[00:00:25] The RefreshLinks job does Linksupdate which does all the link tables and page properties. it is the same class and function that the refreshLinks.php maintenance script calls. Apparrently one does something the other doesn't and I dont' see it. [00:00:44] I know why but I don't see how. [00:01:38] Krinkle, someone has asked me to add a new parameter to jsMessage so VisualEditor's post edit notice can be hidden by ID/class... I told them it's deprecated and mw.notify should be used instead. [00:01:38] Could the VE team change VE to use mw.notify with a tag option so that it has a class? [00:02:06] Krenair: We've already written that code; it's waiting on a core change for us to merge. [00:02:31] Krenair: ... which I would link to if gerrit wasn't down. :-( [00:02:41] Yeah I was about to ask :) [00:02:47] James_F: It's up again [00:02:53] and even seems to work in Firefox \o/ [00:03:15] Seems to be hanging for me [00:03:30] <^d> Caches are cold, give it a minute. [00:03:32] maybe it's overloaded after the reboot [00:03:34] yeah [00:03:36] good night [00:03:42] works now :) [00:04:28] Krenair: https://gerrit.wikimedia.org/r/69856 but again gerrit seems down. [00:04:52] <^d> Should be back now? [00:06:07] ^d: Yeah; thanks. :-) [00:06:37] ok, thanks James_F [00:09:50] James_F: Krenair: One instance of jsMessage has already been updated to mw.notify (as can be seen on https://gerrit.wikimedia.org/r/#/c/69856/1/modules/ve/init/mw/targets/ve.init.mw.ViewPageTarget.js ) the other can be updated as well without the postedit change having to go first. [00:10:15] also, that change nor the one in core gives it a dedicated tag, which is needed for it to have a togglable class [00:12:52] It will have .postedit though, which seems to fit what the user wants (already has .postedit { display: none; }, wants VE postedit notifications to go away as well) [00:16:00] k [00:16:12] New patchset: Ryan Lane; "Add api action for getting console output" [mediawiki/extensions/OpenStackManager] (master) - https://gerrit.wikimedia.org/r/60783 [00:16:50] \o/ [00:17:00] I think I'm actually getting the hang of promise/deferred [00:17:16] and I found a reasonable doc on it :) [00:20:24] Yeah promise/deferred is nice :) [00:21:23] Ryan_Lane: I also did an attempt to explain it during the FrontendMasterClass this january at wmf office. [00:21:43] I'll need to make sure to attend JS classes in the future [00:21:54] We recorded this one [00:21:55] at the time I didn't think I'd be using JS [00:21:57] * Krinkle finds link [00:21:57] ah [00:21:58] cool [00:22:16] http://www.youtube.com/watch?v=2K5PtLTkwhs#t=51m36s [00:22:22] bleh. looks like I rebased this changeset again [00:22:53] New review: EBernhardson; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71740 [00:23:08] in case anyone wants to code review this (https://gerrit.wikimedia.org/r/#/c/60783/20) needs to be checked [00:23:47] and it's a fairly minor change, just following up on a few comments [00:33:11] New patchset: Bsitu; "Mime emails should be sent with UTF-8 charset" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71747 [00:34:17] New patchset: Cmcmahon; "WIP add ve_refs check" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/69128 [00:48:08] New review: Krinkle; "It looks like it is currently doing this:" [mediawiki/extensions/Math] (master) C: -1; - https://gerrit.wikimedia.org/r/69298 [01:12:05] New patchset: Katie Horn; "Starting the Great Form Purge of O-13 Phase 1: Basic cruft removal" [mediawiki/extensions/DonationInterface] (master) - https://gerrit.wikimedia.org/r/71750 [01:14:18] New patchset: Hiong3-eng5; "moved most SpecialPage from app to setup" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/71576 [01:18:45] it sounds more like you want to isolate some particular bug, rather than that you want to know how LinksUpdate works [01:19:40] maintenance/refreshLinks.php is never run on WMF, and we have working categories [01:20:03] so obviously it's not quite as you describe [01:20:40] What did I miss? [01:22:26] New patchset: Mwalker; "Always Display Emergency Priority Banners" [mediawiki/extensions/CentralNotice] (master) - https://gerrit.wikimedia.org/r/71751 [01:22:42] the behaviour you describe in comment 16 sounds completely normal [01:24:10] 10 minutes is not a long enough wait to ensure that everything in the job queue was executed, and you didn't confirm that the job in question was in fact executed [01:26:43] New review: Mwalker; "yay! I'm a bit worried that there might be a hidden CSS dependency -- but this will root it out if t..." [mediawiki/extensions/DonationInterface] (master) C: 2; - https://gerrit.wikimedia.org/r/71750 [01:28:21] 2013-07-02 05:57:55 mw1015 enwiki: refreshLinks Template:Redirect masterPos=db1056-bin.000205/872529929 rootJobSignature=5910fa49190166d8f186f2301298b82e6c001fd8 rootJobTimestamp=20130620135710 t=3800 good [01:29:00] the edit was at 00:33, so it took 5.5 hours [01:29:21] Thanks [01:30:16] as for action=purge doing links updates [01:30:22] there are two ways to purge the page [01:30:26] TimStarling: Can you confirm it wasn't a job caused by a (null) edit to the page directly? [01:30:36] they only differ in whether or not they do links updates [01:30:42] Sure, I understnad. [01:30:46] action=purge doesn't, null edit does [01:30:51] If it just takes a few hours then we don't have a problem. [01:31:05] other than our job queue being a bit slow, but that's not my problem directly. [01:31:23] if the page was null edited, the update would have been done immediately, rather than queued [01:33:14] New review: Parent5446; "(1 comment)" [mediawiki/core] (master) C: 1; - https://gerrit.wikimedia.org/r/71740 [01:35:12] I think the rootJobTimestamp indicates that the job was in response to an edit on June 20 [01:48:37] New review: Mattflaschen; "Not tested, but I understand the issue from using the old version. It will be nice to easily distin..." [mediawiki/vagrant] (master) C: 1; - https://gerrit.wikimedia.org/r/71659 [01:51:30] TimStarling: an edit to a template then, yes [01:52:32] TimStarling: So you're saying the page properties that templatedata adds when is used on a page will end up on Template:Foo when added to Template:Foo/doc, it just didnt' happen yet? [01:52:48] New review: Mattflaschen; "There are definitely parts of the MathJax that are confusing, indirect, and don't meet MW standards...." [mediawiki/extensions/Math] (master) - https://gerrit.wikimedia.org/r/69298 [01:52:59] yes [01:53:36] have we got a job queue length graph somewhere? [01:54:08] I guess the "jobs" property siteinfo gives in the API is not accurate then :P [01:54:28] http://ganglia.wikimedia.org/latest/?c=Miscellaneous%20pmtpa&h=hume.wikimedia.org&m=cpu_report&r=hour&s=descending&hc=4&mc=2#metric_Global_JobQueue_length [01:54:48] Krinkle: that doesn't check redis [01:55:03] self::$jobs = $dbr->estimateRowCount( 'job' ); [01:55:11] ha [01:55:15] I noticed that weeks ago [01:55:27] it relies on the implemention being sql backed [01:55:33] that explains [01:55:55] I wanted to make some performance tweaks for batch queue size detection first...too many things on the TODO list [01:56:09] * Aaron|home has been doing nothing but oauth the last few days [01:56:11] TimStarling: Could you reply on https://bugzilla.wikimedia.org/show_bug.cgi?id=50372 with what you told me? (That it should be working as far as you know, the job queue just took a few hours). [01:56:50] well, a couple of weeks actually [01:59:26] Krinkle: 20 June -> 2 July != "a few hours". [02:00:09] I was citing TimStarling from "5.5 hours". [02:00:43] Aha. [02:00:44] yeah, it happened to execute 5.5 hours after your edit, but it was responding to an edit made 2 weeks previously, not to your edit [02:00:46] Do we have a confirmed sighting that took even longer? Which one was edited on june 20 and only visible until today? [02:00:52] Aha [02:01:06] Which is not good. [02:01:25] : I think the rootJobTimestamp indicates that the job was in response to an edit on June 20 [02:01:27] Missed that one [02:03:16] Do we have statistics on what kind of jobs the job are in the queue (e.g. which type has the most ones) and/or which are taking the most processing time? [02:03:52] refreshLinks takes the most time [02:03:54] It taking several weeks sounds pretty bad. Any possible solutions ops is looking at? [02:04:00] or do you mean what sort of refreshLinks? [02:04:15] I'm not sure, I just looked into that logic for the first time today :) [02:04:48] (I know how it worked roughly, but not how it was called from the ParserOutput) [02:06:38] Krinkle: So apparently for Parsoid Aaron built them their own queue. [02:06:58] $wgJobTypesExcludedFromDefaultQueue[] = 'ParsoidCacheUpdateJob'; [02:07:05] yeah, I spotted that earlier. Fancy :) [02:12:53] queue lengths: http://paste.tstarling.com/p/AMtkHB.html [02:16:00] New review: Ori.livneh; "This works, with the caveat that the PHP call to add the schema module to the page output precedes t..." [mediawiki/extensions/UniversalLanguageSelector] (master) C: 1; - https://gerrit.wikimedia.org/r/68862 [02:16:51] better method, same results: http://paste.tstarling.com/p/idhZKj.html [12:48:13] <^demon|away> Opera has never worked well with Gerrit. [12:48:17] yeah, but it worked better than right now. [12:48:37] <^demon|away> qchris: We should update repo-discuss with what we know too, I fear we need someone who Knows More Than Us to really solve it...I'm fearing some GWT bug. [12:49:11] ^demon|away: Yes, I fully agree. [12:49:21] ^demon|away: However, let us try the new build first. Maybe there are some users who still experience the problem nonetheless :-/ [12:49:32] <^demon|away> Yes, new build is the right first step. [12:49:46] <^demon|away> And if it fixes it for people, it's an acceptable workaround for now pending upstream actually caring. [12:57:46] yurik: you there? [12:58:17] yurik: am i understanding https://gerrit.wikimedia.org/r/#/c/71759/ right? you want people to use URLs like api.php/action=query&prop=whatever? with a