[07:07:21] <apergos>	 I have an error that is blocking all dump runs:
[07:07:22] <apergos>	 Fatal error: Call to a member function getText() on a non-object in /a/usr/local/apache/common-local/php-1.21wmf8/extensions/MobileFrontend/includes/MobileContext.php on line 273
[07:12:25] <apergos>	 it's unclear to me if that's the only MW error but it's certainly the first one reported.
[07:41:00] <gerrit-wm>	 New patchset: Jeroen De Dauw; "Add Ask and ParserHooks" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/48794
[08:17:03] <hashar>	 morning
[08:38:16] <ori-l>	 hey hashar
[08:38:24] <hashar>	 ori-l: howdy ? :-]
[08:38:39] <hashar>	 ori-l: I need to talk with you while I am in SF :-]
[08:38:55] <hashar>	 about random projects / interests we might have in common
[08:38:56] <ori-l>	 you'll be jetlagged, so we'll be keeping the same hours
[08:39:03] <ori-l>	 i think it'll work :)
[08:39:59] <ori-l>	 when are you coming? i vaguely remember there being some platform/ops pow-wow in the works
[08:40:09] <hashar>	 well I usually wake up at 3am, 4:30am, 6:00am 7:30am
[08:40:28] <hashar>	 then show up at office at like 8:30am .  By 3pm I am no more operational :-]
[08:40:44] <hashar>	 I am coming from Feb 25th till March8th
[08:40:47] <hashar>	 (2 weeks)
[08:41:07] <ori-l>	 cool! yeah, i'll be around
[09:16:49] <hashar_>	 !zuul-config
[09:17:10] <hashar>	 !github
[09:42:20] <yurik>	 ori-l, MaxSem hashar - please vote me for  +2 core :)   http://www.mediawiki.org/wiki/Git/Gerrit_project_ownership#Yurik_for_Core_.2B2   ... off to bed now
[09:44:23] <MaxSem>	 yurik, done
[09:46:06] <yurik>	 MaxSem: thx!
[09:47:01] <yurik>	 MaxSem: you wouldn't mind if i change it to the template (+) with your comment?
[09:47:16] <MaxSem>	 I hate templates
[09:47:24] <MaxSem>	 :P
[09:47:55] <yurik>	 makes it much more visual though
[09:51:54] <gerrit-wm>	 New patchset: Hashar; "regroup mediawiki-core-jslint jobs definition" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/48809
[09:53:02] <gerrit-wm>	 New review: Hashar; "Patch Set 1: Verified+2 Code-Review+2" [integration/zuul-config] (master); V: 2 C: 2;  - https://gerrit.wikimedia.org/r/48809
[09:53:02] <gerrit-wm>	 Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/48809
[11:31:13] <Nemo_bis>	 aude: can you clarify this? https://translatewiki.net/wiki/Thread:Support/About_MediaWiki:Wikibase-error-constraint-violation-aliases/en_%283%29
[11:36:31] <aude>	 Nemo_bis: can take a look later
[11:36:39] <aude>	 thanks for poking me
[11:40:42] <Nemo_bis>	 oki
[14:37:41] <sumanah>	 Nikerabbit: I am grateful for http://laxstrom.name/blag/2013/02/11/fosdem-talk-reflections-23-docs-code-and-community-health-stability/ and am looking forward to the next installment
[15:10:09] <^demon>	 qchris: I saw your changes for renaming :)
[15:27:21] <qchris>	 ^demon: The comments were mostly just nitpicking
[15:27:32] <qchris>	 So I guess we'll get that in sooner or later :-)
[15:27:59] <qchris>	 But the search problems are harder :-( The more I read there, the more way I find that search is broken.
[15:28:13] <qchris>	 Maybe we could define what kind of searches we allow and which we do not?
