zijn vlagophangers, wij hebben tijdelijk een heel mooi tafelkleed! [21:12:03] there we goo [21:12:24] Yeah that looks reasonable [21:12:45] wij hebben, tafelkleed, mooi, tijdelijk is an adverb not an adjective, a few other bits and pieces [21:13:04] She's from Brabant apparently [21:13:13] Knegsel [21:13:20] tijdelijk can be either an adverb or an adjective, actually, your usage was valid but means something subtly different [21:13:28] ahh [21:13:42] * RoanKattouw doesn't know what 'knegsel' means, some Southern thing? [21:13:44] it's a place [21:13:48] in Noord-Brabant [21:13:59] Right, OK [21:14:12] * RoanKattouw goes to look up where that is [21:14:56] I've only just learned how to form subordinate clauses from sentences – i.e. "ik denk *dat* hij wil een kopje koffie maakt" [21:14:59] Hah I've actually been on vacation in that area, near Eersel' [21:15:20] "Ik denk dat hij een kopje koffie wil maken" [21:15:21] RoanKattouw: the Dutch here make fun of people from Freisland just as much as they make fun of the Limburgs [21:15:27] Oh I'm sure they do [21:15:29] so no escaping! [21:15:57] RoanKattouw: ahh I got confused with the three verbs. [21:16:05] In that sentence, a true Frisian would have said "maken wil" at the end [21:16:27] forgot that wil is a verb :p [21:16:34] Yeah you can only have one conjugated verb per subordinate clause, there are three verbs so there has to be an infinitive somewhere [21:16:48] right, I forgot about the 'wil' [21:17:12] Also, confusingly, it's SOV rather than SVO in the subordinate clause [21:17:33] But whether that actually happens depends on the word you use to connect the clause ("dat" in this case) [21:17:41] Welcome to my language :D [21:18:54] fixed word order, meh :-P [21:18:54] :-D [21:18:54] It's referred to as "V2" [21:18:58] verb second [21:19:09] Ah Ok [21:19:09] http://en.wikipedia.org/wiki/Verb-second [21:19:24] So let me leave you with what I think is one of the more ridiculous examples [21:19:27] "want" and "omdat" both mean "because" [21:19:39] But the former doesn't trigger V2 while the latter does [21:20:23] So "Ik liep weg omdat ik een kopje koffie wilde zetten" and "Ik liep weg want ik wilde een kopje koffie zetten" mean the same thing, but have different word orders [21:20:23] oh god [21:20:32] ik wilde? [21:20:38] werdna: Wishing you'd stayed? ;-) [21:20:38] Past version of "wil" [21:20:54] I usually use "wou" but technically that's not correct [21:20:56] "singular past indicative and subjunctive of willen." says wiktionary [21:21:10] I don't think I've touched past tense yet [21:21:13] Past is usually +de or +te, except for irregular verbs [21:21:25] I think I pretty much always use the auxiliary form [21:21:25] ("Liep" is the past version of "Loop", that one is irregular) [21:21:32] Auxiliary? [21:21:41] ik heb een kopje koffie gedrankt [21:21:46] or so [21:21:49] Right [21:22:01] whatever you call the one without a progressive aspect [21:22:06] non-progressive? [21:22:12] Present perfect in English, right? [21:22:12] the one with a past participle :p [21:22:20] I believe so. [21:22:29] no wait [21:22:32] past imperfect [21:22:43] in french, passé composée [21:22:43] Yes? [21:22:58] Yes, passé composée, I was about to say that [21:23:06] werdna: "I walked / I was walking" Past perfect vs. Past imperfect. [21:23:15] I'm glad we can agree to describe the attributes of the second language in a third language :) [21:23:22] * James_F snorts. [21:23:29] yes, but in Dutch this is literally "I have walked" [21:23:44] werdna: Oh, yes, the past pluperfect. [21:23:44] Dutch doesn't have the -ing tense [21:23:52] RoanKattouw: Not at all? [21:24:01] no progressive aspect in Dutch [21:24:06] Well, it does kind of have it, but not in quite the same way as English [21:24:15] "I am walking", you just say "I walk", right? [21:24:15] RoanKattouw: