[10:34:40] <none>	 does anyone know if it is possible to add lua modules in Module space to categories?
[10:48:21] <gerrit-wm>	 New review: Hashar; "We will be able to catch errors in JJB config whenever it is installed on gallium ( https://bugzilla..." [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/55300
[12:26:24] <hashar>	 mlitn: ping :-]   AFTv5 is broken in beta because of CentralNotice updater :-D
[12:26:33] <hashar>	 mlitn: update.php dies with a stacktrace hehe
[12:27:04] <mlitn>	 I noticed your comment :)
[12:27:12] <hashar>	 and approved https://gerrit.wikimedia.org/r/#/c/55566/
[12:27:22] <hashar>	 mlitn: on merge, a Jenkins job will deploy it automatically on beta
[12:27:27] <hashar>	 I forgot to announce that
[12:27:38] <mlitn>	 ok great :)
[12:27:46] <hashar>	 the job is https://integration.wikimedia.org/ci/job/beta-mediawiki-config-update/
[12:28:44] <hashar>	 Jenkins will report back to Gerrit whenever I get Zuul upgraded
[12:28:54] <hashar>	 with a message like "Successfully deployed on beta" :]
[12:31:24] <mlitn>	 that would be nice :)
[12:40:04] <hashar>	 hmm
[12:40:08] <hashar>	 I hate php
[12:40:14] <hashar>	 mlitn: the fix should be  https://gerrit.wikimedia.org/r/55567
[12:40:30] <hashar>	 the value is used as a callback for call_user_func_array
[12:44:26] <mlitn>	 lol that's a sad issue :)
[12:46:16] <hashar>	 ahh
[12:46:19] <hashar>	 I am not paying attention
[13:00:09] <hashar>	 (meanwhile on labs: I can't use git :D  )
[13:02:44] <^demon|nopower>	 Over ssh? Over https?
[13:02:56] <^demon|nopower>	 Oh, you said, git, not gerrit :)
[13:03:30] <hashar>	 ^demon|nopower: yeah running locally on a glusterfs partition
[13:03:46] <^demon|nopower>	 Oh, gluster.
[13:03:47] <^demon|nopower>	 :)
[13:03:52] <hashar>	 also I noticed Gerrit is quite slow when pushing a refs for review against mediawiki/core
[13:03:58] <hashar>	 it takes a noticeable amount of time
[13:03:59] <^demon|nopower>	 Indeed, it is.
[13:04:06] <hashar>	 repo too big ?
[13:04:11] <^demon|nopower>	 Needs a gc.
[13:04:19] <^demon|nopower>	 Gonna work with Roan to do that safely.
[13:04:26] <hashar>	 ah
[13:04:30] <^demon|nopower>	 (since last time almost ended in disaster)
[13:04:42] <hashar>	 also I was wondering whether we should change the default merge strategy for mw/core
[13:04:51] <hashar>	 it does "merge if necessary" maybe we could use cherry-pick
[13:05:00] <hashar>	 that would save a few merge commit objects
[13:05:09] <hashar>	 not sure if we will still be able to do a merge --no-ff  though
[13:05:42] <^demon|nopower>	 https://gerrit-review.googlesource.com/#/c/41762/ would also help pushes on repos with large #s of refs.
[13:06:01] <^demon|nopower>	 Personally, I think we should move to ff-only.
[13:06:14] <hashar>	 that force us to rebase every single time
[13:06:27] <^demon|nopower>	 It will implicitly rebase before merge.
[13:06:47] <^demon|nopower>	 From a workflow perspective, it's basically the same I believe.
[13:06:48] <hashar>	 na Gerrit tells you the change could not be merged and need a rebase
[13:06:56] <^demon|nopower>	 Well, if it needs a manual rebase.
[13:07:18] <hashar>	 on the other hand, cherry pick just put the patch on top of master and merge that in
[13:07:26] <hashar>	 but then you get a different sha1 in the repo and in the change
[13:07:36] <^demon|nopower>	 Yeah, I don't like that.
[13:09:44] <Nikerabbit>	 :o nopower
[13:14:19] <^demon|nopower>	 Yeah, snowed overnight. Using my phone right now.
[13:15:50] <mlitn>	 does anyone know if it's possible that, on fenari (eval.php or maintenance script) $wgMemc stuff is stored to different Memcached servers or under different keys (wfMemcKey), than what a web request generates?
[13:16:00] <hashar>	 mlitn: yeah
[13:16:15] <hashar>	 mlitn: there is a bug about it.   fenari is in pmtpa so it loads the memcached configuration for pmtpa
[13:16:25] <hashar>	 whereas apaches are in eqiad and have thus a different memcached configuration
[13:16:35] <hashar>	 definitely a bug we have to fix
[13:16:43] <hashar>	 that cause more and more problem
[13:16:49] <^demon|nopower>	 This is a bug with all maintenance scripts.
[13:16:58] <mlitn>	 so right now, there's not really a reason to change memcached values with a maintenance script?
[13:17:10] <hashar>	 that just change it on pmtpa : /
[13:17:16] <mlitn>	 s/reason/way/
[13:17:18] <^demon|nopower>	 Not if you want the change to do anything live :)
[13:17:22] <hashar>	 which also mean we have a split brain
[13:17:28] <mlitn>	 mmh
[13:19:31] <hashar>	 mlitn: mwscript.php/mctest.php does not know about memcache in both datacenters   https://bugzilla.wikimedia.org/show_bug.cgi?id=46428
[13:20:20] <mlitn>	 thanks - I'll have a read ;)
[13:23:30] <hashar>	 moauuhaha
[13:23:39] <hashar>	 Adding index index_page to table aft_feedback ...A database query syntax error has occurred.
[13:23:41] <hashar>	 I am doomed
[13:24:05] <hashar>	 Database returned error "1061: Duplicate key name 'relevance_page' (10.4.0.53)"
[13:24:22] <hashar>	 mlitn: I guess that is for you :-]   applying the patch  sql/index_page.sql got me a dupe key :(
[13:27:43] <mlitn>	 hashar: correct; https://gerrit.wikimedia.org/r/#/c/55574/ will fix this
[13:27:58] <mlitn>	 my bad
[13:28:08] <hashar>	 :-]
[13:28:55] <hashar>	 I am blocked by CentralNotice again
[13:29:18] * CentralNotice  blocks hashar

[13:29:30] <gerrit-wm>	 New review: Zfilipin; "(5 comments)" [qa/browsertests] (master) C: -1;  - https://gerrit.wikimedia.org/r/55431
[13:29:40] <hashar>	 CentralNotice: :-]
[13:30:18] <hashar>	 YuviPanda: I got you a bug report regarding our licenses in the Wikipedia app.  Seems we should mention we are using JQueryy + Phonegap and list their licenses.
[13:30:40] <YuviPanda>	 hashar: yes - we currently list them *in* the app ('about' has a list of credits + licenses)
[13:31:13] <YuviPanda>	 hashar: i'll copy paste it into the wiki app's source repo at some point,
[13:32:46] <hashar>	 YuviPanda: that would be nice :-]  the bug is https://bugzilla.wikimedia.org/show_bug.cgi?id=46532  , reported to me by the maintainer of F-Droid (a repo of android OSS)
[13:32:59] <YuviPanda>	 yes, I do remember F-Droid :)
[13:33:17] <YuviPanda>	 I've wanted to get Commons app setup on that, but the process seemed a bit too involved
[13:33:29] <hashar>	 YuviPanda: they are nice guys.  Kind of like the Debian project but for android.