slash? [12:58:28] isn't that a really bad idea? [13:18:16] Change merged: jenkins-bot; [mediawiki/extensions/DumpHTML] (master) - https://gerrit.wikimedia.org/r/60194 [13:23:46] New patchset: Hashar; "pupptest-test-runner a wrapper to run puppet tests" [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/71789 [13:24:37] Change merged: Hashar; [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/71789 [13:33:16] New patchset: Hashar; "puppet-test-runner was missing an exit call" [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/71791 [13:33:44] Change merged: Hashar; [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/71791 [13:34:21] New patchset: Hashar; "operations-puppet-test" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/71792 [13:34:50] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/71792 [13:41:11] New patchset: Matmarex; "Use background-image: none instead of background: none" [mediawiki/extensions/Vector] (master) - https://gerrit.wikimedia.org/r/71795 [13:47:43] New patchset: Aude; "Update Wikibase with code for propagateChangesToRepo setting" [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71797 [13:48:42] New patchset: Hashar; "tweaks for puppet-test-runner" [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/71798 [13:49:02] Change merged: Hashar; [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/71798 [13:56:51] New review: Anomie; "You've got two separate changes munged together here: the change to specify the action via pathinfo,..." [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/71759 [14:00:39] Change merged: jenkins-bot; [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71797 [14:02:04] New patchset: Hashar; "puppet-test-runner report success" [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/71799 [14:02:28] Change merged: Hashar; [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/71799 [14:02:43] New patchset: Anomie; "Add getPersonalToolsList to SkinTemplate" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71800 [14:03:11] New patchset: Anomie; "Add getPersonalToolsList to SkinTemplate" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71800 [14:28:22] New patchset: KartikMistry; "Possible license issue with Utkal font" [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/71803 [14:32:06] New patchset: Rasel160; "WIP, do not merge. Will allow dynamic form generation in the details step." [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/66561 [14:32:41] New review: Rasel160; "(1 comment)" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/66561 [14:35:06] New patchset: Hashar; "set ansicolor colormap to vga" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/71805 [14:35:24] New review: Hashar; "pending upstream change" [integration/jenkins-job-builder-config] (master) C: -2; - https://gerrit.wikimedia.org/r/71805 [14:48:50] New patchset: Rasel160; "WIP, do not merge. Will allow dynamic form generation in the details step." [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/66561 [15:00:01] New review: Anomie; "Code looks fine, and the tests seem to run correctly through PHPUnit locally. Did make a few style o..." [mediawiki/extensions/AbuseFilter] (master) C: 1; - https://gerrit.wikimedia.org/r/71571 [15:08:09] New patchset: Rasel160; "WIP, do not merge. Will allow dynamic form generation in the details step." [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/66561 [15:20:39] New review: Alex Monk; "needs rebase" [mediawiki/extensions/OpenStackManager] (master) C: -1; - https://gerrit.wikimedia.org/r/65269 [15:29:29] New patchset: J; "Optional SVG mode with PNG fallback" [mediawiki/extensions/Score] (master) - https://gerrit.wikimedia.org/r/62243 [15:33:13] New review: Matmarex; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71705 [15:33:53] hoo: you alive? [15:34:16] MatmaRex: Sure :D [15:35:43] hoo: i added you to three changes fixing obscure extensions aften the change you merged yesterday :P [15:35:47] after* [15:36:11] ok, I guess I can have a look later today [15:36:22] going to unbreak wikibase firsth, though :P [15:36:54] heh [15:37:00] good luck [15:42:23] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [15:43:20] New patchset: Cmcmahon; "WIP add ve_refs check" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/69128 [15:49:18] MatmaRex, pong [15:49:39] i want people to use api.php/query?prop=whatever [15:49:48] without the "action" [15:51:20] hm. [15:51:29] that's still going to be annoying to use [15:51:40] MatmaRex, why so? [15:51:55] New patchset: Anomie; "Fix mw.ustring edge cases" [mediawiki/extensions/Scribunto] (master) - https://gerrit.wikimedia.org/r/71816 [15:52:07] because i can't use standard URL libraries to construct the entire request URL from a dictionary/hash/object/associative-array [15:52:16] i'd have to handle theaction myself [15:52:18] there is a very good reason for this -- it forces action to always be part of the URL, not POST request, and allows varnish to direct traffic based on action to different clusters [15:52:55] hmmmm, well that's at least be a justification [15:53:20] it'd still be annoying, of course, but that makes sense :) [15:53:29] MatmaRex, the fact that you would have to write one line to extract "action" from the param array and append it to the URL is not that influential :) [15:54:01] parsoid ppl have asked for this [15:54:14] yurik: hey, this is a rather critical detail you didn't mention anywhere :) [15:54:32] sure i did - it has been discussed for a while, including the RFC [15:54:43] s/anywhere/in the commit message/ [15:54:58] i will probably make it a separate commit [15:55:18] anomie is not happy that it is bunched together with the module versioning [15:55:39] so i will make one patch with versioning, and onether that uses that versioning when it sees /action [15:56:14] anomie is hard to please, but he usually makes sense :-P [15:57:54] ;) [15:59:46] * brion starts spouting crazy api ideas just to confuse yurik … :D [16:00:04] allow client to POST multiple JSON objects with request parameters to bundle multiple requests together! [16:00:06] * brion hides [16:00:29] nope, ain't happening :) lots of discussion on that one, there is really no need [16:01:33] its already batched up together as much as we can, and if the multiple requests need to do stuff in sequence, there has to be handling of "step #4 failed, stopping", whereas if its not sequential, do them in parrallel, more efficient [16:01:33] hehehe [16:01:44] http pipelining oughta work anyway [16:01:52] mooeypoo: hi there [16:02:01] plus we won't be able to send stuff to different subclusters based on action [16:02:18] well [16:02:31] we could fan the requests out to multiple backend servers [16:02:36] but let's not get ahead of ourselves :D [16:03:11] New patchset: Anomie; "Make lang:parseFormatterNumber more like tonumber" [mediawiki/extensions/Scribunto] (master) - https://gerrit.wikimedia.org/r/71820 [16:04:03] man i've got a lot of old email to clear out [16:04:08] * brion wanders off to gmail and hits 'archive' a lot [16:05:10] brion, use "e" to archive - by far the most useful button in gmail :) [16:05:23] yurik: yes, keyboard is lovely :D [16:05:30] though i've gotten used to swiping in the android app [16:05:35] * brion misses touchscreen on his laptop [16:06:03] guys, get a real mail client. :P [16:06:30] pine is the one true email client [16:06:32] New patchset: Anomie; "Make lang:parseFormattedNumber more like tonumber" [mediawiki/extensions/Scribunto] (master) - https://gerrit.wikimedia.org/r/71820 [16:06:48] New patchset: J; "(bug 50617) fix iframe output" [mediawiki/extensions/TimedMediaHandler] (master) - https://gerrit.wikimedia.org/r/71821 [16:07:00] sumanah: heyhey [16:07:22] sorry, was busy trying to translate stuff [16:07:34] mooeypoo: no prob! just wanted to wave and ask how your GSoC summer is going. [16:07:59] oh hey, Alpine got a new release this year. https://groups.google.com/forum/#!msg/comp.mail.pine/4w_f1ro6GRA/faiu1Egw_ywJ [16:09:33] great so far! :) very hectic, VE is going through metamorophosis we have to keep up with, but it's awesome seeing things come together [16:10:34] mooeypoo: you're working on the RTL support? [16:11:02] sumanah: the weird thing now is that I realize how much I don't know my technical *hebrew*.. ain't that an irony [16:11:07] brion: aye aye [16:11:11] awesome [16:11:13] or rather, eya eya [16:11:17] mooeypoo: hahaha [16:11:28] ahahah [16:11:41] Dan Kaminsky says "I love being wrong - it means the world is more interesting than I thought it was" [16:11:42] the word "feature" got me completely stumped yesterday on how to say it in Hebrew [16:11:57] I love the many things you are learning :D [16:12:11] New review: CSteipp; "It looks like the logic was correctly duplicated, and the new function looks sane. Working on my tes..." [mediawiki/core] (master) C: 1; - https://gerrit.wikimedia.org/r/71800 [16:12:12] True, I just wish I was wrong *less* in this case, it's still my main language [16:12:20] with luck, VE will be a friendlier environment than markup for bidi text. i've tried, and though it doesn't help not speaking the language, ….. that markup gets scary with mixed latin and hebrew :D [16:12:23] I'm like in language limbo! :D [16:13:02] brion: as a noobie, I have to say, I have NO IDEA how people managed to do mixed hebrew with wiki markup (or even html markup) up until now [16:13:26] I'm translating, and that uses markup -- and that's soooo confusing when it comes to using even things like {{ }} or [] inside hebrew text. [16:13:31] "with great difficulty" [16:14:52] hehe yeah [16:17:00] and I never really noticed how convenient it is that English doesn't have gender in its verbs, by the way. [16:17:09] That's another interesting conclusion from translating. [16:17:32] hmmm [16:17:40] does gerrit allow wildcard searches? [16:17:45] like "status:open project:mediawiki/extensions/*" [16:17:49] (this doesn't work) [16:18:10] "You can click the button to do whatever" is good in English. But in