How do you say "Whilst he was running through the field, a bull attacked and killed him"? [21:24:16] RoanKattouw: I was hoping to get your opinion on https://www.mediawiki.org/wiki/User:Sharihareswara_%28WMF%29/Maintainers [21:24:29] James_F: While he ran ... [21:24:29] and I presume you aren't busy right now [21:24:31] * sumanah may be wrong [21:24:34] RoanKattouw: That's a totally different meaning. :-) [21:24:44] RoanKattouw: So no imperfect past tense at all? [21:24:46] Nope [21:24:49] Shocking. [21:24:58] <^demon> $wgLanguageCode = 'simple'; [21:25:28] :)) [21:25:43] Turns out LANGUAGE IS REALLY COMPLICATED [21:25:58] Which is why I get to spend a year or two of my four studying it :) [21:26:43] <^demon> We should really get on the universal translator thing. [21:27:50] The basic problem is that in many cases the source text doesn't have enough information to make a full rendering in the target language [21:27:59] ^demon: This blew my mind when I first found out about it [21:27:59] http://en.wikipedia.org/wiki/Evidentiality [21:28:10] YES. [21:28:18] * marktraceur remembers talking about that when werdna was in town [21:28:27] :) [21:28:49] <^demon> "Some languages have a distinct grammatical category of evidentiality that is required to be expressed at all times." [21:28:56] <^demon> Pardon me while I go shoot myself. [21:29:02] right [21:29:11] Oh God yeah that [21:29:15] I heard about that [21:29:24] Dutch doesn't have *that* fortunately [21:29:26] put the epistemology in every verb [21:29:38] From the "Typology" section: visual sensory, nonvisual sensory, inferential, reported, heard from known source, direct participation (e.g., Fasu) [21:29:43] Lovely. [21:29:45] sumanah: I think the world would be a better place if we retrofitted English with this. [21:29:53] werdna: You first. [21:29:59] werdna: I too have a disproportionate love for this idea [21:30:08] but will not actually implement it when I am God-Emperor [21:30:30] not for adults, anyway [21:30:33] only for the children [21:31:02] <^demon> "Evidentiality is not considered a grammatical category in English because it is expressed in diverse ways and is always optional." [21:31:08] <^demon> :D [21:31:10] talking of languages... I want a Language object for code "mul". [21:31:38] I wonder if I can just make "mul" a valid language code, or whether it should be treated separately. [21:31:43] 'mul' = multiple? [21:31:53] I guess I should look at how qqx is implemented. [21:32:07] werdna: btw I think you heard that hoo now has +2 on AbuseFilter [21:32:11] sumanah: Aaron for DB is interesting, though I wouldn't say it's invalid without asking Aaron [21:32:12] werdna: multilingual, yes. e.g. for indicating the content language of wikidata items. [21:32:12] DanielK_WMDE_: a piece of advice: use qqm instead? :p [21:32:22] sumanah: yeah, fine by me. [21:32:26] werdna: why? "mul" is in iso 639-2 [21:32:27] werdna: anyone else to nominate? [21:32:36] I'm happy for crowd maintainership because it means less work for me! [21:32:39] it's speced for this exact purpose [21:32:43] sumanah: I assume Chris Steipp is already that [21:32:53] DanielK_WMDE_: then ignore me. [21:33:08] werdna: but only this once ;) [21:33:09] I am clearly talking nonsense about things I don't understand :) [21:33:10] werdna: Chris Steipp has +2 in core and thus in extensions as well, yes [21:33:21] good-o [21:33:23] werdna: but hoo himself will need to get reviewed for a patchset or 2 of his - https://gerrit.wikimedia.org/r/#/c/30110/ is an example [21:33:36] yeah, I am well behind on reviews and so on [21:33:51] I think Krenair_ has a LQT changeset or 5 waiting for you, yeah [21:34:07] oh just 1! [21:34:11] https://gerrit.wikimedia.org/r/#/q/status:open+project:mediawiki/extensions/LiquidThreads,n,z [21:34:27] there are about 5 waiting that have gotten +1s