[13:33:54] <YuviPanda>	 yup! One of these days I want to work on their marketplace android app
[13:33:56] <hashar>	 YuviPanda: you "just" need to submit a pull request for a file named something like   org.wikimedia.commons.txt that contains a few dozen of plain text lines.
[13:33:59] <YuviPanda>	 it could use some love
[13:34:11] <hashar>	 feel free to idlein #fdroid
[13:34:14] <YuviPanda>	 hashar: yeah, I will do that :) Just need to clone, set it up, etc.
[13:34:17] <YuviPanda>	 freenode?
[13:34:21] <hashar>	 yup
[13:35:15] <YuviPanda>	 will do :)
[13:35:35] <hashar>	 aearazeazr
[13:35:45] <hashar>	 I hate our lack of documentation
[13:35:53] <hashar>	 brb
[13:37:56] <YuviPanda>	 hashar: :D
[13:42:55] <hashar>	 stupid central notice patch
[13:42:57] <hashar>	 it is half complete
[13:42:58] <hashar>	 grbmbl
[14:08:38] <hashar>	 I need to relocate to Canada
[14:10:04] <WearyPanda>	 hashar: for testing GeoLocation?
[14:11:36] <hashar>	 WearyPanda: na to get reviewers :-]
[14:11:53] <hashar>	 I could relocate to Quebec (french speaking part of Canada) but that is too cold
[14:12:01] <WearyPanda>	 :D
[14:13:44] <hashar>	 + Quebecian people do not like French people :D  They usually found us to be quite arrogant :-(
[14:14:41] <WearyPanda>	 yeah, I've heard that before
[14:14:58] <WearyPanda>	 I think I've met only one french person ever (you) and didn't find them arrogant.
[14:15:12] <hashar>	 I am not really french :-]
[14:15:16] <hashar>	 well technically I am
[14:15:55] <hashar>	 but I am interacting with so many foreign people that I am barely french anymorehehe
[14:16:40] <^demon|nopower>	 How about New Orleans? They speak cajun french :p
[14:19:05] <WearyPanda>	 HAHA! SOMEONE ELSE WITH NO POWER!
[14:21:24] <hashar>	 ^demon|nopower: the city is too dangerous
[14:21:49] <^demon|nopower>	 WearyPanda: Yeah well, snow.
[14:28:20] <gerrit-wm>	 New patchset: Krinkle; "QUnit: Make builder 'qunit' re-usable." [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/55408
[14:40:01] <gerrit-wm>	 Change abandoned: Krinkle; "Ifef20927889e6af3e7c146fe4" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53534
[14:42:09] <hashar>	 ^demon|nopower: I am getting crazy with a SQL update made by Central Notice :-]
[14:43:06] <^demon|nopower>	 Hmm.
[14:43:14] <^demon|nopower>	 What's the instance & db name?
[14:43:47] <hashar>	 that is on lab
[14:43:53] <hashar>	 I think I will simply revert their patch
[14:44:09] <hashar>	 they changed the schema adding new tables and columns https://gerrit.wikimedia.org/r/#/c/52913/7/CentralNotice.sql
[14:44:36] <hashar>	 and provide a long SQL  patch to bring the new changes
[14:44:47] <hashar>	 which is triggered whenever  cn_controller_mixins table does not exist
[14:44:50] <hashar>	 but it is never created :-]
[14:44:53] <hashar>	 https://gerrit.wikimedia.org/r/#/c/52913/7/patches/CNDatabasePatcher.php,unified
[14:46:51] <hashar>	 I am reverting
[14:52:07] <Krinkle>	 hashar: Can you merge https://gerrit.wikimedia.org/r/#/c/53188/ please ?
[14:52:24] <hashar>	 nop sorry
[14:52:27] <hashar>	 too busy fixing beta
[14:52:28] <hashar>	 :(
[14:53:46] <^demon|nopower>	 Krinkle: I'll take a look.
[14:54:38] <qgil>	 robla, Caltrain cut in Palo Alto due to an accident. I'm at Mountain View station, working.  :)  See you later, hopefully.
[14:55:27] <^demon|nopower>	 Krinkle: +2'd.
[14:57:12] <hashar>	 mlitn: ok after 2 hours of unproductive work on CentralNotice, I am finally hit again by AFTv5 update :-]
[14:59:33] <Krinkle>	 ^demon|nopower: thx
[15:00:19] <^demon|nopower>	 yw
[15:00:53] <mlitn>	 hashar: ^^ let's hope it ends now :)
[15:10:52] <hashar>	 LBFactory_Multi::newExternalLB: Unknown cluster "extension1"
[15:10:57] <hashar>	 I love beta
[15:12:17] <mlitn>	 hashar: that'll need another exception on wmflabs; I'll add that ;)
[15:12:46] <hashar>	 that is from AFTv5 too ?
[15:13:02] <hashar>	 #1 /data/project/apache/common-local/php-master/extensions/ArticleFeedbackv5/ArticleFeedbackv5.backend.LBFactory.php(24): LBFactory_Multi->getExternalLB('extension1')
[15:13:09] <hashar>	 hehe
[15:13:41] <hashar>	 that is a bug in our DB configuration
[15:14:58] <mlitn>	 hashar: https://gerrit.wikimedia.org/r/#/c/55594/
[15:15:11] <mlitn>	 on production, AFTv5 data is on another db
[15:15:21] <mlitn>	 (extension1)
[15:19:35] <hashar>	 but beta should never reference that
[15:19:42] <hashar>	 ahh
[15:19:43] <hashar>	 ok
[15:22:30] <hashar>	 grr
[15:22:37] <hashar>	 need to touch initialiseSettings.php to reset the cache
[15:23:35] <hashar>	 mlitn: that works!   Just add a warning saying  IMPORTANT! Auto-archive is currently disabled. To enable, set $wgArticleFeedbackAutoArchiveEnabled = true.
[15:25:44] <gerrit-wm>	 New review: Cmcmahon; "maintenance" [qa/browsertests] (master); V: 2 C: 2;  - https://gerrit.wikimedia.org/r/55463
[15:25:45] <gerrit-wm>	 Change merged: Cmcmahon; [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/55463
[15:27:17] <^demon>	 qchris: Soo, I got to thinking. Is there anything specific in 2.7-SNAPSHOT we're needing for these plugins? If not, why can't we just build against 2.6-rc0 or 2.6 (final)? They'd be available in the gerrit maven repo.
[15:27:30] <gerrit-wm>	 New patchset: Hashar; "beta: clear config cache on mw-config update" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/55598
[15:27:56] <qchris>	 ^demon No there is nothing specific. It's just that gerrit master is on 2.7-SNAPSHOT
[15:28:03] <^demon>	 Yeah, I knew that.
[15:28:06] <qchris>	 So the plugins' masters should also use that
[15:28:08] <mlitn>	 hashar: that warning is ok; that feature is not supposed to be enabled before maintenance script can manipulate production's caches, so for now it's still off :)
[15:28:20] <^demon>	 qchris: We don't branch/tag plugins (yet) :)
[15:28:23] <qchris>	 Otherwise we cannot build master of plugin against master of gerrit.
[15:28:25] <gerrit-wm>	 Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/55598
[15:28:30] <qchris>	 ^demon: Yes :-(
[15:28:33] <^demon>	 So was just curious how it'd work from $thirdPartyPeople
[15:28:40] <^demon>	 Not that we can't just tag/branch them.