hebrew, you'd have to either use the male-voice (offending women) or the women-voice (alienating men) or rewrite that as a formal "This action can be done by clicking.." [16:18:48] mooeypoo: Wow, I never knew that.. [16:18:58] GorillaWarfare: me neither, and I *speak* the language. [16:19:03] Well, I knew it, I just never really noticed. [16:20:59] mooeypoo: that's why our translation system supports gender… it's complicated :) [16:23:33] brion: true, but then it's exhausting [16:24:05] I'm already getting vertigo from shifting ltr/rtl, if I have to think in two genders, I may well go crazy. [16:24:21] hehe [16:24:28] it's enough to make one give up text and go to all icons [16:24:33] then *nobody* understands what the software does [16:24:38] also, in hebrew I think it makes a bit more sense to go the "formal" tone in user guides, etc. [16:24:43] brion: IKEA. [16:24:47] lol [16:24:49] \o/ [16:25:19] localisation - it is hard [16:25:21] brion, mooeypoo: I think we should just use illustrations, like in airplane emergency exit guides [16:26:10] GorillaWarfare: I fly far too much and am getting to be a connoisseur of those guides [16:26:26] like, there are ones that do not make it super clear whether the overwing exits also provide rafts in the case of a water landing. Bad! [16:27:03] Qantas's was great - it explained *why they want you to do certain things* [16:27:06] with DATA [16:27:28] * sumanah is turning into Roan [16:28:14] MatmaRex: did you already look at the search docs? [16:28:14] GorillaWarfare: "Put your head between your legs / open door and jump onto the big floating yellow boat" works better in icons than "When you click the "References" button, you have the option of adding either a linked reference or a paragraph you can also format to your liking". [16:28:23] MatmaRex: https://gerrit.wikimedia.org/r/Documentation/user-search.html [16:28:26] ... but we *MIGHT* be able to do that with comic strips [16:28:32] ... now that's an idea right there [16:28:51] MatmaRex: I've been able to do, for instance, https://gerrit.wikimedia.org/r/#q,status:open+project:%255Emediawiki.*,n,z "status:open project:^mediawiki.*" [16:29:17] Hey we have a precedent with the upload licensing tutorial [16:29:35] there's an upload licensing tutorial? [16:29:46] mooeypoo: UploadWizard - the cartoon [16:29:56] There's a cartoon!? yay! where? [16:30:11] augh and I thought I was all revolutionary. [16:30:11] https://commons.wikimedia.org/wiki/File:Licensing_tutorial_en.svg [16:30:15] sumanah: yeah,ive found it in the meantime [16:30:19] thanks [16:30:40] mooeypoo: Honestly, I think an illustration of "put your head between your legs" might be helpful for most of our instructional pages [16:30:46] ah! I think I saw that before. [16:30:48] mooeypoo: if you can't be revolutionary, then comfort yourself with the phrase " best practice " [16:30:52] haha GorillaWarfare ++ [16:31:20] דוצשמשי: עםםג שגהןבק [16:31:23] err [16:31:26] sumanah: good advice [16:31:45] GorillaWarfare: mooeypoo - https://meta.wikimedia.org/wiki/Participation:Ocaasi,_Valeriej,_The_wub/Open_Help_Conference/Report a few Wikimedians went to a conference recently to learn about how to do help pages/instructions better in MediaWiki & Wikipedia [16:32:23] interesting [16:32:23] Oh yeah, I remember the wub was working on help pages [16:32:25] * GorillaWarfare looks [16:33:29] Ooh, I'm going to link [16:33:35] Oops [16:33:36] Was the answer - take everything we're currently doing, and do the opposite :P [16:33:44] *Ooh, I'm going to link https://en.wikipedia.org/wiki/Wikipedia:Help_Project/guidelines to some of the helpers in the IRC chan [16:34:17] New patchset: Rasel160; "WIP, do not merge. Will allow dynamic form generation in the details step." [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/66561 [16:34:42] the key thing with help pages is not to let wikipedians keep making them longer and longer [16:34:53] it's a huge temptation to just add one more admonition [16:34:57] it is [16:35:05] constraints are good [16:35:08] bawolff: well, I take everything you guys are LTR'ing and I'm RTL'ing it. So I guess that's pretty much exactly what I'm doing. [16:35:15] hahaha [16:35:32] mooeypoo: you're from the mirror universe! do you have a goatee like Mirror Spock? [16:35:43] :) [16:35:55] rofl [16:36:40] New patchset: Aude; "Update Wikibase, removing propagate moves patch in this branch" [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71827 [16:37:19] Shh... you weren't supposed to tell [16:37:54] New review: Yaron Koren; "Looks good! All the code is well-formatted. I don't know if "cache seconds" is an ideal name for tha..." [mediawiki/extensions/ExternalData] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/71722 [16:37:54] Change merged: Yaron Koren; [mediawiki/extensions/ExternalData] (master) - https://gerrit.wikimedia.org/r/71722 [16:37:57] ^d: Reedy around? [16:38:04] https://commons.wikimedia.org/wiki/File:Identification_Department,_Calgary_-_Criminal_Identification_Book_-_2_%28HS85-10-38280-2%29.jpg confirms what firefly has taught us. All the bad guys wear hats [16:38:09] * aude needs https://gerrit.wikimedia.org/r/#/c/71827/ deployed (sorry to bother) [16:38:54] Yeah [16:39:41] should be all we need until monday [16:39:43] aude: http://www.openitp.org/openitp/circumvention-tech-summit.html it's in Berlin - think you would like to go? [16:39:56] sumanah: i think so :) [16:40:14] it's on a weekend, so would work [16:40:27] you should totally go [16:40:33] and you should encourage other Berlin folks [16:40:43] sure :) [16:41:20] New patchset: Kipcool; "Specialdatasearch" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/71829 [16:42:33] Change merged: Kipcool; [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/71829 [16:45:22] Change merged: jenkins-bot; [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71827 [16:51:12] Change merged: jenkins-bot; [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/71659 [16:56:25] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [17:00:38] New patchset: AzaToth; "Add debian-glue macros" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/71245 [17:01:58] New patchset: Eranroz; "Support ApiPageSet in ApiWatch" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/53964 [17:01:59] * marktraceur loves getting pinged by all these people saying UploadWizard. [17:12:49] New review: Rjain; "I think we can merge this if its OK since there are other commits depending on this and are facing m..." [mediawiki/extensions/Annotator] (master) - https://gerrit.wikimedia.org/r/70794 [17:13:26] New patchset: Yurik; "Update zero ext" [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71831 [17:14:06] New patchset: Yurik; "Update zero ext" [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/71832 [17:16:45] New patchset: MarkTraceur; "Make DI tests nonvoting" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/71833 [17:17:12] Change merged: MarkTraceur; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/71833 [17:18:38] Change merged: Yurik; [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71831 [17:18:40] Change merged: Yurik; [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/71832 [17:21:32] New patchset: Mwalker; "Starting the Great Form Purge of O-13 Phase 1: Basic cruft removal!" [mediawiki/extensions/DonationInterface] (master) - https://gerrit.wikimedia.org/r/71750 [17:24:23] Change merged: jenkins-bot; [mediawiki/extensions/DonationInterface] (master) - https://gerrit.wikimedia.org/r/71750 [17:32:51] Special:RecordImpression ? what's that ? [17:33:39] New patchset: Rasel160; "WIP, do not merge. Will allow dynamic form generation in the details step." [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/66561 [17:33:48] New patchset: Rachel99; "WIP VE links" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/71724 [17:35:06] New patchset: Matmarex; "Fixed $wgCookieExpiration functionality when set to 0." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/61593 [17:36:17] New review: Matmarex; "Simple enough. I'm fixing the docs at https://www.mediawiki.org/wiki/Manual:$wgCookieExpiration too." [mediawiki/core] (master) C: 2; - https://gerrit.wikimedia.org/r/61593 [17:36:41] * SickPanda looks for a marktraceur [17:37:17] * marktraceur hides [17:37:32] wow, we have a LOT of JS requests on an average page... [17:37:34] SickPanda: hey -- re that giant UW ContentHandler patch -- i skimmed through it last night and it LGTM'd to me. are you depending on me for thorough review, or are you covered? [17:37:50] ori-l: i'm looking at marktraceur for review, since it is UW. [17:37:52] SickPanda: I'm working on a patch for mwalker first, then reviewing Rasel160's latest, then you'll get my attention [17:37:57] thedj: don't you say. [17:37:58] marktraceur: yessir [17:38:10] SickPanda: cool, thanks [17:38:14] thedj: WMF should seriously consider removing some extensions, really, eh. [17:38:28] ori-l: I'll poke you when I start moving JsonContent and friends into core. [17:38:43] ori-l: and when I start writing the JSON schema for this one. [17:38:52] ori-l: want to get this merged first, since it is already huge. [17:39:02] hi rachel99 [17:39:04] SickPanda: Just for the record, there's a maintenance script in your patch that will convert things. Right? [17:39:09] SickPanda: yep, i saw your comment, makes sense to em [17:39:10] me [17:39:24] marktraceur: not yet. will be. Don't want to make this patch even bigger. [17:39:35] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/61593 [17:39:42] New patchset: Anomie; "Add 'viewmyprivateinfo', 'editmyprivateinfo', and 'editmyoptions' rights" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/67876 [17:39:50] marktraceur: it's trivially convertible, and I'll add it if you want me to. This patch is already well over 2000LOC changes tho, so... [17:40:11] MatmaRex: well a lot would be solved with actually profiling stuff once in a while. [17:40:12] ^D: hello [17:40:18] instead we only seem to add. [17:40:21] ^d: http://pastebin.com/z9MJLRz6 