[15:28:49] <qchris>	 E.g.:
[15:28:52] <qchris>	 status:new OR status:draft OR status:submitted
[15:28:54] <qchris>	 ^works
[15:29:01] <qchris>	 status:new OR status:draft OR status:submitted OR status:submitted
[15:29:04] <qchris>	 ^does not work
[15:29:18] <qchris>	 -status:merged <-- does not work
[15:29:38] <qchris>	 status:open AND status:open <-- does not work
[15:29:42] <qchris>	 and so on :-(((
[15:31:11] <^demon>	 Yeah, it's kind of known that it's limited.
[15:31:31] <^demon>	 It's assumed lucene would probably make this all way nicer, but nobody's ever finished that work.
[15:31:52] <qchris>	 Yeah! Lucene
[15:32:07] <^demon>	 Shawn and I both attempted, but it's a huge project to do right.
[15:32:17] <qchris>	 Just tell me to try it and I'll go for it :-)
[15:32:48] <qchris>	 But I guess there are more pressing things at hand.
[15:33:01] <^demon>	 Yeah, let's stick to the other stuff we've committed to.
[15:59:59] <hashar>	 I am out see you later
[16:00:08] <apergos>	 tah
[16:00:12] <apergos>	 meh
[16:50:00] <Nikerabbit>	 sumanah: should be coming out tomorrow
[16:50:22] <sumanah>	 Rock!
[17:05:18] <Platonides>	 hey sumanah
[17:14:01] <sumanah>	 Hi Platonides
[17:14:21] <sumanah>	 Platonides: can you talk with me in about 80 min?
[17:18:32] <Platonides>	 I had planned to be away at that time :s
[17:18:36] <Platonides>	 I depend on someone else but...
[17:19:00] <Platonides>	 what about an hour later?
[17:19:05] <Platonides>	 I could cancel it if needed, though
[17:22:05] <sumanah>	 An hour later is fine
[17:22:14] <sumanah>	 sorry for the mixup Platonides
[17:24:44] <Platonides>	 seems you're very busy, don't worry :)
[18:59:24] <sumanah>	 hey there Platonides
[19:02:12] <sumanah>	 I'll be around :-) see you in approx 30 min
[19:09:11] <gerrit-wm>	 New review: Krinkle; "Patch Set 6:" [integration/jenkins-job-builder] (master) - https://gerrit.wikimedia.org/r/44836
[19:09:43] <Krinkle>	 ^demon: Hm.. looks like the irc comment thing is confused. It used to show the (shortened version of) the first comment line
[19:09:59] <Krinkle>	 for some reason it now thinks the path set prefix is part of it, so it doens't really show anything of the actual comment.
[19:14:51] <Krinkle>	 not sure how much of gerrit-wm is our own, filed locally for now: https://bugzilla.wikimedia.org/show_bug.cgi?id=44961
[19:19:47] <^demon>	 I feel like a broken record.
[19:19:53] <^demon>	 Yes, we need to fix the hooks.
[19:20:01] <^demon>	 The irc bot is all us, unfortunately.
[19:40:08] <Krinkle>	 hashar: Hey there
[19:40:15] <Krinkle>	 I'm back in the Netherlands.
[19:42:35] <siebrand>	 AaronSchulz: I see  19:21 logmsgbot: aaron synchronized php-1.21wmf9/maintenance/runJobs.php 'deployed b03384f0ebd95e7f79638fb14ccd55da9c186d97'  9 minutes before https://gerrit.wikimedia.org/r/#/c/48853/ is merged. Is there a clock off somewhere?
[19:43:05] <siebrand>	 AaronSchulz: btw, thanks for the quick help.
[19:43:23] <Krinkle>	 hashar: So things are piling up, I could self-merge but I really don't like that. 1) jshint-checkstyle 2) doc generation.
[19:43:26] <AaronSchulz>	 code and git log looks fine
[19:43:31] <Krinkle>	 Can we get that done this week?
[19:43:35] <AaronSchulz>	 must just be weird timing
[19:43:43] <Krinkle>	 Afaik all code is ready and has been tested by both of us.
[19:43:50] <AaronSchulz>	 siebrand: thanks for those extension bugs
[19:44:07] <siebrand>	 AaronSchulz: I didn't check everything there is. Just what I've got checked out locally.