[15:29:38] <qchris>	 Well ... They'd be on 2.5, 2.6-SNAPSHOT, 2.6-rc0, 2.7-SNAPSHOT. For three of them it'll be broken if we do not start branching :-(
[15:30:00] <qchris>	 I am lacking the rights to at version specific branches. Otherwise, I'd just do that.
[15:30:42] <hashar>	 mlitn: the german version seems to give something (though there are 0 feedbacks)  http://de.wikipedia.beta.wmflabs.org/wiki/Spezial:Artikelrückmeldungen_v5
[15:30:54] <hashar>	 mlitn: the english one still complains about not being enabled for this page http://en.wikipedia.beta.wmflabs.org/w/index.php?title=Special:ArticleFeedbackv5
[15:31:03] <^demon>	 I've got it on delete-project, hooks-(its|bugzilla)
[15:31:37] <qchris>	 ^demon: :-) I'd definitely love to see those branches
[15:31:55] <^demon>	 Worth bikeshedding on the list? I'm curious what other peoples' ideas/plans were.
[15:32:33] <qchris>	 Yes. Sure.
[15:33:11] <^demon>	 Mmk, I'll do that.
[15:34:48] <mlitn>	 hashar: it's "disabled" in JS based on the value that https://gerrit.wikimedia.org/r/#/c/55566/ changes
[15:34:54] <mlitn>	 hashar: mw.config.get('wgArticleFeedbackv5Namespaces'); // outputs []
[15:36:45] <mlitn>	 even though this value should now be [0,12,4] ([NS_MAIN, NS_HELP, NS_PROJECT])
[15:36:56] <mlitn>	 any chance some js is still cached?
[15:44:48] <hashar>	 might be
[15:44:54] <hashar>	 and I have no idea how to clear the RL cache
[15:45:19] <hashar>	 ^demon: in MW installer, do we have a way to track the database schema version of an extension?
[15:45:38] <hashar>	 ^demon: I would like to avoid having a patch for each column deleted / removed  :-]
[15:45:50] <^demon>	 Nope...but we don't in core either.
[15:45:58] <^demon>	 We have to do patches for each schema change.
[15:46:38] <hashar>	 I noticed the setAppliedUpdates which add an entry in updatelog
[15:46:47] <hashar>	 I though we could abuse that to find out whether a patch is up to date :-]
[15:49:53] <xyzram>	 ^demon: do you know how a browser URL like:  http://www.wikidata.org/w/index.php?title=Special%3ASearch&profile=default&search=Boston&fulltext=Search
[15:50:02] <xyzram>	 gets converted to a search URL ?
[15:50:11] <hashar>	 mlitn: the crazy thing is that I do see "9 featured comments…" and some feedbacks. Then the page blink and i get the "Article Feedback page not enabled for this page." message.   I guess it is only disabled on the client side ?
[15:50:15] <xyzram>	 Or which PHP files do it ?
[15:50:17] <^demon>	 xyzram: Yes :)
[15:50:48] <xyzram>	 Great, how/where ?
[15:51:05] <^demon>	 SpecialSearch -> SearchEngine subclass (in our case, from MWSearch) -> lsearchd
[15:51:40] <^demon>	 Lemme look at some stuff.
[15:52:30] <xyzram>	 ^demon: I see ApiQuerySearch.run() as starting things off but can't quite see where that is called from.
[15:52:53] <^demon>	 Well, ApiQuerySearch would be from api.php, not index.php
[15:53:03] <xyzram>	 Oh, I see.
[15:53:23] <^demon>	 ApiQuerySearch would be the SpecialSearch in that chain.
[15:54:05] <hashar>	 mlitn: ah without cache it works http://en.wikipedia.beta.wmflabs.org/w/index.php?title=Special:ArticleFeedbackv5&debug=true :]
[16:34:52] <hashar>	 ^demon: I am heading back home. I left a rant on https://bugzilla.wikimedia.org/show_bug.cgi?id=46526 :-]
[16:34:57] * hashar  waves

[16:36:42] <MatmaRex>	 who is Michelle Lee Kosik? she just cc'd herself on like ten bugs i watch
[17:10:32] <Rahul_21>	 YuviPanda, hello
[17:17:36] <Rahul_21>	 YuviPanda, Hello
[17:21:06] <Rahul_21>	 Reedy, Hello
[17:46:46] <YuviPanda>	 hey Rahul_21
[17:46:56] <YuviPanda>	 sorry, in meetings
[17:46:56] <YuviPanda>	 :(
[17:46:56] <YuviPanda>	 so expect high latency
[17:46:57] <YuviPanda>	 wassup?
[17:47:27] <Rahul_21>	 YuviPanda, i wanted to learn how to test a patch'
[17:48:05] <Rahul_21>	 qgil, hello
[17:48:25] <qgil>	 hello Rahul_21
[17:48:49] <YuviPanda>	 Rahul_21: ? like, apply someone else's patch and test it?
[17:50:04] <Rahul_21>	 YuviPanda, i have a change in mind,not sure how it will look,so how do i test it on my local wiki?
[17:50:19] <Rahul_21>	 qgil, its regarding the gsoc proposal
[17:50:48] <Rahul_21>	 qgil, i had a talk with bawolff(he is interestd in mentoring)
[17:51:40] <Rahul_21>	 qgil, the idea is to introduce a system of recording pronounciation in wikitionary
[17:53:45] <qgil>	 Rahul_21, ok, sounds like an interesting proposal. I *think* there is already a feature request somewhere in Bugzilla.
[17:54:37] <qgil>	 Rahul_21, bawolff is also interested in mentoring other projects. He is great but still he  is only one person with 24h/day...
[17:54:59] <qgil>	 Rahul_21, but none of this is a blocker for presenting a proposal in that direction. Let me check in Bugzilla...
[17:55:56] <Rahul_21>	 qgil, i have started working in presenting a proposal in this direction
[17:56:14] <Rahul_21>	 qgil, i hope its a good proposal
[17:59:02] <qgil>	 Rahul_21, https://bugzilla.wikimedia.org/show_bug.cgi?id=31221
[17:59:53] <Rahul_21>	 qgil, yes but i think no1 has undertaken it
[17:59:53] <qgil>	 Rahul_21, but that is not about recording...
[18:00:31] <Rahul_21>	 qgil, can i explain my proposal just in short maybe for you to have an insight
[18:00:48] <qgil>	 Rahul_21, please do it in that bug report
[18:01:24] <qgil>	 Rahul_21, in that bug report there is an objection to save audio files for Wiktionary words.
[18:01:34] <qgil>	 Rahul_21, you will need to proof that your argument is better
[18:01:51] <qgil>	 Rahul_21, I'm not saying that the commenter is right
[18:02:08] <qgil>	 Rahul_21, but we must check that the Wiktionary community would see a point in this proposal
[18:02:24] <qgil>	 Rahul_21, otherwise you would do a bunch of great work that would never make it to wiktionary.org
[18:03:35] <Rahul_21>	 qgil, bawolff told me they requested it :)
[18:03:44] <qgil>	 Rahul_21, bawolff is already CCed in that report
[18:03:46] <Jarry1250>	 Anyone from the VisualEditor team around?
[18:04:52] <qgil>	 Rahul_21, ok, then you can provide a URL of that request (or ask bawolff to do it) and you will have a strong case you won't need to argue.
[18:05:32] <Rahul_21>	 qgil, bawolf just told me that,he didnt give me any links
[18:07:39] <Rahul_21>	 qgil, any way on the irc to retrieve private talks with users?
