[00:36:18] <ori-l>	 Krinkle: you are so awesome
[00:36:35] <ori-l>	 Re: change set 29724.
[00:37:13] <Krinkle>	 yw!
[00:37:56] <Krinkle>	 I need to re-open my blog, I keep writing long commit messages.
[00:38:10] <Krinkle>	 I closed it a few years ago (or rather, it got lost)
[00:39:45] <ori-l>	 at minimum have something that scrapes the gerrit api for your commit messages and slots them into a blog :P
[01:08:10] <spagewmf>	 kaldari, wfErrorLog( '?' . http_build_query( $event ), $wgClickTrackingLog );
[01:08:37] <spagewmf>	 where http_build_query() just creates the usual encoded params
[01:27:50] <spagewmf>	 kaldari,             wfErrorLog( "Kaldari was here\n" , "/tmp/kaldari.log" ); //  ??  Something must write this stuff
[01:29:26] <kaldari>	 that works, wonder why wfDebugLog() doesn't
[01:29:34] <kaldari>	 oh well
[02:13:33] <spagewmf>	 Krinkle, your  https://gerrit.wikimedia.org/r/#/c/29724/ is pure gold
[02:14:00] <Krinkle>	 thx
[03:42:01] <spagewmf>	 enwiki login didn't ask for a real name a week or so ago.  Now it's asking for a real name.  What changed?
[03:44:10] <Amgine>	 Optional real name has been a part of register an account for a longish time, spagewmf?
[03:46:40] <spagewmf>	 Amgine yes there's a  $wgAllowRealName but it's set to false.
[03:47:35] <Amgine>	 <spuriously wonders if $wgAllowFalseName has been implemented>
[03:47:36] <spagewmf>	 Ahh,  $wgAllowRealName has been deprecated, and in 1.21 was removed.
[04:02:46] <spagewmf>	 but the wgHiddenPrefs global doesn't include realname, so with 1.21 everyone creating an account sees "Real name:" whereas they didn't before.
[05:06:36] <DanielFriesen>	 Yay, my first commit to an ops repo.
[05:06:49] <DanielFriesen>	 spagewmf, since you were on that topic: https://gerrit.wikimedia.org/r/#/c/29741/
[05:49:31] <spagewmf>	 DanielFriesen thanks.  Vs. https://gerrit.wikimedia.org/r/#/c/29742/  :-)
[05:50:13] <ori-l>	 hey TimStarling
[05:50:25] <TimStarling>	 hello
[05:50:52] <DanielFriesen>	 spagewmf, ;) actually your choices are more like mine or both. Because even if it's reintroduced it's till deprecated.
[05:51:50] <ori-l>	 could you look at daniel's change? deprecating it accidentally caused the "real name" input to appear on account creation forms on wikis running 1.21wmf2 (IIRC)
[05:52:33] <ori-l>	 er, "deprecating it" -> removing $wgAllowRealName
[05:53:02] <ori-l>	 culprit: https://gerrit.wikimedia.org/r/#/c/26828/
[05:54:45] <spagewmf>	 TimStarling ,  prior to the change Setup.php would set the wgHiddenPrefs[]  = 'realname' (the right way to do it) based on $wgAllowRealName, which we still set in CommonSettings.php.
[05:55:48] <Brooke>	 [01:55] <gerrit-wm> Change merged: Tim Starling; [operations/mediawiki-config] (master) - https://gerrit.wikimedia.org/r/29741
[06:01:13] <TimStarling>	 I did it
[06:01:55] <TimStarling>	 and apparently also deployed some fundraising change which was merged in 8 hours ago
[06:03:52] <Brooke>	 Thank you. :-)
[06:06:59] <ori-l>	 tim, thanks
[06:07:15] <StevenW>	 +1
[06:07:44] <ori-l>	 we can now all safely revert to our pseudonyms; please address me as MisterSnuggles
[06:07:56] <spagewmf>	 TimStarling, even if I never learn your real name, thx.
