[01:11:37] <kaldari>	 robla: test2 isn't listed in groupOverrides in InitializeSettings. Do you know if uses the values from 'test' or does it just not override anything?
[01:12:06] * robla  looks at initializesettings

[01:12:06] <kaldari>	 er, typo... Do you know if it uses the values from 'test' or does it just not override anything?
[01:13:03] <kaldari>	 I'm guessing it just doesn't override anything
[01:14:09] <robla>	 yeah, I'm pretty sure it's the latter
[01:15:04] <kaldari>	 Will anyone object to me ruining the purity of test2 by turning on upload_by_url permissions?
[01:15:38] <robla>	 I won't.
[01:16:05] <kaldari>	 I guess that's good enough :)
[01:16:38] <robla>	 I think test2 wiki just hasn't been around long enough to accumulate a lot of exceptions that other wikis have
[01:22:17] * robla  packs up to leave a little early for an event

[17:53:24] <Reedy>	 Chrome canary is onto v25..
[18:18:37] <gerrit-wm>	 New review: Hashar; "I will review this whenever I get time to take a look at it then merge and deploy.  Probably during ..." [integration/jenkins] (master); V: 0 C: -1;  - https://gerrit.wikimedia.org/r/32836
[18:22:34] <gerrit-wm>	 New review: Jeroen De Dauw; "Why the -1?" [integration/jenkins] (master) C: 0;  - https://gerrit.wikimedia.org/r/32836
[19:38:04] <gerrit-wm>	 New patchset: Stefan.petrea; "    Removed unicode decoding" [analytics/webstatscollector] (master) - https://gerrit.wikimedia.org/r/33403
[19:43:39] <MaxSem>	 Nikerabbit, how do you look at deploying ext:Solarium and switching Translate to it? is there anything tricky about it that would prevent me from doing it myself?
[19:45:11] <gerrit-wm>	 New patchset: Stefan.petrea; "Fixed bugs on deployment, merged" [analytics/udp-filters] (master) - https://gerrit.wikimedia.org/r/33407
[19:49:09] <Nikerabbit>	 MaxSem: you have deployment access?
[19:49:28] <MaxSem>	 yup
[19:50:15] <Reedy>	 Scary, isn't it?
[19:51:07] <MaxSem>	 i was only interested in not breaking Translate accidentally by deploying latest Translate in process
[19:52:06] <MaxSem>	 I know how to deploy new extensions:)
[19:54:20] <thedj>	 i'm guessing this is not intentional ?
[19:54:22] <thedj>	 http://imagebin.org/235867
[19:55:13] <thedj>	 RoanKattouw: ^
[19:55:13] <marktraceur>	 thedj: Known issue
[19:55:14] <MaxSem>	 RoanKattouw, ^^
[19:55:21] <MaxSem>	 oops)
[19:55:27] <marktraceur>	 thedj: Join #mediawiki-parsoid if you'd like to help :)
[19:55:28] <thedj>	 k thx
[19:55:30] <RoanKattouw>	 Yeah that's a bug
[19:55:32] <RoanKattouw>	 I am working on it
[19:55:39] <RoanKattouw>	 It's a result of a bug in VE that's triggering a separate bug in Parsoid
[19:55:47] <RoanKattouw>	 I'm fixing the VE bug and reporting the Parsoid bug
[19:55:59] <marktraceur>	 Suddenly, bugs! Thousands of them!