[18:10:36] <qgil>	 Rahul_21, if you are logging chats i your channel you should be able to find them by opening a window with that user
[18:10:51] <qgil>	 Rahul_21, I mean if you are logging chats in your IRC client
[18:11:18] <Rahul_21>	 qgil, but since hes not online :(
[18:11:29] <qgil>	 Rahul_21, but you could make a better use of your time asking bawolff [citation needed]  :)
[18:12:06] <Rahul_21>	 qgil, hes not online on the irc?is he?
[18:12:12] <qgil>	 Rahul_21, that works even if the other user is not only, as long as you are logging chants in your client
[18:12:30] <qgil>	 Rahul_21, you still have time, no need to find an answer today
[18:13:33] <qgil>	 Rahul_21, if you want to do somethig rigt now you could post your case to https://bugzilla.wikimedia.org/show_bug.cgi?id=31221 and start testing the waters
[18:14:09] <qgil>	 Rahul_21, I will post a message at the English Wiktionary Vila Pump about this Wiktionary proposal and another one we have about Wiktionary API, that might bring more community feedback to the bug report
[18:14:43] <Rahul_21>	 qgil, to quote bawolff
[18:15:41] <Rahul_21>	 qgil, http://lists.wikimedia.org/pipermail/wikitech-l/2013-March/067572.html
[18:16:05] <qgil>	 Rahul_21, actually yeah, don't post in that bug report. Just reading closer.
[18:16:11] <Rahul_21>	 qgil, fresh request and this is what i was talking about
[18:16:43] <qgil>	 Rahul_21, good!
[18:17:53] <Rahul_21>	 qgil, "You should maybe also ask qgill if he thinks its an appropriate length/scope for a project since he's been going through all the project suggestions and generally seems to be in charge"-bawolff
[18:18:21] <Rahul_21>	 qgil, hehe(makes bawolff sound like a man from the literature0
[18:20:42] <qgil>	 Rahul_21, what I will do is to search Bugzilla better to find if there is a request filed. If not I will create one and I will CC you there. We can continue in that report.
[18:21:44] <Rahul_21>	 Rahul_21, sure :),i hope i dont have to change this idea,alot of thinking went into this
[18:23:21] <qgil>	 Rahul_21, the only thing I want to avoid is that you put even more thinking and plenty of work... only to see your feature stalled at the gates of Wiktionary, not accepted by that community for some reason we ignore today
[18:25:37] <gerrit-wm>	 New review: Krinkle; "For some reason when running qunit/phantomjs on a MediaWiki install built by mw-setup-extensions it ..." [integration/jenkins-job-builder-config] (master) C: -1;  - https://gerrit.wikimedia.org/r/55408
[18:26:52] <YuviPanda>	 anyone with UploadWizard knowledge around?
[18:27:35] <Rahul_21>	 qgil, yea,and i was even thinking of a rating extension to accompany this feature,it would involve some 5 people good at linguistics to rate the pronounciation and the pronounciation with the highest rating will be used
[18:28:15] <Rahul_21>	 qgil, This is not the main idea,but thought that it could go well with the pronounciation feature
[18:29:54] <MatmaRex>	 YuviPanda: depends on the issue. i could probably help with anything frontend-related
[18:30:07] <YuviPanda>	 MatmaRex: UploadCampaigns
[18:30:08] <YuviPanda>	 rather
[18:30:13] <MatmaRex>	 YuviPanda: (i don't have much experience with UW, though)
[18:30:14] <MatmaRex>	 ah
[18:30:40] <YuviPanda>	 I'm about to clone the code and comb through it, but was hoping someone can answer questions and save me that :)
[18:31:09] <YuviPanda>	 MatmaRex: so I'm trying to find out what all UC can do, and if it is possible to extend it to let me get 'list of media uploaded via this campaign'
[18:32:31] <MatmaRex>	 YuviPanda: sorry, i have no idea.
[18:32:54] <YuviPanda>	 MatmaRex: :) Thanks for offering to help MatmaRex :)
[18:33:00] * YuviPanda  gives MatmaRex 3rd party cookies

[18:33:24] <Rahul_21>	 YuviPanda, 3rd Party cookies?hahaha
[18:33:30] <MatmaRex>	 heh
[18:35:25] <YuviPanda>	 :)
[18:35:40] <YuviPanda>	 MatmaRex: on an unrelated note, since you talked about frontend stuff
[18:35:48] <YuviPanda>	 MatmaRex: thoughts on sass/less support with RL?
[18:36:19] <YuviPanda>	 (or just a pluggable pre-processor of some sort)
[18:36:49] <Rahul_21>	 YuviPanda, U said you could do small code reviews right?
[18:36:56] <YuviPanda>	 yep
[18:37:04] <YuviPanda>	 for some definition of 'small' :)
[18:37:16] <Rahul_21>	 YuviPanda, https://gerrit.wikimedia.org/r/#/c/55517/
[18:38:32] <YuviPanda>	 Rahul_21: did you test it?
[18:39:04] <Rahul_21>	 YuviPanda, thats what i was pinging you for,i want to learn how to test my own patch
[18:39:17] <YuviPanda>	 well, do you have a local version of mediawiki running?
[18:39:28] <Rahul_21>	 YuviPanda, yes
[18:39:39] <YuviPanda>	 then... reproduce the cause of the bug as given in bugzilla?
[18:39:44] <YuviPanda>	 and try to see if your patch fixes it?
[18:41:25] <Rahul_21>	 local version of mediawiki ,no :(
[18:41:53] <Rahul_21>	 i have installed mediawiki and its like my own personal wiki with random stuff in it
[18:44:15] <YuviPanda>	 Rahul_21: ? how are you commiting patches without a local version of mediawiki cloned?
[18:44:16] * YuviPanda  is confused

[18:45:19] <Rahul_21>	 it is cloned!ok my bad
[18:45:43] <Rahul_21>	 do i have to place this cloned repo in htdocs?
[18:50:24] <mwalker>	 RoanKattouw: ping!
[18:53:12] <YuviPanda>	 Rahul_21: yes, you have to 'install' it
[18:53:17] <YuviPanda>	 check documentation on installing mediawiki
[18:53:23] <mwalker>	 Krinkle: ping!
[18:53:35] <Rahul_21>	 YuviPanda, installed :)
[18:53:35] <AaronSchulz>	 gwicke|lunch: that looked like a nice lunch there
[18:53:42] <YuviPanda>	 :) test it out then
[18:54:15] <Rahul_21>	 YuviPanda, hmmm
[18:54:23] <YuviPanda>	 Nischayn22: around?
[18:54:37] <Krinkle>	 mwalker: yes?
[18:54:47] <mwalker>	 interesting 'problem' for you
[18:54:53] <gwicke|lunch>	 AaronSchulz: this is what I have been reading: http://symbo1ics.com/blog/?p=827
[18:55:38] <mwalker>	 Fundraising just managed to take down the site because apparently with $.ajax({cache: false }) all jsonp calls get appended with a timestamp
[18:56:07] <Krinkle>	 mwalker: ?
[18:56:18] <Krinkle>	 Who was coding without knowing what they're doing?
[18:56:24] <Krinkle>	 j/k
[18:56:31] <Krinkle>	 mwalker: What's the question?