[19:44:17] <AaronSchulz>	 all the core jobs were fine but it looks like extensions were just returning whatever they felt like in many cases
[19:44:22] <hashar>	 Krinkle: hi :-]   well check style could use the upstream patch :D
[19:44:24] <AaronSchulz>	 odd, since the return value isn't really a new thing
[19:44:24] <siebrand>	 AaronSchulz: That should cover all that is deployed on Wikimedia and then some.
[19:44:27] <AaronSchulz>	 I just started using it
[19:44:33] <Platonides>	 sumanah, pong
[19:44:34] <Krinkle>	 hashar: Do you work from the Gerrit dashboard or should I notify you else where? I am sort of assuming that adding your name is enough, and expect that you look at the the dashboad (and or mail notifications) regularly
[19:44:37] <hashar>	 Krinkle: doc generation, it is bugged for now, pending a patch in Zuul
[19:44:55] <siebrand>	 Can someone check the system time on wikitech wiki?
[19:44:56] <Krinkle>	 hashar: Right, we don't need the local merge. I'll update to remote instead.
[19:45:04] <siebrand>	 I think it may be off by many minutes.
[19:45:12] <sumanah>	 hey Platonides
[19:45:14] <sumanah>	 will pm
[19:45:17] <Platonides>	 ok
[19:45:30] <siebrand>	 !log test for timestamp.
[19:46:12] <siebrand>	 Hmm, wikitech is on time.
[19:46:39] <siebrand>	 As is gerrit.
[19:46:57] <siebrand>	 AaronSchulz: Did you deploy that TranslationNotification change outside of Gerrit?
[19:47:15] <AaronSchulz>	 no
[19:48:02] <siebrand>	 AaronSchulz: That's weird then. I'll check again...
[19:48:28] <siebrand>	 AaronSchulz: I see https://gerrit.wikimedia.org/r/#/c/48853/ merged at 18:30 UTC.
[19:48:54] <AaronSchulz>	 fenari and the servers have the right code
[19:48:57] <siebrand>	 AaronSchulz: https://wikitech.wikimedia.org/index.php?title=Server_admin_log&oldid=56406 is 18:21 UTC, right?
[19:49:03] <AaronSchulz>	 (well the servers I checked)
[19:49:51] <siebrand>	 I'm in the stupidest time zone there is at the moment, UTC+5:30.
[19:50:01] <siebrand>	 So calculating time is a drag.
[20:05:57] <hashar>	 siebrand: there are a few timezones with an offset of 45 minutes
[20:06:18] <siebrand>	 hashar: that sounds even worse :(
[20:06:34] <csteipp>	 hashar: really? I'm totally setting up a server there...
[20:06:36] <hashar>	 such as Nepal which is UTC+05:45 , some part of australia is UTC+08:45
[20:06:47] <hashar>	 and the some islands in the Pacific are UTC+14:00 :-]
[20:06:52] <hashar>	 http://en.wikipedia.org/wiki/UTC%2B14:00
[20:07:20] <hashar>	 they could probably be UTC-10,  I guess they get UTC+14 to be one day ahead
[20:08:06] <hashar>	 ahh that is because the country overlap the changing date line
[20:08:19] <hashar>	 that wikipedia website is awesome
[20:08:52] <hashar>	 csteipp: there is an argument to use atomic time instead of UTC.