gerrit [17:40:27] New patchset: Anomie; "Add 'viewmyprivateinfo', 'editmyprivateinfo', and 'editmyoptions' rights" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/67876 [17:40:36] New review: Rasel160; "(1 comment)" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/66561 [17:40:36] ^d Hello [17:40:36] I'm getting that when it hangs with "working...", not sure if they are related [17:40:44] * thedj wonders if ReferenceTooltips is resourceloader compatible [17:40:48] New patchset: Kipcool; "SpecialDatasearch" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/71838 [17:40:57] ^d : getting an error logging into Gerrit [17:41:07] marktraceur: I've marked TODOs in the comments. Essentially, Maintainance scripts, i18n fixing, and schema verification. will add after i get this merged [17:41:18] thedj: it is, or it almost is [17:41:19] ^d : cannot assign user name [17:41:28] thedj: the version we use at pl.wp is loaded via RL [17:41:34] SickPanda: In that case, the old system is still in place and will work for the time being? [17:42:03] ^d : Can you help with me out? [17:42:09] Like, after I merge this patch, I will still have campaigns in the DB and UW will use the DB to access them, not get them from Campaign:Whatever? [17:42:37] Change merged: Kipcool; [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/71838 [17:42:50] MatmaRex: i'll switch the en.wp version to that as well then, if i can't find a previous revert [17:43:53] thedj: ours might be a modified, though [17:44:08] because when it comes to MW gadgets, nobody ever looks upstream [17:44:52] bawolff, I'm late to the party, but I think the "height constrained no overlay" version of your new gallery is excellent. [17:45:18] ragesoss: thanks [17:45:27] New review: Matmarex; "Wouldn't it be simpler to include the version information in the cache key?" [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/54207 [17:45:45] hey, it's a bawolff [17:45:50] MatmaRex: hmm, i can't even get the thing to work on test.wp [17:46:05] bawolff: wanna look at ^^? it'd be good if someone who actually knows what they're doing reviewed it :) [17:46:05] MatmaRex: I've been here for several hours :P [17:46:23] That's never a good way to introduce please review my patch [17:46:29] Now I think its going to be scary ;) [17:46:36] the patch isn't even mine :D [17:46:40] and it's +5, -1 [17:47:09] thedj: let's do it live! [17:47:17] ragesoss: So far I've gotten that hover is controversial, and one of the two non-hover modes would be best [17:48:03] yeah, I think many times the captions are an essential part of galleries. [17:48:17] users need to be able to see all that info at once, in some cases. [17:48:39] I think it depends on context. In some places it probably doesn't super matter (galleries on category pages where the caption says the file size) [17:48:40] and I much prefer the one that doesn't cover up images with caption. [17:49:08] yes, in some cases, the focus really is on just being a gallery of pretty images related to the page. [17:49:31] Personally I find the current gallery doesn't really show off the beauty of the images [17:49:39] New review: Anomie; "PS11: Remove 'viewprivateinfo' requirement for seeing real name, remove permission requirement for v..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/67876 [17:49:45] particularly on things like galleries of featured pictures [17:49:45] no, the current gallery is awful. [17:50:09] the backgrounds, and the wasted space both within and outside of the backgrounds... really, just terrible. [17:50:18] And commons has some very beautiful pictures [17:50:57] yeah. the hover-captions version would actually make plenty of sense on Commons. [17:51:53] biggest issue with hover captions is that doesn't work well with touch screens [17:52:14] I have it falling back to being always on in that case, but still weird to have two different behaviours [17:55:18] New patchset: Yurik; "Update zero ext" [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/71841 [17:55:42] New review: Alex Monk; "needs rebase" [mediawiki/extensions/OpenStackManager] (master) C: -1; - https://gerrit.wikimedia.org/r/65557 [17:55:58] Change merged: Yurik; [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/71841 [17:56:40] MatmaRex: i see no changes between the .pl and the .en version, so if you never heard a complaint on pl, i'm just gonna switch it to RL [17:56:41] New patchset: Rachel99; "WIP VE links" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/71724 [17:56:43] bawolff: re: touch screens. Isn't most of these stuff not loaded when viewed with MobileFrontend anyway? [17:57:06] That assumes people look at it with mobile frontend [17:57:12] And good question, I don't even know [17:57:17] * bawolff checks [17:57:23] where did my phone go [17:57:45] bawolff: IIRC it is, unless you specifically target mobile with RL [17:58:01] bawolff: also, commons by default doesn't redirect to mobilefrontend, so it does get the desktop site, warts'n'all [17:58:03] Well its largely css [17:58:13] thedj: yeah, it's always worked, and the only change seems to be translation and making it run in all namespaces [17:58:33] wait, yeah guess I can't test my testwiki as it doesn't have mobile frontend installed [17:58:42] New patchset: Yurik; "Update zero ext" [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71842 [18:00:14] Interesting. On current code, on mobile galleries have rounded corners [18:00:44] bawolff: all RL modules are not loaded on MF unless they are specifically targetted to mobile, though. [18:00:52] also, use vagrant! :D [18:03:32] Change merged: Yurik; [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71842 [18:06:02] New review: Yuvipanda; "Create a page in the Campaign: namespace with the following content(https://dpaste.de/BwMkK/) to tes..." [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [18:07:16] marktraceur: i'm going to try and sleep this sickness off. see you tomorrow! :) [18:07:18] New review: Raimond Spekking; "If you prefix all message keys in the i18n files with the extension name like "linkedwiki-..." to av..." [mediawiki/extensions/LinkedWiki] (master) - https://gerrit.wikimedia.org/r/71180 [18:07:36] New review: Bsitu; "(1 comment)" [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/69861 [18:07:48] feel better soon SickPanda! [18:07:54] ty ragesoss. [18:08:03] New patchset: Cmcmahon; "WIP VE links" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/71724 [18:10:00] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [18:11:29] New patchset: Hoo man; "Allow running the AbuseFilter parser tests via phpunit" [mediawiki/extensions/AbuseFilter] (master) - https://gerrit.wikimedia.org/r/71571 [18:12:17] yikes... ULS does a lot of DOM walking and editing. [18:12:35] thedj: it was written by designers [18:12:39] it really should bunch all it's edits that will save a ton of load time i suspect. [18:12:41] or at least that's the only reasonable explanation [18:12:54] thedj: you know why the langauge dialog is/was lazy-loaded? [18:13:01] because it took ten seconds to render in its entirety [18:13:01] New review: Hoo man; "Addressed Anomie's code style comments (thanks for those)" [mediawiki/extensions/AbuseFilter] (master) - https://gerrit.wikimedia.org/r/71571 [18:13:13] New patchset: SuchABot; "Redo campaigns backend to use ContentHandler" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [18:13:32] MatmaRex: not surprised [18:14:33] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [18:15:47] MatmaRex: it might be me, but i don't have any reference tooltips on en.wp [18:16:14] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [18:16:38] thedj: do you have popups? [18:16:45] MatmaRex: disabled [18:16:49] New review: Anomie; "If no one else reviews relatively soon, ping me for +2." [mediawiki/extensions/AbuseFilter] (master) C: 1; - https://gerrit.wikimedia.org/r/71571 [18:16:52] thedj: reftooltips disable temselves if they detect popups [18:16:58] as in, window.pg being truthy [18:17:19] New review: Yuvipanda; "Specific things I'm hoping people could look at:" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [18:18:22] MatmaRex: now i restarted on IE, and it's just showing a border around the ref, it's not creating a popup. [18:18:59] thedj: it does that when you have the reference onscreen already [18:19:08] thedj: it sometimes annoys me as well [18:19:15] but not enough to bother changing that [18:20:37] New patchset: Yaron Koren; "Fixes for partial form handling, thanks to patch from Sebastian Richter" [mediawiki/extensions/SemanticForms] (master) - https://gerrit.wikimedia.org/r/71846 [18:21:44] New review: Matmarex; "Oh please, no more preferences, unless you can convince me it's needed." [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/46790 [18:22:50] nobody's on RT duty? [18:23:00] ah, wrong channel ^^ [18:23:37] we have too many channels. [18:24:13] nah, I just saw a bugzilla discussion and thought that was -tech so -ops is the next window [18:24:42] so in other words, we have too many channels. [18:24:46] Change merged: Yaron Koren; [mediawiki/extensions/SemanticForms] (master) - https://gerrit.wikimedia.org/r/71846 [18:29:13] New patchset: Ori.livneh; "Use '-f' rather than '|| true' trick to suppress error on nonexistant files" [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/71848 [18:29:18] ^ superm401 [18:29:38] Change merged: jenkins-bot; [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/71848 [18:32:12] ori-l, know what to make of this? [18:32:14] https://dpaste.de/92Uqd/ [18:32:15] New review: Nischayn22; "You invited all of this :p" [mediawiki/extensions/UploadWizard] (master) C: -1; - https://gerrit.wikimedia.org/r/70446 [18:32:20] I tried to update the guest additions. [18:34:27] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [18:41:12] New patchset: SuchABot; "Redo campaigns backend to use ContentHandler" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [18:42:03] New patchset: Kipcool; "Specialdatasearch query