[18:56:42] <mwalker>	 well -- we had caching globally disabled in the CentralNotice controller for a long time
[18:56:46] <mwalker>	 but we just took it out
[18:57:07] <mwalker>	 and now it turns out that other extensions/scripts rely on this setting being sane
[18:57:16] <mwalker>	 so... I can either set the global again in CN
[18:57:28] <mwalker>	 or we can go through and fix everything that's not doing the call correctly
[18:57:35] <mwalker>	 or... we can set it somewhere else globally for MW
[18:57:43] <mwalker>	 so -- the question is: what is the appropriate response?
[18:57:56] <Krinkle>	 mwalker: I don't get what you're asking.
[18:58:05] <Krinkle>	 mwalker: A few quick boolean questions
[18:58:20] <Krinkle>	 Do you want / Is it a problem to have server-side caching?
[18:58:28] <mwalker>	 It is not a problem to have server side caching
[18:59:08] <Krinkle>	 which caching was "globally disabled for a long time"?
[18:59:34] <mwalker>	 sorry -- bad wording -- CN set, globally, $.ajax({cache: true })
[18:59:40] <mwalker>	 we removed this code
[18:59:53] <mwalker>	 so now the default, because nothing else sets it, is cache: false
[19:00:25] <mwalker>	 this is not a problem for CN (I can fix all my calls) -- this is a problem for everything else that assumed cache: true
[19:00:35] <Krinkle>	 Why would (or wouldn't) any CN code use the 'ajax' property in an $.ajax call (whether true or false)
[19:00:54] <Krinkle>	 (not a boolean question)
[19:00:56] <Krinkle>	 :D
[19:01:26] <mwalker>	 we load a script dynamically -- we used to do this by injecting a script tag to the DOM
[19:01:59] <Krinkle>	 and now it is a resourceloader module?
[19:02:05] <Krinkle>	 or an API?
[19:02:15] <Krinkle>	 or something else?
[19:02:30] <Krinkle>	 what kind of url is loaded through that script tag?
[19:02:38] <mwalker>	 it's a psuedoapi -- we're loading the script from a MW special page
[19:02:59] <mwalker>	 and I was going to not inject a script tag; but instead use $.ajax({dataType: "script"})
[19:03:18] <Krinkle>	 OK
[19:04:03] <Krinkle>	 So instead of append(createElement(script).src)=special:cn/cn.js you use $.ajax( special:cn/cn.js, { dataType: script })
[19:04:13] <mwalker>	 exactly
[19:04:18] <Rahul_21>	 YuviPanda, i have my repo in my root folder and i have installed mediawiki in /opt/lampp/
[19:04:28] <YuviPanda>	 'root folder'?
[19:04:39] <mwalker>	 Krinkle: so this solves it for CentralNotice -- but Asher is apparently seeing other calls affected by CN removing the global default for caching
[19:04:44] <Krinkle>	 mwalker: Does the cn/cn.js url serve something else for the same url?
[19:04:59] <Krinkle>	 mwalker: Explain "removing the global default for caching"
[19:05:34] <mwalker>	 1st question: yes; there's about 750,000 different things it can serve -- all of them should be cached :)
[19:05:45] <Krinkle>	 I'm trying to understand how the ajax 'cache'  property is relevant, especially if you're coming from adding script tags
[19:05:58] <mwalker>	 http://stackoverflow.com/questions/7054795/adding-a-script-to-the-page-dynamically-with-jquery-never-uses-the-cached-file
[19:06:06] <mwalker>	 apparently jQuery modifies all those DOM calls
[19:06:43] <Rahul_21>	 YuviPanda, yes
[19:07:06] <Rahul_21>	 YuviPanda, root i meant
[19:07:25] <YuviPanda>	 sorry Rahul_21, I don't think I'll be able to help right now :(
[19:07:33] <Krinkle>	 mwalker: Well, one should never append raw html containing a script tag. Soon (maybe 1.9.1) they will be stripped out entirely. It disabling caching is odd, but then again, it shouldn't happen at all imho.
[19:07:35] <YuviPanda>	 Rahul_21: there's documentation on installing and running mediawiki on mediawiki.org
[19:07:47] <Krinkle>	 I meant 1.9, not 1.9.1
[19:07:51] <YuviPanda>	 Rahul_21: and you should be able to repro the steps on the bug to be able to reproduce teh behavior
[19:08:01] <Krinkle>	 mwalker: OK, but you're no longer appending raw html script tags, right?
[19:08:33] <mwalker>	 Krinkle: we didn't do it that way in the first place, but yes that's true (we did $.('<script>'). ...)
[19:08:43] <Rahul_21>	 YuviPanda, take your time,ill be here till 4 am ist!
[19:09:02] <Krinkle>	 mwalker: "750,000 different things [all] of them should be cached" what do you mean? 750,000 unique urls? Or different things that are stored server-side (in sql or memcached) and serves a one for the user (partially at random)
[19:09:21] <Krinkle>	 how do you prevent being caught by squid and everybody getting the same
[19:09:21] <mwalker>	 750,000 different variances on the URL
[19:09:25] <Krinkle>	 without hitting php every time
[19:09:39] <mwalker>	 it's project=XX&language=XX&country=XX&...
[19:09:43] <Krinkle>	 Sure
[19:09:56] <Krinkle>	 but that's only for that
[19:10:01] <Krinkle>	 that doesn't differentiate the user
[19:10:17] <mwalker>	 the only differentiating we need to do server side is if they're logged in or not
[19:10:29] <mwalker>	 so we have an additional parameter "anonymous=true/false"
[19:10:45] <Krinkle>	 so all logged-out users in a certain country in a certain language on a certain wiki all see the same banner for 30 days?
[19:10:53] <Krinkle>	 (obviously no)
[19:10:58] <mwalker>	 we override caching -- it's 15minutes
[19:11:06] <mwalker>	 but essentially yes
[19:11:07] <Krinkle>	 ok
[19:12:34] <Krinkle>	 mwalker: So previously, when you still appended raw script tags directly, how did you get it to not disable cache?
[19:12:38] <Krinkle>	 the ajaxPrefilter?
[19:13:31] <mwalker>	 nope; we did ( function ( $, mw ) { $.ajax({ cache: true; }); } )(); on initial resourceloader call
[19:13:55] <mwalker>	 which is what I meant by setting it globally
[19:14:20] <Krinkle>	 mwalker: ? That makes an request to the current page
[19:14:31] <Krinkle>	 an http request*
[19:14:41] <mwalker>	 erm; hang on then -- let me look at the code
[19:14:42] <Krinkle>	 You mean $.ajaxSetup ?
[19:15:15] <mwalker>	 probably
[19:15:57] <mwalker>	 yes; we called $.ajaxSetup
[19:16:02] <mwalker>	 https://gerrit.wikimedia.org/r/gitweb?p=mediawiki/extensions/CentralNotice.git;a=blob;f=modules/ext.centralNotice.bannerController/bannerController.js;h=64f03e35c3d3b2f5bcece89c88f67751c56a13c6;hb=f35b1c5479ea721214da6b0007576baa16f6d05e
[19:16:07] <Krinkle>	 mwalker: OK, I'm not sure that works now-adays (ajaxPrefilter might override ajaxSetup(defaultSettings))
[19:16:15] <Krinkle>	 mwalker: anyway, so that worked, right?
[19:16:17] <mwalker>	 yes
[19:16:22] <mwalker>	 until we removed it :)
[19:16:39] <Krinkle>	 mwalker: You removed it *after* switching to $.ajax instead of append('<script>')
[19:16:51] <Krinkle>	 Right?