or no marks [21:35:08] werdna: what can I do to help you be less behind? do you think maybe we could nominate someone for +2 on LQT? Alex maybe? [21:35:42] Let's see how I go over the coming week or so. The last month or two has been disrupted. First I had exams, then I broke my laptop and only got it back a few days ago. [21:36:00] So if I'm still behind by the end of the week maybe we can figure out a solution but otherwise I think I'll be able to catch up [21:36:37] werdna: okay. but even if you do catch up in the next week, there will be future disruptions too, so a sustainable course of action would end up with multiple maintainers for every bit of codebase you currently maintian [21:36:39] maintain* [21:36:42] AF, LQT, etc [21:36:56] nod. I don't know who it would be though. [21:37:41] werdna: it doesn't have to be Now. We can train people up [21:38:06] Nod. [21:38:11] we can ask a volunteer, for example, whether they'd be interested in setting a goal of being a comaintainer of foo in n months [21:38:36] sumanah: For installer, might I suggest Chad? Langconv for Siebrand I'd ask him. For logging, Niklas? Maintenance scripts, add Chad maybe? RL, add Timo & Trevor. Revdel, ask Aaron. Unit tests, possibly add Chad. CentralAuth, Chris S? ExtensionDistributor, Tim? Gadgets, me&Timo. GlobalUsage, BryanTM, but he's inactive. MobileFrontend, mobile team. OpenSearchXml, Brion? PagedTiffHandler,... [21:38:36] ...Kinzler? ParserFunctions, Tim? PoolCounter, Tim and/or Asher? RSS, Mark Hershberger? SecurePoll, Tim? WikiEditor, me&Trevor. WikimediaMaintenance, Sam? WikimediaMessages, i18n team / multichill? [21:38:50] RoanKattouw: I say go ahead and add them, unless you'd rather I do so [21:39:05] (the ones who are active) [21:39:35] <^demon> Somehow I got saddled with maintenance scripts, even though I didn't write almost any of them. [21:39:35] sumanah: Well >half of those suggestions were "ask X whether it's appropriate", and I want to actually get some code written today, so if you wouldn't mind .... [21:39:50] <^demon> Unless it's a generic issue affecting all maintenance scripts, in which case I'm responsible. [21:39:53] RoanKattouw: I shall, worry not. [21:40:05] Thanks [21:40:28] * sumanah hits edit [21:40:48] RoanKattouw: you mean "RevDel, ask chris?" [21:41:22] AaronSchulz is washing his hands :p [21:41:41] Sure whatever :) [21:42:07] AaronSchulz: Chris Steipp? [21:50:28] yes [21:53:48] jobqueue question - AaronSchulz is this a very high priority? https://bugzilla.wikimedia.org/show_bug.cgi?id=41656 [22:05:02] RoanKattouw, where did you see a poolcounter commit? [22:05:44] I didn't [22:05:59] I was responding to sumanah RoanKattouw: I was hoping to get your opinion on https://www.mediawiki.org/wiki/User:Sharihareswara_%28WMF%29/Maintainers [22:06:27] ah, I thought it was for assigning bugs [22:06:42] I think so? I don't know [22:06:46] I hadn't seen the question and made it up in my mind :P [22:06:48] I suppose we should add your name to PoolCounter, right? [22:06:57] (Or not, if you don't want to) [22:07:12] probably, given that I wrote the C daemon [22:07:42] go ahead Platonides [22:07:49] why does that page link to github but not to gerrit? [22:07:57] add yourself, if you're ok with dealing with questions & review requests :-) [22:08:15] Platonides: because it was late and when I was coding the template I could easily construct 1 but not the other [22:08:20] feel free to change the template [22:08:57] I remember I had a reason at the time, but it may be entirely bogus to someone who's actually slept recently [22:09:43] note that it's 11pm here [22:10:03] it may also be bogus in general [22:13:29] I added a couple of links [22:13:32] nod [22:13:34] they apparently work fince [22:13:36] *fine [22:14:05] Yay! thanks [22:15:38] :) [22:57:41] robla: https://bugzilla.wikimedia.org/show_bug.cgi?id=41656 - looks like an urgent JobQueue problem [22:58:57] I'll look in a bit. doesn't look like it's a problem on the cluster (yet), just translatewiki.net [22:59:41] okay [22:59:42] i think it might be a problem on the cluster [23:00:51] AaronSchulz: it looks like refreshLinks2, etc jobs are broken again [23:02:41] hrm, they're getting run, although there are 20 hour old jobs in the enwiki queue [23:03:42] binasher: would it make sense for me to add you to the "jobqueue" part of https://www.mediawiki.org/wiki/User:Sharihareswara_%28WMF%29/Maintainers ? [23:04:02] It looks like its dropped off again... [23:04:43] sumanah: no, i'm currently acting as would any generic ops engineer [23:05:02] okay [23:06:11] Ugh [23:06:12] Number of entries to runJobs is tiny [23:06:41] hasn't gotten an update in the last hour either [23:06:41] http://ganglia.wikimedia.org/latest/graph_all_periods.php?c=Miscellaneous%20pmtpa&h=spence.wikimedia.org&v=40548&m=enwiki_JobQueue_length&r=hour&z=small&jr=&js=&st=1352156786&z=large [23:06:45] thank you kaldari [23:06:45] What changed at 21:50? [23:06:57] NP [23:07:32] http://ganglia.wikimedia.org/latest/?r=hour&cs=&ce=&s=by+name&c=Jobrunners%2520pmtpa&tab=m&vn= [23:07:41] They look relatively busy [23:09:01] Reedy: load dropped around 22:00 [23:09:04] Fatal error: wikiversions.cdb has no version entry for ` [23:09:04] Fatal error: Class 'Memcached' not found in /usr/local/apache/common-local/php-1.21wmf3/includes/objectcache/MemcachedPeclBagOStuff.php on line 57`. [23:09:19] More apaches missing hte module [23:09:27] nope [23:09:44] thats on mw6 when trying to start jobs-loop.sh [23:09:45] but [23:09:50] root@mw6:~# php -m | grep memc [23:09:50] memcached [23:09:58] That's no use [23:10:13] oh, ignore th at [23:10:16] Is it running in apache though? [23:10:19] no [23:10:28] the job runners don't run apache [23:10:36] not anymore ;) [23:12:27] ++ php -n MWScript.php nextJobDB.php --wiki=aawiki --type=MoodBarHTMLMailerJob [23:12:27] + db=' [23:12:27] Fatal error: Class '\''Memcached'\'' not found in /usr/local/apache/common-local/php-1.21wmf3/includes/objectcache/MemcachedPeclBagOStuff.php on line 57' [23:12:57] yeah, I noticed that wtf piecing together the log entry [23:12:57] php -i shows memcached loaded and configured.. what else could cause that? [23:13:21] also seeing Fatal error: wikiversions.cdb has no version entry for ` [23:13:28] * robla reads backlog [23:15:06] php -r "var_dump( class_exists( Memcached ) );" [23:15:07] bool(true) [23:15:28] reedy@mw6:/usr/local/apache/common$ php multiversion/MWScript.php runJobs.php enwiki [23:15:28] 2012-11-05 23:15:04 htmlCacheUpdate Wikipedia_talk:WikiProject_Video_games/to_do table=templatelinks start=13554131 end=14544256 STARTING [23:15:28] 2012-11-05 23:15:09 htmlCacheUpdate Wikipedia_talk:WikiProject_Video_games/to_do table=templatelinks start=13554131 end=14544256 t=5236 good [23:15:44] oh! [23:16:08] /usr/local/bin/jobs-loop.sh calls "php -n" [23:16:22] wtf [23:16:29] =ignore php.ini [23:16:29] lol [23:16:30] duh, I missed that too [23:16:46] how is jobs-loop.sh deployed? [23:17:16] via scap, it's in a MW extension [23:17:19] WikimediaMaintenance [23:17:50] it seems to be in puppet though too [23:18:01] ... [23:18:16] Reedy: are we using the one in puppet then? [23:18:31] nfi [23:18:38] i think the puppet version is what's on the newer non-apache job runners? notpeter would know [23:19:06] probably [23:19:07] modules\mediawiki_new\templates\jobrunner\jobs-loop.sh.erb [23:19:46] anyone know if there was good reasoning for the -n at some point? [23:20:47] OK, I'm gonna leave it to you folks, thx for hashing it out [23:21:01] sumanah: this isn't related to that bug [23:21:17] oh. [23:21:53] orly? [23:22:02] binasher: http://svn.wikimedia.org/viewvc/mediawiki/trunk/tools/jobs-loop/jobs-loop.sh?r1=14781&r2=15611 [23:22:08] blame tim? ;) [23:22:53] 2006 tim! [23:23:06] ok, i'm going to merge https://gerrit.wikimedia.org/r/#/c/31968/1 [23:24:03] notpeter: ppl were wondering if the jobs-loop.sh in puppet is the only one thats really in use [23:24:25] should be [23:24:37] AaronSchulz says there's a version that's installed via scap too [23:24:52] it's probably unused then [23:24:54] * AaronSchulz will nuke that [23:25:44] robla: I need to go to dinner. can I ask you to shepherd https://bugzilla.wikimedia.org/show_bug.cgi?id=41656 to some kind of .... proto-pre-resolution? [23:26:03] binasher: when you say "this isn't related to that bug", do you mean a) you're not working out why the JobQueue isn't running jobs on the Wikimedia cluster, or b) you're not working out the general MediaWiki problem that's hitting translatewiki.net [23:26:11] sumanah: yup, I'm on it [23:26:12] hashar: We may need some help on https://gerrit.wikimedia.org/r/25838 [23:26:30] sumanah: go eat :) [23:26:45] :) [23:26:55] robla: i just fixed the issue with the jobqueue on the wmf cluster [23:27:10] hashar: Not that I think you should be working at this hour, but throw it into a TODO list if you don't mind [23:27:10] just confirmed. the one in the mediawiki_new module is the only in play on the cluster [23:27:11] that ticket predates our job queue breaking [23:27:41] the translatewiki issue sounds like the problem AaronSchulz fixed last week [23:28:42] notpeter: thanks! [23:29:12] ah, ok, got it. yeah, I was admittedly more concerned about our job queue too :) [23:31:44] hmm, so far just enotifNotify jobs are running now that the job queue is alive again [23:32:08] that doesn't bode well [23:33:17] which log do php errors on job runners go to? [23:33:46] marktraceur: will look at it tomorrow hopefully. [23:34:05] binasher: that was fixed on twn as well by the same fixes, but broke again it seems [23:34:15] robla: i don't think they're handled any differently [23:34:15] binasher: heh, look at tim's mw core summary for ec277ae [23:34:45] binasher: I'm pretty sure the last time we did this, we found the errors in some log that we hadn't been looking at [23:34:53] hashar: Great! sumana will be pleased. [23:35:04] Perhaps Harry will be more pleased. [23:35:46] * robla pulls up the #wikimedia-operations log from Nov 1 [23:36:08] robla: was that discovered by one of Reedy or AaronSchulz? [23:36:22] actually, I think apergos found it [23:36:39] * robla can't remember for sure [23:40:38] marktraceur: i think it is related to sqlite. [23:41:22] hashar: Really? The database fails to update, or...? [23:41:38] hashar: Because "the wiki doesn't allow SVG uploads" seems like a reasonable explanation to me [23:42:19] binasher: found it. it's not a separate log, it's a way of running the job runner [23:42:34] one could hmm set -x on this bash script, run it with output to a file for a bit [23:43:25] see http://bots.wmflabs.org/~petrb/logs/%23wikimedia-tech/20121031.txt for the full conversation, starting at [21:34:28] [23:48:23] marktraceur: that might be the reason. SVG upload should be enforced by the test. [23:48:39] marktraceur: too late for now, will have a look at it tomorrow; -) [23:48:41] Yes, do that [23:48:44] Sleep well, hashar! [23:49:01] so apergos was pointing out that the runner script only runs enotifNotify jobs so long as any wiki has any of those in its queue [23:50:53] sounds like tim's comment in ec277ae [23:52:09] AaronSchulz: where is that? pasting into gerrit's search box gives nada [23:52:29] try the github mirror! :) [23:53:45] grr, there was another "php -n" that i missed [23:53:58] ahh, right there are two indeed [23:57:37] ok, that fixed non-priority jobs [23:57:49] \o/ [23:58:02] so bug 41656 is still something different [23:59:15] !b 41656 [23:59:15] https://bugzilla.wikimedia.org/show_bug.cgi?id=41656