[20:09:15] <hashar>	 csteipp: since atomic time is not subject to leap seconds. That could prevent servers from hitting a leap second bug in their OS kernel
[20:09:33] <csteipp>	 Yeah, but what would be the fun of that ;)
[20:09:49] <hashar>	 here is the timezone definition: https://gist.github.com/dolmen/3163703
[20:10:11] <hashar>	 I got a nice explanation in french http://www.bortzmeyer.org/leap-second-or-not-leap-second.html
[20:10:23] <hashar>	 anyway the whole point is to get rid of leap seconds
[20:10:32] <hashar>	 % TZ=Etc/TAI date && date -u
[20:10:33] <hashar>	 Fri Jul  6 21:00:16 TAI 2012
[20:10:34] <hashar>	 Fri Jul  6 20:59:41 UTC 2012
[20:12:54] <Denny_WMDE>	 hashar, csteipp - there was also some work by google to basically replace leap seconds with a smearing second
[20:13:06] <Denny_WMDE>	 i.e. make the other seconds a bit longer on that day or hour or sth
[20:13:17] <Denny_WMDE>	 http://googleblog.blogspot.de/2011/09/time-technology-and-leaping-seconds.html
[20:14:46] <csteipp>	 fascinating.. I hadn't heard that one
[20:34:05] <gerrit-wm>	 New patchset: Krinkle; "Implement publisher.checkstyle" [integration/jenkins-job-builder] (master) - https://gerrit.wikimedia.org/r/44836
[20:34:05] <gerrit-wm>	 New patchset: Krinkle; "Fixing Inject to not create empty tags" [integration/jenkins-job-builder] (master) - https://gerrit.wikimedia.org/r/48880
[20:34:05] <gerrit-wm>	 New patchset: Krinkle; "publishers: added option to define groovy postbuild script" [integration/jenkins-job-builder] (master) - https://gerrit.wikimedia.org/r/48881
[20:35:40] <Krenair>	 Anyone around who knows much about database performance? The comment in the middle of this makes me think we can't add this feature with misermode, but maybe it's okay? https://gerrit.wikimedia.org/r/#/c/48661/3/includes/logging/LogPager.php
[20:36:18] <gerrit-wm>	 Change merged: Krinkle; [integration/jenkins-job-builder] (master) - https://gerrit.wikimedia.org/r/44836
[20:36:18] <gerrit-wm>	 Change merged: Krinkle; [integration/jenkins-job-builder] (master) - https://gerrit.wikimedia.org/r/48881
[20:36:18] <gerrit-wm>	 Change merged: Krinkle; [integration/jenkins-job-builder] (master) - https://gerrit.wikimedia.org/r/48880
[20:37:56] <Krinkle>	 hashar: I've force-pushed gerrit/HEAD to the right hash, so that we have matching sha1 hashes again. We didn't have any overrides that weren't upstream already
[20:38:15] <hashar>	 Krinkle: of which repo ? :-D
[20:38:21] <Krinkle>	 hashar: jjb
[20:38:52] * hashar  update it

[20:39:05] <Krinkle>	 pushing them via git-review would have caused merge conflicts for no reason
[20:39:19] <Krinkle>	 hashar: Can you merge/deploy https://gerrit.wikimedia.org/r/#/c/43773/ after you've recompiled jjb?
[20:39:23] * Krinkle  also updating

[20:39:41] <hashar>	 Krinkle: nice to see your support for check style landed in
[20:39:49] <Krinkle>	 yep
[20:40:05] <hashar>	 that does not excuse the fact that the Violations plugin is severely bugged ahah
[20:40:39] <Krinkle>	 yeah
[20:40:49] <hashar>	 oh yeah that huge patch
[20:40:55] <hashar>	 Krinkle: got some minutes ahead to talk about it ?
[20:41:10] <Krinkle>	 sure
[20:41:25] <gerrit-wm>	 New review: Krinkle; "Patch Set 16:" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/43773
[20:42:15] <hashar>	 ahah
[20:42:26] <hashar>	 infinite recursion is fun
[20:42:36] <Krinkle>	 ?