[19:16:53] <mwalker>	 nope
[19:17:03] <mwalker>	 which is why the site died today when that got deployed to enwiki
[19:17:26] <Krinkle>	 mwalker: Why was it removed?
[19:17:45] <mwalker>	 concerns from Mobile -- they didn't know if it would affect how they did their ajax calls
[19:18:09] <mwalker>	 probably because they, like us, didn't realize exactly wtf that setting did
[19:18:22] <Krinkle>	 mwalker: btw, earlier you suggested something calls $.ajax({ cache: false, url: .... }) is that really correct? Or did you mean jquery internally calling that for $('<script'>)
[19:18:36] <AaronSchulz>	 gwicke: "This would not matter, it would just be more ignorant tripe about Haskell by dim but loquacious bloggers, of which gigabytes are already in circulation" LOL
[19:19:32] <mwalker>	 Krinkle: hang on; let me get my patch up into gerrit
[19:21:06] <mwalker>	 Krinkle: https://gerrit.wikimedia.org/r/#/c/55626/1
[19:22:36] <Krinkle>	 mwalker: OK, that looks good (aside from the empty success: function(){}, which I guess you're going to remove)
[19:23:03] <mwalker>	 yep
[19:23:27] <Krinkle>	 mwalker: So yeah, put "cache: true" back in the default $.ajax settings for the short term and start converting code that relies on it and then move it back out asap.
[19:23:41] <Krinkle>	 but convering it should be fairly high priority because it's quite fragile
[19:23:45] <Krinkle>	 anything can change that
[19:23:55] <Krinkle>	 any module, any extension, any gadget etc.
[19:23:56] <mwalker>	 do we want to do that in CentralNotice; or in some other sitewide js file?
[19:24:06] <Krinkle>	 Wherever it was before should be good for the short term.
[19:24:10] <mwalker>	 kk
[19:24:36] <Krinkle>	 that is has worked so far is extremely lucky.
[19:24:46] <mwalker>	 no kidding!
[19:25:13] <Krinkle>	 mwalker: btw, single quoties :)
[19:25:36] <mwalker>	 good call :)
[19:32:55] <Rahul_21>	 qgil, should i post my proposed solution there?
[19:34:16] <qgil>	 Rahul_21, if you wish. You are a potential GSOC student, your deadlines are still weeks away. But if you start drafting the all this will count
[19:35:02] <gerrit-wm>	 New patchset: Reedy; "Add Thanks extension" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/55654
[19:51:32] <Rahul_21>	 qgil, sure :)
[20:09:45] <mwalker>	 pgehres: want to do a deploy for me this afternoon?
[20:10:04] <Jarry1250>	 Anyone from the VisualEditor team around?
[20:10:05] <pgehres>	 mwalker: you guys should really get a new cluster deployer :-)
[20:10:13] <pgehres>	 mwalker: K4-713 should do one
[20:10:20] <TrevorParscal>	 Jarry1250: hello
[20:10:31] <hashar>	 mwalker: hey :) Sorry for the revert while you were sleeping :/
[20:10:34] <TrevorParscal>	 we hang in #mediawiki-visualeditor
[20:10:35] <mwalker>	 pgehres: she should -- but she's never done one before
[20:10:39] <mwalker>	 hashar: oh; no worries
[20:10:45] <Jarry1250>	 Hi TrevorParsal, I'll join there, thanks
[20:10:54] <hashar>	 mwalker: I did reply to your latest mail, will test your new patchset tomorrow while you are sleeping.  I am busy tonite
[20:10:58] <mwalker>	 hashar: that was the least of CentralNotice's problems today *shakes head*
[20:11:05] <mwalker>	 hashar: cool :)
[20:11:51] <mwalker>	 pgehres: so -- want to mentor K4 to deploy code?
[20:12:01] <pgehres>	 sure :-)
[20:26:25] <Rahul_21>	 qgil, Drafting the all?
[20:28:11] <rachel99>	 chrismcmahon: Hi!  Sorry I didn't get to finish the last code review you sent me.  I was working on it, but the ruby code is all new to me, so it took awhile.
[20:30:03] <chrismcmahon>	 hi rachel99 at some point I thought we might run down the next steps as a pair-programming sort of thing (for some version of "pair")
[20:31:38] <rachel99>	 chrismcmahon: Sure, I would like that.  Right now, I only have about 45 mins time available, but maybe we could start?
[20:32:35] <chrismcmahon>	 rachel99: OK.  It would be great if zeljkof were around also, he's really good at this...
[20:33:12] <rachel99>	 chrismcmahon:  Also, I am stuck on an git issue.  I need to pull down search.feature.  It isn't there though the other files got pulled down.
[20:34:30] <chrismcmahon>	 rachel99: that's odd.  what happens if you do 'git pull' from the /browsertests directory?  or you could maybe clone the /browsertests repo from scratch into a new directory
[20:38:14] <rachel99>	 chrismcmahon: yes, I may clone it.   What happened is I had my own search.feature that I modified and it gave me trouble merging the master to it, so I did git -rm on my search.feature.   But the new one did not pull down for some reason.
[20:38:55] <rachel99>	 chrismcmahon: Would you know how to fix that?
[20:41:35] <chrismcmahon>	 rachel99: so I did a pull on the browsertests repo this morning and I have the search.feature file containing  "Search with accent..." scenarios.
[20:43:17] <chrismcmahon>	 rachel99: I'm not sure what that 'git -rm' did (or if you can reverse it) but you may want to clone from scratch and modify search.feature from the new clone.  hacky, but it should work
[20:43:38] <rachel99>	 chrismcmahon: Ok. Could you give me the clone command again?
[20:44:29] <chrismcmahon>	 rachel99:  git clone ssh://USERNAME@gerrit.wikimedia.org:29418/qa/browsertests
[20:49:52] <rachel99>	 chrismcmahon: That worked, and I have all the feature files now.
[20:51:37] <rachel99>	 chrismcmahon: Do you want to start any of the programming now, or should I contact zeljkof tomorrow morning?  I am available then.
[20:53:04] <rachel99>	 vi
[20:53:22] <Rahul_21>	 chrismcmahon, hi
[20:53:37] <chrismcmahon>	 rachel99: you have a few minutes to get an idea of what we'll do next.  run 'bundle exec cucumber features/search.feature' and watch what Cucumber does with the output for Scenarios that have no _steps.rb file.  Also, try 'bundle exec cucumber features/search.feature:15' to run only the scenario at line 15
[20:53:49] <chrismcmahon>	 hi Rahul_21
[20:54:08] <Rahul_21>	 chrismcmahon, i have a problem with my local wiki,can you help me out
[20:55:16] <chrismcmahon>	 Rahul_21: I'm pretty sure I won
[20:55:38] <chrismcmahon>	 Rahul_21: I'm pretty sure I won't  be able to tell you anything you haven't already heard, but you can ask.
[20:55:51] <Rahul_21>	 A database query syntax error has occurred. This may indicate a bug in the software. The last attempted database query was:
[20:55:52] <Rahul_21>	 (SQL query hidden)
[20:55:52] <Rahul_21>	 from within function "ShortUrlUtils::encodeTitle". Database returned error "1146: Table 'my_wiki.w1_shorturls' doesn't exist (127.0.0.1)"
[20:56:24] <Rahul_21>	 i get ^ this message after i installed smw and hence the database tables got updated
[20:58:15] <chrismcmahon>	 Rahul_21: I don't know much beyond what you can read at http://www.mediawiki.org/wiki/Extension:ShortUrl
[20:59:04] <Rahul_21>	 BadJokePanda, your extension is causing troubles :)
[20:59:18] <BadJokePanda>	 ?