speed" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/71849 [18:42:05] New patchset: SuchABot; "Redo campaigns backend to use ContentHandler" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [18:42:32] New patchset: Mattflaschen; "Add more detail to settings documentation." [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/71850 [18:42:35] Change merged: Kipcool; [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/71849 [18:46:07] GorillaWarfare: how goes the fight? [18:46:24] It goes, mwalker :) [18:46:37] Working on a screencast of uploading files to Wikisource, per Raylton's suggestion [18:46:48] To record the current workflow [18:47:11] *thumbs up* (that'll probably help me too :) because I have no idea how that works!) [18:48:10] did anyone get anywhere on determining what metadata should go where? [18:48:23] mwalker: I haven't heard back from the email group [18:48:55] mwalker: I have https://meta.wikimedia.org/wiki/Book_management/Technical_notes#Integration_with_ProofreadPage onwiki, to record all the questions a little more concisely [18:50:21] *nods* reading now [18:50:56] New patchset: Kipcool; "forgot to actually use $join... trying with a second straight_join for syntrans" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/71853 [18:51:09] Change merged: Kipcool; [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/71853 [18:53:52] New patchset: Yurik; "API versioning" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71759 [18:54:53] New patchset: Yurik; "API versioning" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71759 [18:55:06] Hi Somebody on the BrowserQA ? [18:55:59] chrismcmahon: ^ [18:56:07] New review: Yurik; "Revised parameter model, will split into multiple patches in the next commit (want to preserve comme..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71759 [18:56:13] hi karima [18:56:25] thanks mwalker [18:56:43] Hi [18:56:58] New patchset: Bsitu; "Remove Echo preference change eventlogging" [mediawiki/extensions/Echo] (master) - https://gerrit.wikimedia.org/r/71854 [18:56:59] I tested the patch https://gerrit.wikimedia.org/r/#/c/70044/3,publish [18:57:42] now the step When(/^I click link Create$/) do is undefined [19:00:38] New review: Tpt; "See my comments in https://gerrit.wikimedia.org/r/#/c/71172/7/includes/page/ProofreadPageContent.php" [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) C: -1; - https://gerrit.wikimedia.org/r/71172 [19:00:46] karima: as I understand it, you want to run this test against a vagrant instance, is that correct? [19:01:28] yes with the last version of vagrant or VE is not intalled [19:02:52] class DoesNotExistPage is necessary for my extension but there are not link Create source [19:03:55] karima: right now those tests are intended to run in environments with VisualEditor in place. You can get them to pass when they run in your environment by altering them locally. [19:04:31] New patchset: Yurik; "API versioning in module~2 format" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71759 [19:04:35] ok [19:04:41] thanks [19:05:23] New review: Anomie; "(1 comment)" [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/53964 [19:06:03] karima: take a look at https://github.com/wikimedia/qa-browsertests/blob/master/features/support/pages/does_not_exist_page.rb . If you change "text: 'Create source'" to "text: 'Create'" then the tests that use that page will pass in your environment. [19:06:18] New review: MarkTraceur; "(1 comment)" [mediawiki/extensions/CentralNotice] (master) C: 2; - https://gerrit.wikimedia.org/r/71751 [19:06:50] yes I know but it's strange [19:07:48] I think I copy the class DoesNotExistPage directly in my project [19:08:36] thedj: Re https://bugzilla.wikimedia.org/show_bug.cgi?id=50677 , does that error display in non-debug mode? My understanding is it should be suppressed by a try-catch in that case [19:08:49] Change merged: jenkins-bot; [mediawiki/extensions/CentralNotice] (master) - https://gerrit.wikimedia.org/r/71751 [19:09:50] greg-g: can I get a deploy window :) [19:09:56] oh ai [19:10:01] RoanKattouw: lemme check [19:10:02] - [19:10:05] (there's a space there) [19:10:20] mwalker: so, yeah, just you and VE team asking for 'em today. [19:10:30] I assume VE isn't ready yet, so are you ready? [19:10:33] mwalker: And I'm scapping, so you could hitch a ride [19:10:34] thedj: https://bugzilla.wikimedia.org/show_bug.cgi?id=50677 is being discussed on #mediawiki-visualeditor right now [19:10:38] (At 4pm) [19:10:47] hurm... I dont need a scap [19:10:50] and i wont be around at 4 [19:10:56] but if you're not doing anything now I can go [19:11:03] right, that's what I was thinking [19:11:12] (now-ish) [19:11:16] New review: Mdale; "Tested and working in production mode. Giving this +2 because it fixes a pretty critical bug 50617." [mediawiki/extensions/TimedMediaHandler] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/71821 [19:11:18] Change merged: Mdale; [mediawiki/extensions/TimedMediaHandler] (master) - https://gerrit.wikimedia.org/r/71821 [19:11:23] RoanKattouw: so, you definitely just want to take the 4pm slot? [19:11:42] thedj: might wanna join? [19:11:44] New review: Rtdwivedi; "I used the variables so that the cases being tested become clear. Or is it clear enough without the ..." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [19:12:33] greg-g: I don't need anything earlier in the day, no [19:12:40] RoanKattouw: cool. [19:12:54] mwalker: go forth and make banners display [19:13:04] mwalker: mind adding to deploy calendar? [19:13:08] will do! [19:13:12] thanks! [19:14:33] New patchset: Bsitu; "Remove invalid maintenance scripts" [mediawiki/extensions/Echo] (master) - https://gerrit.wikimedia.org/r/71856 [19:16:44] New review: Tpt; "I think it's clear enough without variable names." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [19:20:03] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [19:20:42] jdlrobson, awjr_away: I just noticed -- what I'm about to push out for CentralNotice includes the patch enabling it on the mobile stable site [19:20:50] (if that didn't already go out on the train) [19:20:58] New patchset: Yurik; "PATH_INFO-based API actions" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71858 [19:21:03] mwalker: that shouldn't be a problem [19:21:17] just wanted to heads up you guys in case there was a problem :) [19:24:47] New review: Matmarex; "(1 comment)" [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/71530 [19:24:49] New review: Brian Wolff; "Not really a fan of this idea. API is for machines, they can deal with ugly urls. PATH_INFO doesn't ..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71858 [19:25:55] anomie, MatmaRex - api was split into two patches, justlike anomie wanted :) [19:26:23] New patchset: Rtdwivedi; "Added unit tests for Page pages." [mediawiki/extensions/ProofreadPage] (pagePagesRefactoring) - https://gerrit.wikimedia.org/r/71172 [19:27:54] yurik: and see, bawolff iscomplaining abouthe same thing i complained about [19:28:07] yurik: commit summaries should be like novels, you know ;) [19:28:26] MatmaRex, i wrote more text this time :) [19:28:45] two whole lines :P [19:29:19] I can haz a tech talk hangout? [19:30:24] MaxSem: there's a youtube stream: http://www.youtube.com/watch?v=cVkItxI0I0c [19:30:27] MatmaRex: nice comment on the design-l thread [19:31:11] since I do Solr stuff too, I'd like a two-way mean of communication:) [19:31:42] MaxSem: you could ask Chip for the G+ address? cndiv_ [19:31:55] What is today's tech talk? [19:32:03] Solr [19:32:04] bawolff: heh [19:32:10] when is it? [19:32:25] tech talk did not start yet? [19:32:26] NOW!!! [19:32:44] MatmaRex: I resisted very hard about making a use foss tools comment ;) [19:33:12] Both xcf and svg are allowed on commons... [19:33:21] bawolff: well, i use lots of closed-source software myself :P [19:33:31] opera, sublime text [19:33:34] windows. /me hides [19:33:35] <^d> <3 sublime [19:33:42] manybubbles: are we expecting Rainman today? [19:33:45] hell yeah <3 sublime. [19:34:45] New patchset: Mwalker; "Update CentralNotice to master" [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/71861 [19:35:01] qgil: tech talk did not start yet? [19:35:21] rainman-sr: could you PM me the Google ID you'd like to join with? [19:35:23] <^d> zeljkof: Just about to start. [19:35:31] ^d: thanks [19:36:05] New patchset: Mwalker; "Update CentralNotice to master" [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71862 [19:36:43] now [19:37:05] http://www.youtube.com/watch?v=cVkItxI0I0c [19:37:05] Solr based search [19:37:24] qgil: I am der [19:37:26] I'm in qgil ^d [19:37:35] robla: ping [19:37:59] good! [19:38:29] Change merged: Mwalker; [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/71861 [19:39:27] I do not see the slides in youtube stream [19:39:33] Oh no, the youtube stream is doing the "An error occurred, please try again later" thing again :( [19:40:03] anomie: youtube stream works for me [19:40:18] not seeing slides, just a blank profile? [19:40:23] me too [19:40:26] me too [19:40:29] me too [19:40:31] I can hear audio fine though [19:40:31] slides are back for me [19:40:32] now [19:40:40] not for me [19:40:44] nevermind [19:40:57] Hmm, does not like my Firefox for some reason [19:40:59] back [19:41:02] Chromium works here. [19:41:12] i can see chad now [19:41:18] mglaser: same here :) [19:41:34] Change merged: jenkins-bot; [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71862 [19:43:08] Reedy: I'm unable to rebase https://gerrit.wikimedia.org/r/?dbg=1#/c/33713/ , could you help again pleeeeeeeeeeease? [19:43:28] Has it changed that much again? :/ [19:43:35] if you have any questions about the tech talk please ping me and I will ask. thanks! [19:44:04] Reedy: dunno, it seems someone added a mail class at the end of the file, where I add mine [19:44:07] qgil: will do :) [19:44:22] plus some hundreds of other changes in the repo of course [19:44:52] or else we can comment over there I guess qgil ? [19:45:27] configured via http