[19:56:10] <thedj>	 k, just wanted to make sure it was known
[19:56:33] <RoanKattouw>	 I found out about it just minutes before you reported it :)
[20:12:45] <Reedy>	 Krinkle|detached: RoanKattouw
[20:12:46] <Reedy>	      28 Warning:  array_map() [<a href='function.array-map'>function.array-map</a>]: An error occurred while invoking the map callback in /usr/local/apache/common-local
[20:12:47] <Reedy>	 /php-1.21wmf4/includes/resourceloader/ResourceLoaderFileModule.php on line 0
[20:12:47] <Reedy>	      24 Warning:  Illegal offset type in isset or empty in /usr/local/apache/common-local/php-1.21wmf4/includes/resourceloader/ResourceLoader.php on line 378
[20:12:47] <Reedy>	      24 Warning:  Illegal offset type in isset or empty in /usr/local/apache/common-local/php-1.21wmf4/includes/resourceloader/ResourceLoader.php on line 377
[20:12:49] <Reedy>	      20 Warning:  array_map() [<a href='function.array-map'>function.array-map</a>]: An error occurred while invoking the map callback in /usr/local/apache/common-local
[20:12:51] <Reedy>	 /php-1.21wmf4/includes/resourceloader/ResourceLoaderFileModule.php on line 607
[20:12:53] <Reedy>	 bleh
[20:13:00] <Reedy>	 I know the array_map ones is that stupid php bug
[20:13:54] <RoanKattouw>	 Hmm I'll look at the illegal offset ones
[20:13:57] <RoanKattouw>	 But first, lunch
[20:14:42] <Krinkle>	 ResourceLoader.pph:377-279
[20:14:45] <Krinkle>	 	public function getModule( $name ) {
[20:14:45] <Krinkle>	 		if ( !isset( $this->modules[$name] ) ) {
[20:14:45] <Krinkle>	 			if ( !isset( $this->moduleInfos[$name] ) ) {
[20:15:03] <Krinkle>	 php:376-378 *
[20:15:21] <ori-l>	 must mean $name is array or object
[20:16:16] <Krinkle>	 I see a few occurences where getModules() is loped over and calling getModule($name) in the loop
[20:16:21] <Krinkle>	 Reedy: recent regression?
[20:16:40] <Reedy>	 Seems to be newly appearing in 1.21wmf4
[20:17:26] <Reedy>	 the error logs are full of stupid warning spam currently
[20:18:10] <Krinkle>	 Reedy: any particular url to trigger it?
[20:18:17] <Reedy>	 They're not fatals
[20:18:20] <Reedy>	 So we don't get stack traces etc
[20:18:27] <Krinkle>	 k
[20:19:43] <Krinkle>	 I've checked the last 2 weeks of commits in ./includes/resourceloader/
[20:19:48] <Krinkle>	 (commitdiff)
[20:19:52] <Krinkle>	 nothing stands out
[20:20:11] <Krinkle>	 must be a callee problem from outside resourceloader (it is a public function)
[20:20:20] <Reedy>	 Some extension probably
[20:20:45] <Krinkle>	 Reedy: Is it feasible to get a stacktrace on it?
[20:20:55] <Reedy>	 You'd have to live hack in some debugging
[20:20:56] <Krinkle>	 maybe a non-fatal stacktrace hack to get some info.
[20:21:01] <Krinkle>	 right
[20:21:21] <Krinkle>	 I can't reproduce it locally with master and (most) of the wmf extensions
[20:21:48] <Krinkle>	 hm.. maybe on labs?
[20:21:51] <ori-l>	 one of the calls to getModule passes it $row->mr_resource
[20:22:15] <ori-l>	 in messageblobstore (i'm just grepping ,sorry if this isn't useful) i see: array( 'mr_resource' => $modules, (...)
[20:22:20] <ori-l>	 modules sounds suspiciously plural
[20:22:44] <Krinkle>	 but row fields can only be strings or null afaik.
[20:23:07] <Krinkle>	 Reedy: This would be a fine case for labs, right? I mean, it should be reproducible on labs.
[20:23:24] <Krinkle>	 Do we have the same logging infrastructure there yet? (if it is puppetised, it is)
[20:25:37] <ori-l>	 $modulesWithMessages = array();
[20:26:05] <ori-l>	 $res = $dbr->select( 'msg_resource', array( 'mr_resource', 'mr_timestamp' ), array(  'mr_resource' => $modulesWithMessages, ...
[20:26:22] <ori-l>	 $this->getModule( $row->mr_resource )
[20:27:40] <ori-l>	 i don't know that the relationship is between mr_resource in the param and the resulting row, probably a red herring
[20:45:30] <gerrit-wm>	 Change merged: Ottomata; [analytics/webstatscollector] (master) - https://gerrit.wikimedia.org/r/33403
[20:46:03] <Reedy>	 https://gerrit.wikimedia.org/r/#/c/33413/
[20:46:09] <Reedy>	 ^ Very easy review is very easy
[20:47:01] <MaxSem>	 Reedy, done
[20:47:16] <Reedy>	 cheers
[21:02:27] <pgehres>	 Reedy: done with your window?