[20:59:18] <BadJokePanda>	 which one?
[20:59:24] <Rahul_21>	 short url
[20:59:34] <BadJokePanda>	 Run the update script which will automatically create the necessary database tables that this extension needs.
[20:59:36] <BadJokePanda>	 did you do that?
[20:59:44] <BadJokePanda>	 https://www.mediawiki.org/wiki/Manual:Update.php
[21:01:51] <Rahul_21>	 i am getting this now A copy of your installation's LocalSettings.php
[21:01:52] <Rahul_21>	 must exist and be readable in the source directory.
[21:02:18] <Rahul_21>	 i have my LocalSettings.php in the source dirsctory
[21:04:09] <Rahul_21>	 BadJokePanda,
[21:04:47] <MatmaRex>	 Rahul_21: so clearly the file is not readable to the update script
[21:05:26] <BadJokePanda>	 Rahul_21: is it readable by the apache user?
[21:05:26] <BadJokePanda>	 and is it the right place?
[21:07:18] <Rahul_21>	 BadJokePanda, yes it is,it is in the sameplace where the default index.php is saved
[21:08:00] <Rahul_21>	 BadJokePanda, and on top of that i have changed the logo and a few other things in my LocalSettings.php which reflect in my wiki properly
[21:08:35] <BadJokePanda>	 well, I suppose that it's readable by apache but not the update script?
[21:19:40] <rachel99>	 chrismcmahon:  I ran both scenarios as you suggested.
[21:21:55] <Rahul_21>	 BadJokePanda, yes:)
[21:23:38] <rachel99>	 chrismcmahon:  For the scenarios undefined in _steps.ruby, it takes the cuke commands and prints out Given and When statements using your scenario, and pending statements for the regular expressions needed.
[21:24:12] <rachel99>	 chrismcmahon: Thats pretty cool.
[21:25:36] <chrismcmahon>	 rachel99:  all the lines that say "pending" will be replaced by code.  you can see existing examples under the step_definitions directory
[21:26:24] <rachel99>	 chrismcmahon: So, we generate the code for those pending lines?
[21:26:48] <chrismcmahon>	 rachel99: we'll write it :)
[21:26:57] <rachel99>	 chrismcmahon: Very cool.
[21:27:42] <rachel99>	 chrismcmahon: I have to go now. Tomorrow morning should be free for me..so I will try to get in touch with Zeljko.
[21:28:36] <chrismcmahon>	 rachel99: about 10AM Mountain/9AM Pacific would probably be good
[21:29:46] <rachel99>	 Ok...will be there..11am Eastern, my time.  Thanks.
[21:37:47] <binasher>	 is anyone here familiar with ActiveUsersPager?
[21:40:46] <^demon>	 binasher: Sup?
[21:41:05] <^demon>	 (I've looked at it before, so that's more than anyone else who's responded ;-))
[21:45:55] <binasher>	 ^demon: i noticed that one of the enwiki slaves was lagging a bunch and getting a bunch of ActiveUsersPager queries.. looks like an uncommonly called page, but the query results in just about a full scan of recentchanges with an on disk temp table
[21:46:20] <binasher>	 takes around 2 minutes on a busy server if all pages are in the bufferpool
[21:46:24] <binasher>	 and 8 if not
[21:46:42] <binasher>	 but i also wonder wtf it's even there for
[21:46:59] <^demon>	 For looking at a list of active users.
[21:47:01] <binasher>	 i mean, it shows users names sorted by user name with a limit 51, not sorted by edits
[21:47:17] <binasher>	 its never going to work for more than seeing a small initial list of active users
[21:47:20] <binasher>	 on enwiki
[21:47:21] <^demon>	 But I've always thought it was kind of silly. The algorithm is rather dumb and arbitrary.
[21:47:29] <binasher>	 and it barely works for that
[21:47:58] <MaxSem>	 disable it the hell out of the sky?
[21:48:04] <^demon>	 So yeah, I'm in favor of that.
[21:48:08] <binasher>	 yes plz!
[21:48:15] <^demon>	 I can see some people being upset and saying "fix it instead so it's not slow."
[21:48:30] <MaxSem>	 some shit can't be fixed
[21:48:34] <^demon>	 Indeed.
[21:48:38] <^demon>	 Some shit never should've happened.
[21:48:47] <binasher>	 that :)
[21:48:47] <MaxSem>	 or it can, but nobody cares...
[21:49:10] <binasher>	 the feature could be implemented differently, but not as an sql query
[21:54:55] <binasher>	 ^demon: if i open a bugzilla ticket, will you kill ActiveUsersPager with fire?
[21:56:36] <ori-l>	 binasher: could be a good use case for redis: use a sorted set (user_id members, timestamp scores); use zremrangebyscore to specify time window
[21:57:15] <ori-l>	 could note that in the bug in case someone feels like taking a stab
[21:58:57] <ori-l>	 well, I guess we don't want to implement core functionality in redis because we don't want to make it prerequisite for running mediawiki.
[21:59:53] <Rahul_21>	 MatmaRex, i tried a few alternatives but still the same problem
[22:02:52] <Rahul_21>	 BadJokePanda, tried a few alternatives,still the same problem
[22:02:55] <binasher>	 ori-l: the integration with ipblocks would complicate a redis reimplementation a bit, but really, i don't think the feature is at all worth the effort
[22:03:02] <BadJokePanda>	 Rahul_21: must be a permission issue
[22:03:11] <BadJokePanda>	 sorry, doubt I can help much - haven't had to deal with mw installs for ages :(
[22:03:16] <BadJokePanda>	 Rahul_21: post on wikitech-l?
[22:03:41] <ori-l>	 binasher: yeah, probably right
[22:03:48] <Rahul_21>	 BadJokePanda, nope il do it now
[22:05:00] <binasher>	 ori-l: recentchanges could be entirely moved to redis though
[22:06:33] <gerrit-wm>	 New patchset: Hashar; "mwext-Survey-testextensions-master now voting" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/55803
[22:06:49] <gerrit-wm>	 Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/55803
[22:06:58] <MaxSem>	 binasher, ^demon: https://gerrit.wikimedia.org/r/55804
[22:07:20] <^demon>	 Heh, was about to push a change to just remove the whole feature :p
[22:07:23] <Reedy>	 ONOUDIDNT
[22:07:50] <MatmaRex>	 lol irc cabal
[22:07:57] <binasher>	 MaxSem: YAY
[22:07:58] <^demon>	 Doesn't isExpensive() already handle $wgMiserMode erroring out?
[22:08:05] <MaxSem>	 meh
[22:08:20] <binasher>	 i don't have to open an bugzilla ticket! wooooo
[22:08:26] <Rahul_21>	 YuviPanda, do review that bug it looks fine
[22:08:48] <YuviPanda>	 I'll, at some point :) a bit swamped now, sadly :(
[22:08:51] <YuviPanda>	 Rahul_21: do mail wikitech!
[22:09:16] <Rahul_21>	 YuviPanda, yes doing that !
[22:09:29] <Reedy>	 ^demon: 	 * Expensive pages are disabled or cached in miser mode.  Originally used
[22:09:40] <MaxSem>	 ^demon, isExpensive() doesn't do the trick
[22:09:49] <MaxSem>	 just checked
[22:09:55] <Reedy>	 Why not?