[07:51:38] <gerrit-wm>	 New patchset: Hashar; "mw core jobs using the Zuul" [integration/jenkins] (zuul) - https://gerrit.wikimedia.org/r/29746
[07:52:36] <gerrit-wm>	 Change merged: Hashar; [integration/jenkins] (zuul) - https://gerrit.wikimedia.org/r/29746
[09:40:41] <gerrit-wm>	 New patchset: Hashar; ".gitreview file" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29757
[09:41:03] <gerrit-wm>	 New patchset: Hashar; "import layout.yaml from labs" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29758
[09:44:03] <gerrit-wm>	 Change abandoned: Hashar; "already in :-)" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29757
[09:44:27] <gerrit-wm>	 Change abandoned: Hashar; "(no reason)" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29758
[09:45:00] <gerrit-wm>	 New patchset: Hashar; "import layout.yaml from labs" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29759
[09:45:31] <gerrit-wm>	 Change abandoned: Hashar; "bah will force push" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29759
[09:46:29] <gerrit-wm>	 New patchset: Hashar; "testing change creation in gerrit" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29760
[09:46:53] <gerrit-wm>	 Change abandoned: Hashar; "(no reason)" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29760
[09:47:01] <gerrit-wm>	 Change restored: Hashar; "(no reason)" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29759
[09:47:10] <gerrit-wm>	 Change restored: Hashar; "(no reason)" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29757
[09:47:21] <gerrit-wm>	 Change abandoned: Hashar; "(no reason)" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29757
[09:47:35] <gerrit-wm>	 Change abandoned: Hashar; "(no reason)" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29759
[09:48:03] <hashar>	 stupppiiid gerrit
[09:48:03] <hashar>	 :(
[13:38:41] <gerrit-wm>	 New patchset: Stefan.petrea; "Fixed anonymize" [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/29779
[13:56:57] <gerrit-wm>	 New review: Diederik; "Ok." [analytics/udp-filters] (master); V: 1 C: 1;  - https://gerrit.wikimedia.org/r/29779
[14:17:03] <gerrit-wm>	 New patchset: Hashar; "import layout.yaml from labs" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29783
[14:18:27] <gerrit-wm>	 Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/29783
[14:38:36] <Platonides>	 chrismcmahon, the IE problem should be fixed
[14:38:53] <Platonides>	 IE was capitalising the element names
[14:40:46] <chrismcmahon>	 ha!
[14:41:00] <chrismcmahon>	 I don't think I've heard that one before.
[14:41:38] <Platonides>	 html element names are case-insensitive
[14:41:45] <Platonides>	 although the rule was body .gender...
[14:41:59] <Platonides>	 when queried, the interface returned that the rule was BODY .gender-...
[14:42:11] <Platonides>	 a one-letter fix :P
[14:45:59] <chrismcmahon>	 Platonides: spot check on IE6, looks good now
[14:47:08] <Platonides>	 :)
[14:48:17] <Platonides>	 I have got an error report, but without specifying browser, so in NEEDSINFO
[14:49:38] <Platonides>	 btw, the goal of this is for running http://toolserver.org/~platonides/wlmsurvey/wlm2012.php
[14:58:24] <zeljkof>	 this is probably simple, but I just can not find a way how to do it: how do I import existing git repo to empty gerrit repo?