[20:43:09] <hashar>	 reading your comment in https://gerrit.wikimedia.org/r/#/c/43773/16/macro.yaml,unified
[20:44:01] <hashar>	 so here is how I review the stuff
[20:44:09] <hashar>	 I fetch the jjb-config repo under ./config/
[20:44:26] <hashar>	 I create an empty directory "output"
[20:44:35] <hashar>	 then do:
[20:45:02] * Krinkle  (re)wrote https://www.mediawiki.org/wiki/CI/JJB

[20:45:13] <hashar>	 jenkins-jobs test ./config/ -o ./output/   &&    cp -r output output-prev
[20:45:15] <Krinkle>	 I've done it once to test the Checkstyle thing in production
[20:45:21] <hashar>	 then I fetch the patchset under config
[20:45:24] <hashar>	 rerun jjjb
[20:45:33] <hashar>	 and do a diff of the two output dirs:    diff -u -r output-prev output
[20:45:46] <hashar>	 I will make that a jenkins job wheneverI find out a way to deploy jjb on gallium
[20:45:51] <Krinkle>	 I have the same setup, except that I have jjb and jjb-config both in the same directory and use a symlink instead (so that I have them all in root ~/Development)
[20:46:06] <hashar>	 Krinkle: ho man, you have put shortcuts on mw.org  NICE!!!
[20:46:14] <Krinkle>	 :)
[20:46:38] <csteipp>	 Hey, can get an easy +2 https://gerrit.wikimedia.org/r/#/c/48070/ ?
[20:47:02] <hashar>	 Krinkle: I should probably have initiated  a hangout to lead you in JJB arcane
[20:47:36] <Krinkle>	 csteipp: Is there consensus? A bug ticket?
[20:48:06] <hashar>	 Krinkle: I will review the CI/JJB article tomorrow
[20:48:18] <Krinkle>	 e.g. that this extension featuer has been throughouly reviewed for production deployment and that we want it deployed beyond test wiki (namely meta.wikimedia and medaiwikilorg)
[20:48:42] <csteipp>	 Krinkle: Reviewed by Tim about 6 months ago
[20:49:00] <csteipp>	 And the stewards have all been talked to about it
[20:49:25] <Krinkle>	 Well, the change doesn't mention, refer or proof any of that.
[20:49:33] <Krinkle>	 But someone else might feel lucky.
[20:50:50] <hashar>	 Krinkle: I had to split the lint and jshint tasks in two different jobs :/  https://gerrit.wikimedia.org/r/#/c/48422/
[20:51:13] <hashar>	 Krinkle: the reason is jshint does not pass on REL1_19, so whenever the branch is REL1_19 I am not triggering the jshint job
[20:54:32] <gerrit-wm>	 New patchset: Hashar; "JSHint: run from bash and with checkstyle magic." [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/43773
[20:55:07] <Krinkle>	 hashar: The blocking bug for doc, is that this one? https://gerrit.wikimedia.org/r/#/c/44384/1/layout.yaml
[20:55:10] <gerrit-wm>	 New review: Hashar; "Patch Set 17:" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/43773
[20:55:16] <Krinkle>	 I'm not sure I follow
[20:55:32] <Krinkle>	 are you saying there is no way to run a job after a merge completes?
[20:55:36] <hashar>	 Krinkle: na it is an issue in Zuul itself.  I haven't started poking it at it.
[20:55:48] <Krinkle>	 post vs. postmerge, what's that about
[20:56:08] <Krinkle>	 Is there a bug report?
[20:56:25] <hashar>	 Krinkle: I got an issue with 'post' jobs not including a branch which is used by the Git Plugin as a branch specifier.   I will either fix zuul or find a workaround :-)
[20:56:53] <hashar>	 Krinkle: anyway regarding your change  https://gerrit.wikimedia.org/r/#/c/43773/17/macro.yaml,unified
[20:57:06] <hashar>	 Krinkle: any reason to have a jshint-in macro that takes a dir as a parameter ?
[20:57:18] <hashar>	 Krinkle: could probably be named "jshint"  and have the shell part  cd "$WORKSPACE"
[20:57:30] <hashar>	 Krinkle: the change itself is fine though, just wondering
[20:57:30] <Krinkle>	 hashar: tell me you didn't rebase and make changes at the same time again
[20:57:38] <hashar>	 again ?
[20:57:39] <hashar>	 wtf?