[22:10:48] <Platonides>	 we should support redirecting special pages to an url
[22:10:52] <MaxSem>	 because isExpensive() controls caching
[22:11:02] <Platonides>	 that way we could redirect them to the toolserver
[22:11:08] <MaxSem>	 and I wanted to blow it outta the sky
[22:11:38] <Reedy>	 git rm
[22:12:11] <binasher>	 wgMiserMode is used for other special pages
[22:12:41] <Reedy>	 Not to disable them..
[22:12:44] <binasher>	 and places like ApiQueryAllImages.php… wow, that exists
[22:13:04] <Reedy>	 		if( $wgMiserMode ) {
[22:13:04] <Reedy>	 			unset( $fields['like'] );
[22:13:04] <Reedy>	 		}
[22:13:22] <Platonides>	 :O
[22:13:48] <MaxSem>	 poor ppl without miser mode...
[22:13:50] <binasher>	                         if ( $wgMiserMode  ) {
[22:13:51] <binasher>	                                 $this->dieUsage( 'MIME search disabled in Miser Mode', 'mimesearchdisabled' );
[22:14:14] <Reedy>	 Yeah, that's not a special page ;)
[22:15:14] <^demon>	 I like my patch better: https://gerrit.wikimedia.org/r/55807
[22:15:33] <binasher>	 oooh
[22:15:45] <binasher>	 real killing with fire :D
[22:15:50] <MaxSem>	 ^demon, release notes
[22:15:55] <^demon>	 Pfft.
[22:15:59] <^demon>	 Nobody cares.
[22:16:17] <Platonides>	 ^demon, I'm sure someone does in wikipedia
[22:16:35] <Reedy>	 This one time
[22:16:38] <Reedy>	 In Wikipedia..
[22:16:53] <Platonides>	 try replacing the special page with "Please leave a message to User:^demon if you want this page to work" :)
[22:17:28] <MatmaRex>	 6 files changed, 0 insertions(+), 277 deletions(-)
[22:17:48] <MaxSem>	 MatmaRex, flawless victory!
[22:18:20] <MatmaRex>	 MaxSem: :D
[22:18:30] <MaxSem>	 Platonides, "Think on the kittens"?
[22:18:34] <^demon>	 Abandoned?
[22:20:50] <^demon>	 Oh, misread the irc.
[22:20:52] <^demon>	 heh
[22:24:52] <binasher>	 did we ever decide on a server in eqiad to run mwscript from?
[22:43:23] <Eloquence>	 Given the issues reported on https://en.wikipedia.org/wiki/Wikipedia:VPT#Errors_with_editing_interface_and_sidebar it may be necessary to pull back enwiki to wmf11
[22:45:12] <greg-g>	 Eloquence: I'm still unable to reproduce in Fx, Chromium, or Epiphany :/
[22:45:24] <Eloquence>	 well, it's clearly widespread enough to be serious
[22:45:32] <Eloquence>	 possibly gadget timing issues involved as well
[22:45:41] <greg-g>	 yeah, that is my guess
[22:46:44] <Platonides>	 @fundraising guys, email received on otrs: «Hey I wanna know if you really are collecting funds or if it's just a message someone put to blackmail»
[22:47:13] <Reedy>	 RoanKattouw: https://en.wikipedia.org/wiki/Wikipedia:VPT#Javascript_Slideshow_Widget_for_Board_Games
[22:47:22] <qgil>	 apergos, thank you for http://www.mediawiki.org/wiki/Mentorship_programs/Possible_projects#XML_dumps - I will add it right away
[22:47:25] <greg-g>	 Reedy: can you verify/take a look at this? https://en.wikipedia.org/wiki/Wikipedia:Village_pump_%28technical%29#CSS_errors_with_search_box.2C_etc..3F
[22:49:18] <Platonides>	 apergos, do you have an otrs account?
[22:56:23] <greg-g>	 Reedy: any luck with the issue reported I linked to above? This might be a rollback-worthy issue.
[22:58:10] <greg-g>	 Susan: can you reproduce that issue a few lines up? ^^
[22:58:22] <Susan>	 Hello.
[22:58:26] <greg-g>	 hi
[22:58:33] <Susan>	 Are you following the imagerotate bug?
[22:58:40] <Susan>	 I'm more concerned about that at the moment.
[22:58:44] <ori-l>	 greg-g: Can't reproduce exactly, but I do see 'Unexpected CSS token: * ' in dev console in Chromium nightly, coming from skins/common/commonPrint.css, line 257. But git-blame says it has been there since 2011-02-24, so maybe it's a red herring.
[22:59:02] <Susan>	 Red is my favorite type of herring.
[22:59:11] <Susan>	 This is only happening on Commons?
[22:59:24] <Susan>	 It looks like some screwy JavaScript.
[22:59:46] <Susan>	 Hmm.
[23:00:15] <Susan>	 That thread seems pretty muddy.
[23:00:25] <Susan>	 Bushranger seems to be having his own set of issues.
[23:01:01] <greg-g>	 https://bugzilla.wikimedia.org/show_bug.cgi?id=46547 ??
[23:01:23] <Susan>	 No, the other one.
[23:01:42] <greg-g>	 https://bugzilla.wikimedia.org/show_bug.cgi?id=46548 ??
[23:01:49] <greg-g>	 :)
[23:01:53] <Susan>	 greg-g: https://bugzilla.wikimedia.org/show_bug.cgi?id=46549
[23:02:11] <Susan>	 The loss part was kind of making me :-/.
[23:02:28] <Susan>	 Commoners, y'know.
[23:03:03] * greg-g  nods

[23:03:09] <Susan>	 Are other people besides Bushranger having issues?
[23:03:51] <greg-g>	 Susan: haven't really confirmed that.
[23:03:56] <Susan>	 When he logged out, he wasn't having this issue, it seems.
[23:04:07] <Susan>	 > The problem appeared to go away after I log out and clear my cache - but it reappeared the moment I logged in, and when I logged out again, it was still there until I cleared my cache, at which point, logged out, the search box worked fine - until I logged in again.
[23:04:16] <greg-g>	 yeah
[23:04:29] <Platonides>	 that may be the source for a "some searches don't work" email I recently replied
[23:04:30] <Susan>	 Right... I'd wait for further reports.
[23:04:39] <Susan>	 https://en.wikipedia.org/wiki/User:The_Bushranger/monobook.js
[23:04:57] <Susan>	 https://en.wikipedia.org/wiki/User:The_Bushranger/monobook.css
[23:04:59] <Susan>	 https://en.wikipedia.org/wiki/User:The_Bushranger/common.js
[23:05:01] <Susan>	 Lots of variables here.
[23:05:41] <Susan>	 What was deployed today?
[23:05:58] <Susan>	 enwiki, ah.
[23:51:00] <AaronSchulz>	 TimStarling: can you look at https://gerrit.wikimedia.org/r/#/c/55002/ ?
[23:52:24] <TimStarling>	 $ok = ( $source === $dest ) ?: copy( $source, $dest );
[23:52:47] <TimStarling>	 I don't think that is on, I'm going to put a comment on that
[23:55:12] <AaronSchulz>	 on?
[23:57:29] <TimStarling>	 that is to say, it's not on
[23:57:57] <TimStarling>	 you know, http://www.urbandictionary.com/define.php?term=not+on
[23:58:26] <Reedy>	 and if it's not on, it's off
[23:59:50] <AaronSchulz>	 what is not on?