[14:58:49] <zeljkof>	 existing repo is https://github.com/zeljkofilipin/Page-Object-WMF-spike
[14:59:11] <zeljkof>	 empty gerrit repo is https://gerrit.wikimedia.org/r/#/admin/projects/qa/browsertests
[14:59:40] <zeljkof>	 I have cloned existing repo at my machine
[14:59:49] <zeljkof>	 I can clone gerrit repos at my machine
[15:01:32] <zeljkof>	 "git review" says
[15:01:46] <zeljkof>	 $ git review
[15:01:46] <zeljkof>	 Errors running git rebase -i remotes/gerrit/master
[15:01:46] <zeljkof>	 fatal: Needed a single revision
[15:01:46] <zeljkof>	 invalid upstream remotes/gerrit/master
[15:02:17] <zeljkof>	 I have added gerrit as remote to the repo cloned locally
[15:02:20] <zeljkof>	 $ git remote -v
[15:02:20] <zeljkof>	 gerrit	ssh://zfilipin@gerrit.wikimedia.org:29418/qa/browsertests.git (fetch)
[15:02:21] <zeljkof>	 gerrit	ssh://zfilipin@gerrit.wikimedia.org:29418/qa/browsertests.git (push)
[15:08:38] <^demon>	 zeljkof: Just push the existing branch to the gerrit remote.
[15:08:43] <^demon>	 `git push gerrit master` for example.
[15:09:06] <zeljkof>	 ^demon: tried that too :)
[15:09:20] <^demon>	 Probably don't have permissions on it. I wasn't told it had existing history when I made the repo :)
[15:09:36] <zeljkof>	 ^demon: I get this error
[15:09:41] <zeljkof>	 …remote: ERROR:  committer email address cmcmahon@wikimedia.org
[15:09:42] <zeljkof>	 remote: ERROR:  does not match your user account….
[15:10:03] <^demon>	 Yeah, you need the "Forge Committer" permission :)
[15:10:11] <^demon>	 Try now.
[15:10:11] <zeljkof>	 who do I ask for that?
[15:10:19] <zeljkof>	 will do
[15:10:35] <zeljkof>	 the same error
[15:10:52] <^demon>	 Whoops, my fault.
[15:10:54] <^demon>	 Try now.
[15:10:58] <zeljkof>	 ^demon: :)
[15:11:12] <zeljkof>	 whoa!
[15:11:14] <zeljkof>	 ^demon: done!
[15:11:30] <zeljkof>	 ^demon: thanks a lot!
[15:12:00] <^demon>	 You're welcome.
[15:13:32] <zeljkof>	 ^demon: you can remove "forge commiter" permission for me now
[15:13:38] <zeljkof>	 I do not think I will need it in the future
[15:13:52] <zeljkof>	 I plan to follow the usual workflow now
[15:16:17] <chrismcmahon>	 'forge committer' is new to me, ha
[15:18:04] <Platonides>	 chrismcmahon, can you test on Google Chrome 22 on Windows? Preferably 64 bit Windows 7
[15:18:07] <Platonides>	 he reports an uncaught error on User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.4
[15:18:07] <Platonides>	 (KHTML, like Gecko) Chrome/22.0.1229.94 Safari/537.4 (Google Chrome
[15:18:07] <Platonides>	 22)
[15:18:34] <Platonides>	 although works fine for me with Linux Chromium 22.0.1229.94
[15:19:02] <chrismcmahon>	 Platonides: checking...
[15:21:45] <chrismcmahon>	 Platonides: no sign of an error for me with Chrome/Win7 at your /gendertest page
[15:26:43] <Platonides>	 and at http://toolserver.org/~platonides/wlmsurvey/wlm2012.php?userlang=be-tarask ?
[15:26:59] <Platonides>	 (changing the options on question 3)
[15:27:23] <Platonides>	 there shouldn't be a difference, it's the same code, but that's where he saw it
[16:00:34] <Reedy>	 Subscribed to 107 mediawiki repositories
[16:00:34] <Reedy>	 Subscribed to 337 mediawiki repositories
[16:00:38] <Reedy>	 Thanks ^demon
[16:00:39] <Reedy>	 :P
[16:01:23] <^demon>	 Yeah I got that too.
[16:01:25] <^demon>	 Silly github.
[16:07:08] <Krenair>	 Hm. I don't think I did...