[21:14:17] <pgehres>	 AaronSchulz: you guys done with wmf4?
[21:14:57] <AaronSchulz>	 I think so
[21:15:35] <pgehres>	 okay, i'm going to go ahead then, let me know otherwise
[21:17:36] <RoanKattouw>	 pgehres: Feel free to run over till 2:30 or so, I'm late with my prep stuff anywya
[21:18:23] <pgehres>	 RoanKattouw: i hope i don't need extra time, will let you know if i go past 2p, thanks :-)
[21:20:42] <bsitu>	 Is job queue running on test2?
[21:21:11] <Reedy>	 it should be
[21:21:27] <Reedy>	 reedy@fenari:/home/wikipedia/common$ mwscript showJobs.php test2wiki
[21:21:27] <Reedy>	 37
[21:21:31] <Reedy>	 little backed up
[21:21:55] <bsitu>	 Some echo jobs are sitting in the queue for quite a while
[21:23:28] <Reedy>	 probably gets a low weighting due to a small queue..
[21:23:32] <Reedy>	 mwscript runJobs.php test2wiki
[21:23:39] <Reedy>	 ^ You can use htat on fenari to force them ot be processed
[21:24:24] <bsitu>	 Reedy: thx
[21:31:41] <pgehres>	 kaldari: can i borrow you for a sec? http://test.wikipedia.org/wiki/Special:FundraiserStatistics
[21:39:27] <pgehres>	 kaldari: they want it out.  so we will fix and redeploy when we can
[21:42:41] <pgehres>	 RoanKattouw: I'm all done for now
[21:42:48] <RoanKattouw>	 Ok thanks
[21:43:39] <pgehres>	 we have a little bug that was a non-blocker.  no idea how long it will take to find, hopefully we can push a quick update once you guys are done
[21:45:32] <RoanKattouw>	 That's fine
[21:45:39] <RoanKattouw>	 I'm still getting Trevor to review my code right now
[21:47:28] <pgehres>	 RoanKattouw: okay, looks like the config that I thought went out, may not have, syncing CmmonSettings real quick
[21:47:36] <RoanKattouw>	 OK go ahead
[21:49:32] <pgehres>	 RoanKattouw: done :-)
[23:09:09] <anomie>	 Anyone around familiar with the QUnit tests?
[23:10:01] <RoanKattouw>	 anomie: I am, somewhat
[23:10:10] <RoanKattouw>	 Krinkle is the expert
[23:10:28] <RoanKattouw>	 lol
[23:10:29] <RoanKattouw>	 gerrit-wm	New patchset: Krinkle; "Fix intermittent QUnit failure" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/33485
[23:10:53] <anomie>	 RoanKattouw- Exactly. Patchset 2 on the change I was looking for someone to review.
[23:11:03] <Krinkle>	 anomie hit the nail right on its head
[23:11:13] <Krinkle>	 that failure was bothering me, I couldn't reproduce it consistenty
[23:11:24] <Krinkle>	 Thanks for that patch.
[23:11:39] <Krinkle>	 As soon as jenkins-bot is ready, I'll merge it.
[23:12:17] <Krinkle>	 getting closer to having mw front-end pass in testswarm...
[23:12:56] <Krinkle>	 RoanKattouw: Remember jquery.textSelection failing in IE6-7? Well, it's back somehow. And has been for several months apparently. Now in IE8 and Opera 12 (just released this month) as well.
[23:13:14] <RoanKattouw>	 Ugh
[23:13:18] <Krinkle>	 Same bug now triggered in Opera.
[23:14:03] <^demon>	 Ugh, Opera makes me angry.
[23:15:19] <Krinkle>	 RoanKattouw: I found out about opera at the hackathon last week. I re-installed TestSwarm on labs, which incidently meant that the browsermatrix was up to date again and now included Opera 12.
[23:15:31] <Krinkle>	 Eh. meant to ping Reedy, as he was there.
[23:16:01] <Krinkle>	 Anyway, we'll see. I hope to get it up and running from Jenkins within a week or two. At least on a manual trigger from Jenkins (not fast enough yet for the main job)