[20:57:41] <hashar>	 :D
[20:57:44] <Krinkle>	 :/
[20:57:54] <hashar>	 I just rebased it
[20:57:57] <Krinkle>	 But I got a cli trick now, thanks to roan
[20:57:58] <Krinkle>	 Hm..
[20:58:08] <Krinkle>	 Wait, it *is* a rebase.
[20:58:10] <Krinkle>	 Great
[20:58:11] <hashar>	 yeah
[20:58:16] <hashar>	 I usually do the rebase first
[20:58:20] <Krinkle>	 sure
[20:58:22] <hashar>	 then submit another patchset to fix the issue
[20:58:26] <Krinkle>	 did it have conflicts?
[20:58:27] <hashar>	 unless that is a basic conflict
[20:58:30] <hashar>	 na no conflict
[20:58:37] <Krinkle>	 since the rebase is on your name, not Gerrit
[20:58:53] <hashar>	 that basically rebase it to a point that include https://gerrit.wikimedia.org/r/#/c/48422/  (which separated  JSHint to  -jslint jobs)
[20:59:03] <hashar>	 ah I did the rebase manually
[20:59:03] <hashar>	 l
[20:59:11] <hashar>	 git rebase master && git-review -R
[20:59:20] <hashar>	 faster for me than switching to the browser and clicking rebase
[20:59:37] <hashar>	 (since if there is a conflict I don't have the alert box and can fix it right away)
[21:00:38] <hashar>	 so hmm
[21:00:50] <hashar>	 jshint-in with a dir parameter is nice
[21:01:19] <hashar>	 but  jshint  and   cd "$WORKSPACE"  would be simpler
[21:01:21] <hashar>	 not a big deal though
[21:02:37] <hashar>	 Krinkle: whenever you are around, I will +2 and get you to deploy the change :-]
[21:02:49] <Krinkle>	 ok
[21:02:54] <Krinkle>	 which change
[21:03:28] <Krinkle>	 we don't need to separate lint/jshint before deploying jshint-checkstyle. It is working fine without the separation. I've been waiting for a month now.
[21:03:53] <gerrit-wm>	 New review: Hashar; "Patch Set 17: Code-Review+1" [integration/jenkins-job-builder-config] (master) C: 1;  - https://gerrit.wikimedia.org/r/43773
[21:04:18] <hashar>	 Krinkle: then the -lint job running both php lint / jshint will choke when being run on mediawiki/core@REL1_19
[21:04:36] <hashar>	 by separating jshint to its own job, I could make it ignore the REL1_19 branch
[21:05:27] <hashar>	   - name: mediawiki-core-jslint
[21:05:27] <hashar>	     voting: true
[21:05:28] <hashar>	   # (bug 44846) JSHint is always wrong in MediaWiki REL1_19
[21:05:29] <hashar>	     branch: (?!REL1_19)
[21:05:30] <Krinkle>	 hashar: that's been that way for almost a year
[21:05:31] <hashar>	 (in zuul-config )
[21:05:32] <Krinkle>	 and you self-merged again
[21:05:34] <Krinkle>	 I can't work this way
[21:05:38] <Krinkle>	 waiting for a month when I need something, and you have something and just push it and move on
[21:06:03] <hashar>	 you know that CI is my core responsibility and I have to fix the stuff right away ?
[21:06:08] <hashar>	 that was blocking merges in 1.19
[21:06:17] <hashar>	 so I repair right away
[21:06:17] <Krinkle>	 Yes
[21:06:29] <hashar>	 I wish we could talk more often but given our timezone diff that is not really idealy
[21:06:29] <hashar>	 ideal
[21:06:32] <Krinkle>	 But was it happening right now? And couldn't it wawit 10 minutes for a review?
[21:07:07] <hashar>	 anyway
[21:07:15] <Krinkle>	 I watch my gerrit dashboard regularly, if you push something I can review within a day, usually earlier.