[16:10:42] <Reedy>	 I suppose, it's better than 444 seperate emails
[16:12:52] <Krenair>	 How am I supposed to watch all of these for pull requests? :P
[16:15:10] <Reedy>	 STOP CHANGING MEH CODESZZZZ
[16:15:29] <^demon>	 Krenair: You'll get e-mails :)
[16:25:13] <Krenair>	 ^demon, well after I saw Reedy's notes about his 'Subscribed to x MediaWiki repositories' emails I assumed I'd get one to confirm I was watching the repos
[16:25:46] <Krenair>	 But I haven't, and the Github web interface doesn't show me as watching them
[16:25:54] <^demon>	 Hmm.
[18:36:00] <Jeff_Green>	 hashar: around?
[18:36:32] <hashar>	 Jeff_Green: yup
[18:37:15] <Jeff_Green>	 rob'la suggested I talk with you about a problem I'm mucking with re. payments, multiple datacenters, and mw
[18:37:39] <Jeff_Green>	 said you've been working on something very similar for beta labs
[18:37:43] <hashar>	 kind of
[18:38:07] <hashar>	 we have a second sql master in beta which is dedicated to wikivoyage
[18:38:11] <Jeff_Green>	 I'm trying to come up with an elegant way to select different database pools based on site (where site is eqiad vs pmtpa)
[18:38:15] <hashar>	 so we had to setup a database load balancer
[18:38:27] <hashar>	 https://gerrit.wikimedia.org/r/#/c/29344/3/wmf-config/db-wmflabs.php,unified
[18:39:19] <hashar>	 do you want something like  payment1 db to be sent to an eqiad master  and payment2 db to a pmtpa master ?
[18:39:52] <Jeff_Green>	 not exactly
[18:39:55] <Jeff_Green>	 replication is:
[18:40:40] <hashar>	 (note we are on a public channel )
[18:40:47] <hashar>	 ;-]
[18:41:15] <Jeff_Green>	 that's ok. nothing too sensitive here
[18:41:16] <Jeff_Green>	 payments1 is master
[18:41:30] <Jeff_Green>	 pmtpa slaves: payments[2-4]
[18:41:39] <Jeff_Green>	 eqiad slave payments1001
[18:41:53] <Jeff_Green>	 then payments1001 -> payments100[2-4]
[18:42:13] <Jeff_Green>	 all mw instances should use payments1 as the master,
[18:42:34] <Jeff_Green>	 but they should choose from slaves at the same datacenter for everything else
[18:42:56] <Jeff_Green>	 oh, I guess that's what you said
[18:43:01] <hashar>	 kind of :-]
[18:43:31] <Jeff_Green>	 so . . . i was thinking I'd puppetize a simple bit of config to be sourced before choosing the db cluster
[18:44:01] <Jeff_Green>	 but for other reasons I don't want to puppetize all the mw config
[18:44:23] <hashar>	 so
[18:45:42] <hashar>	 you want to use a common $wgLBFactoryConf
[18:46:00] <hashar>	 and based on the value of $cluster, get different 'sectionLoads'
[18:46:09] <hashar>	 assuming a load of 0 is for the master
[18:46:09] <hashar>	 let me read the doc
[18:46:23] <hashar>	 we have $wgDBServers too, not sure it is still used though
[18:46:52] <Jeff_Green>	 i guess more the question is how to set a different $cluster sanely
[18:47:30] <Jeff_Green>	 i.e. it needs to be accessible to mw via apache and for scripts
[18:47:49] <hashar>	 in prod puppet write a file in /etc/wikimedia-cluster
[18:47:52] <hashar>	 grr
[18:48:04] <hashar>	 it echo the value of the cluster in /etc/wikimedia-cluster
[18:48:09] <Jeff_Green>	 ah ha!
[18:49:36] <hashar>	 you also get puppet realm in /etc/wikimedia-realm
[18:49:40] <hashar>	 so that might fit your needs
[18:50:15] <Jeff_Green>	 yup
[18:50:20] <hashar>	 Jeff_Green: are you using the operations/mediawiki-config.git repo for your mw configuration?