puts? [19:45:27] harshkothari, well, if you want that the question / answer happens in the video stream. Otherwise of course just comemnt here. [19:45:33] What's with the weird http methods [19:45:55] bawolff, is that a question or a comment? (just checking) :) [19:46:02] more a comment [19:46:15] I'll ping you if its a real question [19:46:17] ok [19:46:32] New patchset: Cmcmahon; "First iteration complete for test2wiki please review" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/69128 [19:49:34] New patchset: Cmcmahon; "VE refs first iteration complete for test2wiki please review" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/69128 [19:51:01] oh noes, no slides again :( [19:51:08] New patchset: CSteipp; "Basic MediaWiki integration, library cleanup" [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/70316 [19:51:23] yes no slide :( [19:51:28] zelikof, same for me [19:51:33] ahhh [19:51:41] greg-g: since nothing seems to have blown up -- I'm calling my deployment complete [19:53:05] mwalker: *knock on wood* [19:53:27] indexing the fully rendered page: at BlueSpice, we came to the same conclusion :) [19:54:09] polling? [19:54:27] New patchset: Ragesoss; "switch to correct mailTo address for crash reports" [apps/android/commons] (master) - https://gerrit.wikimedia.org/r/71866 [19:54:45] http://solr-mw.instance-proxy.wmflabs.org/wiki/Main_Page [19:55:13] ^^^^click that link :) [19:55:29] New patchset: Kipcool; "specialSuggest using element attributes instead of an hidden input for each variable" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/71867 [19:55:45] http://8983.solr1.instance-proxy.wmflabs.org/solr/#/jawiki_shard1_replica1/analysis?analysis.query=%E6%97%A5%E6%9C%AC%E8%AA%9E&analysis.fieldname=title&verbose_output=1 [19:56:00] ^^^^ Solr's admin console [19:56:21] http://ganglia.wmflabs.org/latest/?r=hour&cs=&ce=&c=solr&h=solr0&tab=m&vn=&mc=2&z=medium&metric_group=NOGROUPS_|_nikwiki_|_memory [19:56:54] time for questions! [19:58:22] I have a hard time understanding robla [19:58:27] Our caching layer must already know when a page needs to be re-rendered… could we reindex at the same time? Would that solve the mystery of template/transcluding changes? [19:58:33] can somebody type he question here? [19:58:58] andrewbogott, +1 [19:59:07] he introduced the developer of the current search engine still in production, rainman-sr [19:59:13] (who is talking now) [19:59:20] andrewbogott: smart [19:59:38] g [19:59:42] qgil: thanks [20:00:05] I wonder if there should be a "MediaWiki extensions > CirrusSearch" component in Bugzilla already to handle bugs, or if things are still too experimental. But that's probably stuff to ask later on, via email :P [20:00:09] andrewbogott, I'll ask when there is a chance [20:00:37] qgil: can you give a microphone to robla? :) [20:00:46] he has one :) [20:01:03] qgil: can you give him another one, this one sucks :) [20:01:21] New review: Kipcool; "@Hiong3-eng5: could you check that this works for you?" [mediawiki/extensions/WikiLexicalData] (master) - https://gerrit.wikimedia.org/r/71867 [20:01:25] <^d> andrewbogott: That was my thought already, I just hadn't dug into it yet [20:01:36] this time better? [20:01:41] I'm curious whether the new setup will help enabling cross-wiki search (i.e. searching from all wikis at once)? [20:02:11] qgil: better, but he is still hard to understand [20:02:15] hmm, if I listen very closely I can understand RobLa [20:02:18] <^d> andre__: Yes, we should eventually. It's still a bit early so far since nobody has it deployed yet. [20:02:27] ^d: Yeah, I have no idea how that decision is made, just that it must be a solved problem. [20:02:40] <^d> Nikerabbit: I'll let manybubbles say more, but I think it's probably in the realm of what's possible. [20:02:49] andre__: I can understand just fine everybody else, except robla [20:02:59] zeljkof, same here [20:03:10] ok, we are ditching that mic [20:03:16] qgil: thanks [20:03:20] I'll steal qgil's next time I speak :-) [20:03:30] try [20:03:34] robla: please do :) [20:03:49] It's sharing, not stealing!! :P [20:03:57] <^d> andrewbogott: Yes. I just haven't looked if I can hook in there yet. [20:04:57] qgil: now I can hardly understand you :( [20:05:05] New review: MarkTraceur; "OK, now I'm going to say that I'm pretty much happy with the code and move on to functionality. So t..." [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/66561 [20:05:06] oops, asked and answered :) [20:05:20] but robla's microphone is great now :) [20:05:52] I imagine you'd want LinksUpdate not general HTMLCacheUpdate purges [20:06:58] New patchset: Hashar; "option to strip requested prefix index in the list" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71911 [20:07:17] zeljkof, is the problem volume or distortion? [20:07:30] file proper bugs, man! [20:07:33] :) [20:07:49] <^d> bawolff: I thought that. But just hadn't yet. [20:07:50] Right now sound is fine again. need another testcase! [20:07:59] qgil: both, the volume is low, and there is slight distortion, like you are far away from the microphone [20:08:39] any more questions? [20:08:41] If you do it on LinksUpdate, you would already have the ParserOutput around, so that would probably be most of the cpu cycles already [20:09:09] <^d> bawolff: Yup, you're reading my mind :) [20:09:18] marktraceur: morning. do you have access rights to reload Zuul ? [20:09:24] Yup [20:09:30] marktraceur: :-] [20:09:31] I have done it a few times in the past 24 hours [20:09:51] hashar: Hopefully I didn't blow things up [20:10:09] well if you look at the layout diff :-) you should be fine [20:10:14] *nod* [20:10:17] Now I know [20:10:20] plus Jenkins is supposed to complain / block if the layout is invalid [20:10:27] Yeah, it does [20:10:33] I haven't seen it, but I know that test is voting [20:10:35] and if all fails, git revert and repeat :-] [20:11:13] *nod* [20:11:30] I certainly made sure I didn't see any egregious errors in the logs, as the wiki page requires. [20:13:38] New patchset: Aaron Schulz; "[WIP] Added consumer request and approval pages" [mediawiki/extensions/OAuth] (master) - https://gerrit.wikimedia.org/r/71156 [20:15:51] New patchset: Cmcmahon; "Test for page elements for VisualEditor References UI review" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/69128 [20:15:59] New review: Parent5446; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71911 [20:16:09] New patchset: Karima Rafes; "Add table.feature" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/71913 [20:16:39] New review: Yurik; "There are very specific reasons why we need this: it would allow sub-clusters to process API actions..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71858 [20:17:01] New review: TheDJ; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71530 [20:17:43] it's a wrap! [20:17:48] thank you everybody! [20:17:57] thanks a lot! [20:18:08] rainman-sr, thank you as well! And sorry for the initial misunderstanding [20:18:15] thanks everybody, very interesting! [20:18:16] thanks everyone for all of your time. [20:18:21] qgil, no problems, it was good talking to you all again! [20:18:22] thanks manybubbles ^d [20:18:38] I'm already running Solr for few wikis, so I am interested in trying out Solr for wiki search as well [20:18:43] New review: Matmarex; "(And this should really be in the commit message.)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71858 [20:19:12] <^d> Nikerabbit: We're still a little bit alpha, but we would love to have you use it :) [20:19:38] ^d: I'm going to add to our list of thing to try setting up a copy of enwiki - I'm pretty sure we don't have the resources for that at our disposal right now. [20:20:08] <^d> Probably not, I doubt we'd be able to get it functioning even in the largest labs install. [20:20:15] <^d> Might need actual physical hardware. [20:20:19] New review: MarkTraceur; "Scratch the coordinates - those are part of something else entirely. Leave them out of the template ..." [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/66561 [20:21:03] ^d: I'll follow the progress [20:23:29] thedj: gods, i hate tablesorter. [20:24:09] MatmaRex: it just needs to do a lot. [20:24:31] MatmaRex: especially because we had all kinds of implied behavior in the old version. [20:25:57] thedj: also you might want to fix "colspan property" in commit message [20:26:20] also, wtf, fixed tests appear to work with new code [20:26:28] or RL is caching things too well [20:27:10] New patchset: TheDJ; "Tablesorter: colSpan property was incorrectly accessed" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71530 [20:29:01] New patchset: Karima Rafes; "I forget to remove gems" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/71917 [20:29:53] thedj: yeah, fixed tests still pass with broken code [20:30:00] how. [20:30:04] New patchset: Ryan Lane; "AJAXify delete instance action" [mediawiki/extensions/OpenStackManager] (master) - https://gerrit.wikimedia.org/r/65557 [20:30:59] so either the tests are shit [20:31:03] or they are not actually run [20:31:37] jdlrobson: uhhh... thought! regarding zero... if centralnotice is running via the mobile hook -- does that mean it's also running on zero.wm.o? [20:32:53] New patchset: Ryan Lane; "AJAXify delete instance action" [mediawiki/extensions/OpenStackManager] (master) - https://gerrit.wikimedia.org/r/65557 [20:33:02] New review: MarkTraceur; "See change Ib9b381816d18f2723a9f715c54f6a64ff4b69847 for long- and short-form text fields." [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/66561 [20:33:52] thedj: apparently jQuery normalizes .prop( 'rowspan', '3' ) in a way that it works. [20:34:06] New patchset: Zfilipin; "Test for page elements for VisualEditor References UI review" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/69128 [20:34:26] MatmaRex: yeah, but that is deprecated [20:34:30] Yippie, build fixed! [20:34:30] Project browsertests-en.wikipedia.org-linux-chrome build #164: FIXED in 5 min 18 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.org-linux-chrome/164/ [20:34:40] New review: Karima Rafes; "Zfilipin," [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/71917 [20:35:04] thedj: yeah, but that's not my point [20:35:09] these tests shouldn't have passed [20:35:17] or they're completely useless [20:35:29] New patchset: Mattflaschen; "Added the db schema" [mediawiki/extensions/Annotator] (master) - https://gerrit.wikimedia.org/r/70794 [20:35:35] yurik: question regarding zero -- I just pushed CentralNotice to the mobile stable site -- it looks like it's also running on the zero domains... which is probably a problem? [20:36:21] MatmaRex: can we be more specific about which tests ? [20:37:15] New review: Zfilipin; "Fixed whitespace in visual_editor_page.rb and renamed visual_editor_references.feature and visual_ed..." [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/69128 [20:37:21] New patchset: Ryan Lane; "Show more informative instance status" [mediawiki/extensions/OpenStackManager] (master) - https://gerrit.wikimedia.org/r/71920 [20:37:23] uh, wait [20:37:28] Change merged: Zfilipin; [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/69128 [20:37:32] so the tests you touched are not actually relevant to the fix? [20:37:43] the tests in general pass with the broken code [20:38:18] ah yes, they would do that. [20:38:54] there is no testcase that would have covered this problem. [20:39:01] bleh. [20:39:09] can you add one, pretty-please? :) [20:39:16] sorry, should have told you that. [20:39:35] i should have checked [20:39:40] New patchset: Ryan Lane; "Show more informative instance status" [mediawiki/extensions/OpenStackManager] (master) - https://gerrit.wikimedia.org/r/65269 [20:40:00] anyone feel like a review? https://gerrit.wikimedia.org/r/#/c/65557/ :) [20:40:27] Yippie, build fixed! [20:40:27] Project browsertests-en.wikipedia.org-linux-firefox build #153: FIXED in 5 min 57 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.org-linux-firefox/153/ [20:40:44] looking [20:42:03] similarly: https://gerrit.wikimedia.org/r/#/c/65269/ [20:42:04] :) [20:44:47] MatmaRex: k, -1 it and i'll add a testcase, but i don't have the time to do it today. [20:46:09] okay. i really don't feel like touching tablesorter again myself, or i'd add it myself [20:46:11] Krenair: https://gerrit.wikimedia.org/r/#/c/68126/6/OpenStackManagerEvent.php,unified <— do you mean delete the function and rename the other one to something more generic? [20:46:21] (also, i wouldn't be allowed to merge if i did it,so there's that) [20:46:35] and add a param that can specify the event type? [20:46:59] Yes [20:47:02] Sorry if it wasn't clear [20:48:25] New review: Hashar; "Need to give it a try :)" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/71245 [20:49:20] New review: Mwalker; "(1 comment)" [mediawiki/extensions/OpenStackManager] (master) - https://gerrit.wikimedia.org/r/65557 [20:49:23] MatmaRex: i'm not sure if I can actually find a testcase. I doubt this was actually affecting something in live code... [20:50:21] MatmaRex: you see, this.column isn't used by any other code. [20:51:11] New review: Mattflaschen; "This looks good to me." [mediawiki/extensions/Annotator] (master) C: 2; - https://gerrit.wikimedia.org/r/70794 [20:51:15] Change merged: jenkins-bot; [mediawiki/extensions/Annotator] (master) - https://gerrit.wikimedia.org/r/70794 [20:51:30] undoing the headerToColumns[columnToHeader[this.column] would fail testcases. [20:51:38] Project browsertests-en.wikipedia.org-windows-internet_explorer_6 build #125: FAILURE in 6 min 1 sec: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.org-windows-internet_explorer_6/125/ [20:52:34] i can add a testcase that checks the value of this.column of course... [20:52:47] hm [20:52:59] New review: Brian Wolff; "Personally I think either approach is fine." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/54207 [20:52:59] so this doesn't fix any bugs, it just cleans up the code? [20:53:04] yes [20:53:12] hm [20:53:17] New patchset: Hashar; "Add debian-glue macros" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/71245 [20:53:22] 'twould be nice. [20:53:27] i'll think about it tomorrow [20:53:30] :P [20:53:46] there is actually another way to fix it. if you rename this.column to this.headerIndex and leave the code the same, you also fix it. [20:54:07] New review: Hashar; "I have added the 'pybal' repository as well." [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/71245 [20:54:13] (but reading the original code, that's not what was intended). [20:54:17] New patchset: Ryan Lane; "AJAXify delete instance action" [mediawiki/extensions/OpenStackManager] (master) - https://gerrit.wikimedia.org/r/65557 [20:54:17] Krenair: ok. cool [20:55:29] New review: Ryan Lane; "(1 comment)" [mediawiki/extensions/OpenStackManager] (master) - https://gerrit.wikimedia.org/r/65557 [20:58:28] New review: MarkTraceur; "All. The. Fail." [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [20:59:08] New review: Hashar; "The generated job names yield:" [integration/jenkins-job-builder-config] (master) C: -1; - https://gerrit.wikimedia.org/r/71245 [21:02:37] New patchset: Hashar; "gerrit project (no jobs yet)" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/71923 [21:02:49] New patchset: Hashar; "gerrit project (no jobs yet)" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/71923 [21:03:12] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/71923 [21:06:09] New patchset: Hashar; "Jenkins job validation (DO NOT SUBMIT)" [gerrit] (wmf) - https://gerrit.wikimedia.org/r/71924 [21:07:39] New review: Hashar; "And running it on labs needs a slightly different setup since the Zuul git repository is not there :-]" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/71245 [21:08:13] New patchset: Rachel99; "WIP VE links" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/71724 [21:13:07] New review: Hashar; "We will need the Zuul references (a merge commit of the patchset and the tip of the branc) to be ava..." [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/71245 [21:15:27] New review: Mattflaschen; "Can we just backport mediawiki.ui to 1.20 and 1.21 instead, to avoid needing this extension? See al..." [mediawiki/extensions/Agora] (master) C: -1; - https://gerrit.wikimedia.org/r/69458 [21:18:23] New patchset: Karima Rafes; "add prefix linkedwiki- in the i18n files" [mediawiki/extensions/LinkedWiki] (master) - https://gerrit.wikimedia.org/r/71926 [21:20:17] New review: Karima Rafes; "Raimond" [mediawiki/extensions/LinkedWiki] (master); V: 1 C: 1; - https://gerrit.wikimedia.org/r/71926 [21:24:49] New patchset: MarkTraceur; "licenseInput: Handle redirects for license templates" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/50755 [21:25:40] New review: MarkTraceur; "I got tired of waiting, implemented a bunch of asyncness in places that should have been async alrea..." [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/50755 [21:36:14] AaronSchulz: we were just monitoring the Parsoid job insertion rate after editing a template with 16k users, but did not see any spike [21:36:54] are you aware of any general issues with range splitting in HTMLCacheUpdate? [21:37:08] zz_SickPanda: Hoi Yuvi, I did your review, you caused fatal errors everywhere which is just awesome. In return I wonder if you can run through https://gerrit.wikimedia.org/r/68835 a bit - another big commit, on the client-side [21:43:21] New patchset: GWicke; "PATH_INFO-based API actions" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71858 [21:44:40] marktraceur: you need to define the namespaces in your LocalSettings.php. I should've mentioned that. [21:45:57] New review: Yuvipanda; "You need to define the namespaces in your LocalSettings.php" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/70446 [21:46:17] New review: MarkTraceur; "Next patch should address most of the issues." [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/68835 [21:46:33] New patchset: MarkTraceur; "Massive refactor" [mediawiki/extensions/UploadWizard] (master) - https://gerrit.wikimedia.org/r/68835 [21:46:39] Yeees you should've [21:48:52] SickPanda: Next up, http://paste.marktraceur.info/102 [21:49:12] But at least it loads normally now [21:50:14] * SickPanda checks whats up [21:50:24] marktraceur: when are you getting that? [21:50:25] * SickPanda isn't [21:50:43] When visiting Special:UW with a campaign in the query string [21:50:54] A campaign that has existed on the wiki and worked before your patch [21:50:55] marktraceur: does that campaign exist? [21:50:58] Yup [21:51:15] marktraceur: I mentioned that 1. it doesn't do migrations yet 2. gave a sample campaign and how to activate it to test with [21:51:16] Not in the namespace, in the existing database [21:51:25] SickPanda: You can't just break campaigns. [21:51:33] you want an even bigger patch? [21:51:55] this isn't going to be deployed as is. I'll write those as followup patches [21:52:02] I put this in the comments! [21:52:03] Well, no. But there's not much we can do about it - I don't want to break campaigns... [21:52:21] I am not breaking campaigns [21:52:26] I'm just splitting that into another commit! [21:52:34] Well [21:53:04] I'd much rather have that commit ready to get merged at the same time than have the repository in a broken state while you write it... [21:53:49] interesting definitions of broken state, but fine. I'll update it with a maintenance script, and it'll be a huge patch nobody can really understand (squashed from 40+ commits), but I guess it serves some purpose of ideological purity [21:53:57] marktraceur: I'll add that tomorrow. Can you test the rest of the patch? [21:54:06] SickPanda: Sure sure [21:54:07] it requires a maintenance script. [21:54:36] I will do so under protest that this is a grotesque svn like model imposed on us by the horrible monstrosity that is Gerrit [21:54:41] but I'll do it tomorrow :) [21:54:56] SickPanda: You don't need to write it as part of this patch if you don't want to - a new patch would be fine, just so long as master can