[21:08:01] <hashar>	 anyway it is getting late   I would like to get the check style merged in / deployed
[21:08:06] <Krinkle>	 k
[21:08:15] <hashar>	 so the jshint-in builder accept a dir parameter
[21:08:37] <hashar>	 the macro could be changing to the workspace directly   cd "$WORKSPACE"
[21:08:40] <hashar>	 but that is not a big deal
[21:08:44] <hashar>	 I was just curious :-]
[21:09:08] <hashar>	 (note that cd is not even needed since the shell is launched in the workspace already)
[21:09:56] <Krinkle>	 hashar: I was following the same pattern as the other macro, iirc you told me to do it this way. I had a plain jshint macro at first and converted to jshint-in due to different directories for extensions
[21:10:15] <hashar>	 yeah that is what I thought :-]
[21:10:40] <hashar>	 lets deploy it
[21:10:48] <hashar>	 do you have a jenkins_jobs.ini  file for production?
[21:10:50] <Krinkle>	 hashar: anyway, I understand your problem, but you understand my frustration a bit when I have to wait weeks/months to get something deployed?
[21:11:00] <hashar>	 Krinkle: yeah I know about it
[21:11:03] <Krinkle>	 hashar: I think so, yes.
[21:11:10] <hashar>	 Krinkle: I got the same issue with ops from time to time :-]  but lets deploy !
[21:11:14] <Krinkle>	 Got the token from the jenkins profile of jjb bot
[21:11:22] <Krinkle>	 mine didn't work
[21:11:43] <hashar>	 the one at https://integration.mediawiki.org/ci/user/krinkle/configure  then  [Show API token]
[21:12:01] <hashar>	 user is either krinkle or Krinkle I guess
[21:12:06] <hashar>	 should be the labs username
[21:12:08] <Krinkle>	 hashar: I know where to get it
[21:12:13] <hashar>	 hehe
[21:12:17] <Krinkle>	 That's how I got the one for the jjb account
[21:12:20] <hashar>	 making sure I know you know about it
[21:12:22] <Krinkle>	 I;m saying it doesn't work under my name
[21:12:22] <hashar>	 ahh
[21:12:34] <hashar>	 so just use  jobbuilder-bot username
[21:12:43] <Krinkle>	 I can get an API token for my account, but using it rejects the api request
[21:12:46] <Krinkle>	 yep
[21:12:49] <hashar>	 that is apparently the user I am using :/
[21:12:55] <Krinkle>	 that's how I deployed jshint-checkstyle last month
[21:13:19] <Krinkle>	 which was then undone after I verified by submitting master again
[21:13:36] <hashar>	 ahh
[21:13:43] <hashar>	 so what I usually do:
[21:13:47] <hashar>	 I send the patch in Gerrit
[21:13:59] <hashar>	 update one of the production job
[21:14:01] <hashar>	 test it out
[21:14:12] <hashar>	 if that fail, amend && send to Gerrit && update job && test
[21:14:17] <hashar>	 rinse until the job works fine
[21:14:31] <hashar>	 then I self merge the change notifying something like "deployed"
[21:14:44] <hashar>	 so the git repo is essentially used as a log of the actions
[21:15:04] <hashar>	 Krinkle: I think what will work the best is that you self merge / deploy yourself
[21:15:17] <hashar>	 Krinkle: you can alway revert back to a previous working state if there is a trouble
[21:15:29] <hashar>	 and I can do the same by just looking at the git log :-]
[21:16:24] <hashar>	 note that your change modify ALL the jslint jobs. So you might want to only update one of the extension
[21:16:34] * hashar  should setup  test/mediawiki 

[21:17:12] <Krinkle>	 hashar: well, the way I do it: send to gerrit, run jjb locally for one job to assert that it works, make a test commit etc., the undo the local jjb submission in production and wait for you to merge and then you or I will deploy permanently
[21:17:32] <hashar>	 Krinkle: yeah
[21:17:42] <hashar>	 Krinkle: so to make it faster:  be bold and merge!
[21:17:43] <hashar>	 -:]
[21:17:53] <hashar>	 it is probably useless to wait for me to review
[21:17:58] <hashar>	 if you know it is working, just merge / deploy
[21:18:16] <hashar>	 that makes everything faster in this context.