[18:50:29] <hashar>	 cause the $cluster is hardcoded to 'pmtpa' there
[18:50:35] <Jeff_Green>	 no, payments has its own puppet instance
[18:50:43] <hashar>	 nice
[18:50:58] <Jeff_Green>	 i morphed realm.pp into something called globals.pp
[18:51:14] <Jeff_Green>	 which sets a whole bunch of global variables based grouped by subnet
[18:51:46] <hashar>	 so the /etc/ files are written from base.pp
[18:51:46] <hashar>	 https://gerrit.wikimedia.org/r/#/c/12377/3/manifests/base.pp,unified
[18:52:22] <hashar>	 hmm
[18:52:25] <hashar>	 it is no more in puppet :(
[18:52:39] <Jeff_Green>	 was it /etc/wikimedia-realm ?
[18:52:40] <hashar>	 yeah
[18:53:03] <Jeff_Green>	 ah it's still there
[18:53:24] <hashar>	 maybe  /etc/wikimedia-cluster was a change in the test branch that did not get merged back in production branch :/
[18:53:43] <hashar>	 not an issue for 'beta' though since we use the realm and then hardcode $cluster
[18:53:53] <Jeff_Green>	 i may make it wikimedia-site since that's how we've named the variable in puppet
[18:54:08] <hashar>	 probably better
[18:54:12] <Jeff_Green>	 so we just source the file from mw wherever we need to know the site?
[18:54:26] <hashar>	 make sure to trim() the value when reading it :-]
[18:54:31] <Jeff_Green>	 ha
[18:54:31] <hashar>	 since there might be a trailing newline
[18:54:38] <Jeff_Green>	 kthx
[18:54:44] <hashar>	 so yeah
[18:54:52] <hashar>	 puppet will provide the value in a file
[18:55:06] <hashar>	 and you can read from it by simply reading from the file (eventually file_exist() first)
[18:55:11] <hashar>	 do make sure to provide a sane default though
[18:55:21] <Jeff_Green>	 ha
[18:55:37] <hashar>	 something like:  $site = 'pmtpa';   // do your stuff with /etc/wikimedia-site  //
[18:55:55] <Jeff_Green>	 right right
[18:56:23] <hashar>	 you could even add a switch( file( '/etc/wikimedia-site' ) ) {   case 'pmtpa': $site=pmtpa;   default:  $site=pmtpa  }
[18:56:35] <hashar>	 aka some kind of whitelisting
[18:56:56] <Jeff_Green>	 yep
[19:27:42] <hashar>	 ^demon: I guess we can merge AaronSchulz  job queue documentation :)     https://gerrit.wikimedia.org/r/29625
[19:27:48] <hashar>	 we can always improve it later if needed :-]
[19:29:07] <hashar>	 thanks :)
[20:01:04] <Krenair>	 <Krenair> Getting this on the latest REL1_20:
[20:01:05] <Krenair>	 <Krenair> from within function �RecentChange::save.� Database returned error �1054: Unknown column 'rc_moved_to_ns' in 'field list' (localhost).�
[20:01:10] <Krenair>	 Krenair> Is this because I usually use master and therefore my database is too up-to-date?
[20:01:31] <^demon>	 Most likely. I think that field was dropped.
[20:02:30] <Krenair>	 Ugh. Guess I'll have to make a new MW instance to test REL1_20 then
[20:06:45] <Krenair>	 tewwy, I saw that you nominated me for Echo +2. Thanks! :)
[20:12:14] <gerrit-wm>	 New patchset: Stefan.petrea; "Temporary branch with LF fixed" [analytics/wikistats] (master) - https://gerrit.wikimedia.org/r/29870
[20:13:44] <gerrit-wm>	 New review: Diederik; "Ok." [analytics/wikistats] (master); V: 1 C: 2;  - https://gerrit.wikimedia.org/r/29870
[20:13:44] <gerrit-wm>	 Change merged: Diederik; [analytics/wikistats] (master) - https://gerrit.wikimedia.org/r/29870
[20:27:24] <gerrit-wm>	 New patchset: Stefan.petrea; "Actual merge for ipad and tests and commandline." [analytics/wikistats] (master) - https://gerrit.wikimedia.org/r/29871
[20:33:29] <gerrit-wm>	 New patchset: Stefan.petrea; "Fixed anonymize" [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/29779
[21:23:53] <ori-l>	 Krinkle: wjat
[21:24:08] <ori-l>	 er, what's the cleanest way of adding setup/teardowns to the qunit test environment?