be pretty much functional *most* of the time [21:55:11] it is functional. it just reads data from a new place [21:55:29] Well, and breaks if you don't update. And breaks in a pretty nasty way. [21:55:31] there needs to be migration done. I don't see why it has to be the same patch. [21:55:43] yes. running master of software is expected to break. [21:55:51] else we'll only have commits on master that go from v1.1 to v1.2 [21:55:53] and so forth [21:55:54] Not usually, with UW [21:55:58] we have 'releases' for a reason [21:56:02] as I said [21:56:15] * marktraceur doesn't understand why this is so heated [21:56:15] you want me to add that to this patch, I'll do that. But again, I think that is a bad bad idea. [21:56:21] atomic commits!Q [21:56:30] SickPanda: I said you don't have to - you can add a new patch with the maintenance script [21:56:37] fine, I'll add a new one then :) [21:57:02] apologies if I sounded edgy. [21:57:08] No, it's no problem [21:57:17] I understand Gerrit can be a frustrating workflow [21:57:27] I don't think the residual gerrit/svn hate will go away, ever. [21:58:16] *shrug* same here re: GitHub, so no hard feelings [21:58:42] heh [21:58:51] that is officially a flame war, so let's not go there. [21:59:47] Exactly. [22:01:30] marktraceur: okay, so new patch for the maintenance script :) moving on... [22:01:37] * SickPanda starts writing that patch [22:01:37] *nod* [22:01:51] SickPanda: Didn't you go to sleep, like, 6 hours ago? Get enough rest? [22:02:10] marktraceur: I had a horrible nightmare, and don't want to go back to sleep right now [22:02:13] Aw. [22:02:26] marktraceur: look at -mobile to read what that is. very wiki related, very scary :| [22:02:35] (and very short) [22:02:55] Wow [22:03:22] yeah [22:04:29] marktraceur: it also had references to marriage, and stuff. The overwhelming feeling was one of hate. I should also stop reading comments on right-wing websites. [22:05:43] *nod* sounds like it [22:06:08] marktraceur: anyway, I'll be around for a bit more :) Still nauseous and feeling terrible, but rolling around in bed with weird visions is worse than flamewarring about Gerrit :P [22:07:07] @notify Elsie [22:07:07] This user is now online in #wikimedia-dev. I'll let you know when they show some activity (talk, etc.) [22:07:44] marktraceur: are you familiar with deployment prep? [22:09:34] aude: Relatively, why? [22:10:09] well, no one is around and http://en.wikipedia.beta.wmflabs.org/ is broken [22:10:33] somehow the sites table got cleared and we want to run a maintenance script to repopulate [22:10:45] mwalker|away, hmm, checking... [22:10:48] but the script won't run for me [22:10:57] * aude is in #wikimedia-labs asking [22:11:20] Oh dear. [22:12:19] it's like a holiday already, i see [22:12:23] most people are gone [22:12:32] mwalker|away, i don't see it [22:14:44] marktraceur: whelp, looks like I was wrong. Going away now, will probably be back shortly but. Thanks for the review so far :) [22:15:08] SickPanda: My pleasure! Be well. [22:21:41] gwicke: not that I know of [22:21:51] Project MobileFrontend-en.m.wikipedia.beta.wmflabs.org-linux-firefox build #51: FAILURE in 1 min 3 sec: https://wmf.ci.cloudbees.com/job/MobileFrontend-en.m.wikipedia.beta.wmflabs.org-linux-firefox/51/ [22:22:05] AaronSchulz: k [22:22:13] gwicke: you can try playing with Title::getBackLinkCache() in eval.php for the template [22:22:18] what template? [22:23:00] unreferenced section [22:23:14] AaronSchulz: unreferences section [22:23:14] [[Template:unreferenced section]] [22:23:59] marktraceur: hey, why is this page empty, along with the history? http://etherpad.wmflabs.org/pad/p/MWCoreReview/timeslider [22:24:25] I, uh [22:24:28] gwicke: enwiki? [22:24:28] I do not know [22:24:33] Is it not supposed to be? [22:24:35] AaronSchulz: I won't get to it before Monday, but will try it then [22:24:48] AaronSchulz: yes [22:24:55] marktraceur: we did a lot of things there yesterday, the important part copied to a wiki for safe keeping, but... i wanted to check something :/ [22:25:07] * marktraceur will endeavour to find it for ye [22:25:14] thanks sir [22:26:40] gwicke: the BacklinkCache functions seem to work fine [22:27:28] Uh [22:27:43] Hm. [22:31:31] alright solved our problem [22:31:49] beta deployment is okay again [22:32:22] how do i log what i did? do we do that? [22:33:29] aude: If you did something in the deployment project in labs, then go to #wikimedia-labs and type !log deployment [22:33:32] greg-g: I have no idea what's going on with this [22:33:41] alright did it and it worked :) [22:35:01] marktraceur: ok [22:35:05] Hm [22:35:12] It looks like it was still there about 12 minutes ago [22:35:20] now I need to use my *brain's* memory? [22:36:11] Or maybe I'm misreading it [22:37:05] * aude notes that http://en.wikipedia.beta.wmflabs.org/wiki/Main_Page has no logo and there is something in the log about logo [22:37:22] * aude not fixing that and it's unrelated to what i did [22:37:43] greg-g: I see the text in the log, would that be helpful? [22:37:47] oh, now i see the logo :) [22:38:22] Hm, it's sort of...JSON encoded [22:41:31] marktraceur: ori-l: Got a minute for https://gerrit.wikimedia.org/r/#/c/70787/ later today? Though there's only 1 dependency now, I'd like to have more interactive @example ones in our documentation. [22:42:22] Krinkle: Maybe? I'm leaving in ~15 minutes so I can fly, but I might be able to review from the airport if ori-l is busy [22:43:11] greg-g: You have PM [22:43:53] safe flight, marktraceur :) [22:43:57] Grazie [22:44:00] Krinkle: it depends what kind of review you want; i read it once and figured it'd take me a week to figure out exactly why you're doing everything you're doing. it looks sane and sensible enough to me, and it's not a critical code paths, so i'm OK with +2ing it [22:44:03] Just a little hop over to Vegas, so [22:44:06] Should be OK [22:44:10] if you want me to actually run it locally it might take me a bit [22:44:22] marktraceur: fourth of july 'weekend'? [22:45:12] SickPanda: Something like that. Family's there. [22:45:17] ah, nice! [22:45:25] * SickPanda goes to try sleep again [22:53:19] New patchset: Bsitu; "Add HTML email support to Echo notification" [mediawiki/extensions/Echo] (master) - https://gerrit.wikimedia.org/r/70112 [22:55:32] can i again ask what Special:RecordImpression does ? [22:56:49] spying [22:56:54] tracking [22:56:54] nsa [23:01:51] Yippie, build fixed! [23:01:52] Project MobileFrontend-en.m.wikipedia.beta.wmflabs.org-linux-firefox build #52: FIXED in 59 sec: https://wmf.ci.cloudbees.com/job/MobileFrontend-en.m.wikipedia.beta.wmflabs.org-linux-firefox/52/ [23:03:14] thedj: I'm not sure what it does, but what sort of request keys are being added to it? I'm wondering what extension that's in and can't seem to find it... [23:03:58] CentralNotice [23:04:23] i think it's registering my response to fundraiser centralnotices or something. [23:04:45] New patchset: Catrope; "Update VisualEditor to master" [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71933 [23:04:47] Sounds plausible [23:05:06] New patchset: Catrope; "Update VisualEditor to master" [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/71934 [23:05:08] "Null endpoint. This is a workaround to simplify analytics." [23:05:11] Change merged: Catrope; [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/71934 [23:05:13] It really has no code to it [23:05:18] Change merged: Catrope; [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71933 [23:24:03] New patchset: Aaron Schulz; "Fixed file purge script to handle NULL archive names" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71937 [23:27:41] New patchset: Catrope; "Update VE to master again to pick up b21fe5f" [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71938 [23:27:51] New patchset: Catrope; "Update VE to master again to pick up b21fe5f" [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/71939 [23:27:58] Change merged: Catrope; [mediawiki/core] (wmf/1.22wmf8) - https://gerrit.wikimedia.org/r/71939 [23:28:06] Change merged: Catrope; [mediawiki/core] (wmf/1.22wmf9) - https://gerrit.wikimedia.org/r/71938 [23:28:17] James_F: ping ping [23:28:48] hoo: Pong. [23:29:31] https://bugzilla.wikimedia.org/show_bug.cgi?id=32216 what information do you exactly like there? Wikitext? Html? Message keys? [23:30:03] I've already got it working, but I'm not sure waht you really need [23:30:18] huh, another old bug getting resolved thanks to VE needing it? [23:30:40] that's like getting money in academia [23:30:45] :P [23:30:55] you'll only get it if it can be considered military [23:31:01] :) [23:32:45] James_F: ^ [23:33:19] MatmaRex: Yeah. :-) [23:33:46] hoo: Whatever AbuseFilter inserts for EditPage.php should instead be in the API (and ideally EditPage.php should be a client of said API, but that's a different bug). [23:34:27] hoo: So in this case the message to display to the user ("You can't edit because $1"). Should be the final (minimal) HTML to display, I'd suggest? [23:35:28] James_F: I guess so... so I'll just give you the html we usually dump into the UI [23:36:54] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71937 [23:44:30] New review: Krinkle; "fixme: Add release notes for deprecations." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/71591 [23:44:37] hoo: [23:44:59] Krinkle: ? [23:46:31] oh, RL [23:46:33] * RN [23:46:48] aude: https://gerrit.wikimedia.org/r/71591 see last comment [23:49:21] hoo: Works for me. :-) [23:49:42] :) [23:51:00] Why isn't 'Security' on the product list under Advanced search? I might not be able to see bugs there by default but I can see some [23:51:12] I have to change the resulting URL to have product=Security instead [23:56:33] hoo: Having aude fix it is fine, but it is your responsibility as reviewer to ensure it (and if needed, fix or revert if there is nobody fixing it). Of course revert for a RN would be a bit exxagerated, but you get the idea :) [23:57:07] Sure... I guess she'll be around tomorrow and will fix it then