[21:18:27] <hashar>	 the setup is pretty simple to debug and only us do modifications to it
[21:18:41] <hashar>	 that will get you out of despair
[21:18:52] <hashar>	 (and I am sorry to have made you wait for it for so long)
[21:20:09] <Krinkle>	 hashar: well, I'm sorry but I can't do that. I'd like your view on it. Even if I would know jjb very well (I'm getting better at it now that I've made a plugin and got it reviewed upstream), I'd still like your view before it goes viral.
[21:20:26] <gerrit-wm>	 New review: Hashar; "Patch Set 17: Verified+2" [integration/jenkins-job-builder-config] (master); V: 2 - https://gerrit.wikimedia.org/r/43773
[21:20:48] <hashar>	 Krinkle: I don't mind having it viral
[21:21:05] <hashar>	 Krinkle: virii are fun, they tend to spread out :-] too many control and we are just slowing done.
[21:21:17] <hashar>	 hoesntly I have the same level of knowledge regarding jjb as you
[21:21:27] <hashar>	 the tricky parts I do them in labs, that is the only difference probably
[21:21:33] <hashar>	 but yeah
[21:21:46] <hashar>	 I will make sure I respond to your future review requests in a timely manner
[21:21:55] <hashar>	 FYI, I read all Gerrit  notification on a daily basis
[21:22:06] <hashar>	 and do read the dashboard from time to time
[21:22:58] <hashar>	 I just have a biit too many incoming and outgoing reviews :(
[21:23:21] <Krinkle>	 I know the feeling, there's probably a few commit authors feeling the same way about mine.
[21:25:17] <hashar>	 so your jslint change, go ahead and deploy it :-]
[21:25:32] <hashar>	 if that screw up somehow feel free to mail me about it and I will help you find out the issue
[21:25:37] <hashar>	 meantime that is bed time for me :-D
[21:25:52] <hashar>	 (you know the story: wife, daughter waking up at 7am …)
[21:26:06] <hashar>	 again sorry for the review delay :/
[21:52:46] <gerrit-wm>	 New review: Krinkle; "Patch Set 17: Verified+2 Code-Review+2" [integration/jenkins-job-builder-config] (master); V: 2 C: 2;  - https://gerrit.wikimedia.org/r/43773
[21:52:46] <gerrit-wm>	 Change merged: Krinkle; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/43773
[22:24:10] <TimStarling>	 anomie: it wasn't possible to do "":usub() before my changes, I tested it
[22:41:32] <mwalker>	 marktraceur: you wrote a little blurb on writing an extension that pulls in code from git/sourcerepo (https://www.mediawiki.org/wiki/Mentorship_programs/Possible_projects#Write_an_extension_for_pulling_files_from_a_git_repository)
[22:41:36] <mwalker>	 did anything ever happen with that?
[22:41:48] <marktraceur>	 mwalker: As a matter of fact, someone is working on it right now
[22:41:55] <Reedy>	 Isn't someone working on something like that now?
[22:42:19] <marktraceur>	 mwalker: She's hanging out in #mediawiki, if you want to chat with her
[22:42:19] <mwalker>	 cool! where can I get more details?
[22:42:26] <mwalker>	 name?
[22:42:29] <marktraceur>	 mwalker: It's called Git2Pages, it's in Gerrit
[22:42:32] <Reedy>	 Git2Pages in our git repo ;)
[22:42:39] <marktraceur>	 mwalker: The intern's nick is terrrydactyl
[22:42:50] <marktraceur>	 And Dereckson is mentoring her
[22:42:53] <mwalker>	 awesomely possumely
[22:52:08] <anomie>	 TimStarling- You need parens, (""):usub(). But it works for me locally on both LuaStandalone and LuaSandbox if I check out 2d9e3c74ba0b11fa84e5869af5d80ec89e814b90
[23:02:29] <TimStarling>	 hmm, I wonder what I tested then