[21:25:30] <ori-l>	 ah, nevermind -- i see that QUnit.newMwEnvironment takes an environment as a param. cool.
[21:26:40] <sumanah>	 ori-l: :) have you ever heard of the teddy bear rule, or the stapler rule, in tech support?
[21:29:03] <ori-l>	 the teddy bear rule applies when you already have the requisite knowledge for answering the question and are held back by a foggy understanding of the problem.. i didn't have the requisite knowledge. it occurred to me to look at  QUnit.newMwEnvironment after asking, but that wasn't a consequence of me asking the question.
[21:29:32] <ori-l>	 not to put too fine a point on it, heh.
[21:30:01] <Krinkle>	 ori-l: yeah, QUnit.newMwEnvironment takes it, or more precise, QUnit.module takes the setup/teardown which QUnit.newMwEnvironment passes on inside the general mediawiki setup/teardown.
[21:31:04] <sumanah>	 ori-l: I'm sorry, I did not mean to be passive-aggressive
[21:31:08] <ori-l>	 Krinkle: oh, so i don't even need to invoke QUnit.newMwEnvironment myself? i current have module ('foo', QUnit.newMwEnvironment(fooEnv) ), but you're saying module(foo, fooEnv) would work as well?
[21:31:13] <sumanah>	 I literally wanted to share a funny story
[21:31:19] <ori-l>	 sumanah: didn't interpret it as such :)
[21:31:30] <Krinkle>	 ori-l: it does help to put a need/quest/question in words though. Even if you didn't need to ask it, sometimes the brain comes up with great ideas when spitting out what you already know
[21:31:42] <Krinkle>	 ori-l: yes/no
[21:32:21] <Krinkle>	 ori-l: Yes, module needs it  (that's not an option, it is the only entry point where setup/teardown works, module _needs_ to have it for it to be called before/after each test()
[21:32:45] <Krinkle>	 ori-l: No, pass it to QUnit.newMwEnvironment, other wise the default setup/teardown I created for mediawiki won't run.
[21:32:57] <Krinkle>	 e.g. restoring mw.config and what not.
[21:33:10] <Krinkle>	 passing the object to QUnit.newMwEnvironment is like subclassing it.
[21:33:40] <Krinkle>	 pass the return value to module() like module('foo', QUnit.newMwEnvironment({ setup: function () { .. } });
[21:34:16] <Krinkle>	 it will return { setup: function () { /* general mw setup */ your setup(); }, teardown: ... }
[21:34:38] <ori-l>	 right, makes sense
[21:34:57] <ori-l>	 the environment returned by newMwEnvironment calls localEnv.setup() localEnv.teardown() if it gets a local env
[22:28:46] <gerrit-wm>	 New patchset: Stefan.petrea; "Actual merge for ipad and tests and commandline." [analytics/wikistats] (master) - https://gerrit.wikimedia.org/r/29871
[22:58:45] <Nemo_bis>	 O_o https://gerrit.wikimedia.org/r/#/c/29879/
[23:18:18] <gerrit-wm>	 New review: Erik Zachte; "Thanks for cleaning up my dirty line breaks. " [analytics/wikistats] (master); V: 1 C: 2;  - https://gerrit.wikimedia.org/r/29871
[23:18:18] <gerrit-wm>	 Change merged: Erik Zachte; [analytics/wikistats] (master) - https://gerrit.wikimedia.org/r/29871