[00:00:24] New review: Krinkle; "(1 comment)" [integration/jenkins-job-builder-config] (master) C: -1; - https://gerrit.wikimedia.org/r/53534 [00:19:17] TimStarling: anomie|away [00:19:25] 41 Exception from line 285 of /usr/local/apache/common-local/php-1.21wmf11/includes/Hooks.php: Detected bug in an extension! Hook ScribuntoHooks::showDocSubpageHeader failed to return a value; should return true to continue hook processing or false to abort. [00:21:56] https://bugzilla.wikimedia.org/show_bug.cgi?id=46092 [00:22:30] i guess as the code does return true at the end, it's not getting that far in the code [00:37:54] hello [00:46:17] Reedy, how could it not be reached? [00:46:43] the only reason would be that it throwed an exception [00:46:49] but that would appear differently on the error log [00:47:09] my only guess is that it's related to the rename from showDocSubpageHeader to showDocPageHeader [00:47:13] somehow with APC [00:47:41] If it threw and exception, it wouldn't reach the return true [00:49:45] it wouldn't reach the throw at Hooks.php, either [00:55:31] Reedy- Yeah, that error makes no sense [00:55:37] Platonides- That change isn't deployed yet [00:56:11] public static function showDocPageHeader( &$article, &$outputDone, &$pcache ) { [00:56:16] in the stuff on fenari [00:57:43] They seem to be dropping away slowly [00:57:49] Did someone deploy that? [00:58:00] Chad did earlier, I think [00:58:00] hmm, yeah, I guess they did. [00:58:06] ok then [00:58:55] But then it complaining about showDocSubpageHeader *really* doesn't make any sense, since that function doesn't even exist anymore. [00:59:06] Is it half deployed somewhere? [00:59:16] wouldn't be the first time [00:59:19] one way to try and confirm [01:02:24] From #wikimedia-operations: [01:02:25] 23:17: !log demon Started syncing Wikimedia installation... [01:02:25] 23:23: !log no one sync apaches, we wre doing crazy magic on them [01:02:25] 23:41: !log demon Finished syncing Wikimedia installation... [01:02:25] That's probably not good... [01:03:04] They were messing with certs/apache config [01:03:07] not mw stuff [01:08:28] ... Bah. First occurrence of the error in the logs is at 22:44:34, last is at 22:44:42. It probably was half deployed, for 8 seconds during the sync-dir. [01:13:33] lol [02:01:32] !g I0259a6e8b68b8b35bcc1e12c5426f558c5af2097 [02:01:32] https://gerrit.wikimedia.org/r/#q,I0259a6e8b68b8b35bcc1e12c5426f558c5af2097,n,z [02:19:38] New patchset: Hashar; "pep8 & pyflakes now vote for ops/debs/ircecho" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/53715 [02:19:59] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/53715 [02:43:47] hashar: ^ woot, nice! [02:47:47] :-] [02:47:57] ori-l: moooore pep8 :-] [02:48:11] ori-l: python tox is not available in Precise :( [02:49:34] hm, well, I think the essential thing is just to run the test suites, period [02:49:38] (for those repos that have them) [02:49:59] there is nothing about the tests that makes them dependent on tox; tox is simply a way to run the tests against multiple versions of the python interpreter [02:55:20] ori-l: I can't remember wich repo you wanted tests for [02:55:59] eventlogging, but it's not too important [03:10:27] git version 1.8.2 !! [03:26:54] https://github.com/git/git/blob/master/Documentation/RelNotes/1.8.2.txt [03:27:02] hashar: Thanks [03:27:13] yeah reading that right now [03:27:31] "git commit" can be told to use --cleanup=whitespace [03:27:39] seems to be automatic whitespace removal [03:27:43] running `brew upgrade` though that brought me from 1.8.1.3 to 1.8.1.5 [03:28:13] https://github.com/mxcl/homebrew/commit/583272dcd70d886f97416b4945535546cd5dbddc [03:28:22] Looks like I'l have to run brew update first :) [03:29:13] yay 1.8.2 [03:29:21] \O/ [03:34:15] hashar: The only interesting point I see so far is about "git rm $submodule" sucking less from now on. [03:35:18] ... which is in 1.8.1 [03:37:19] yay, globstar! [03:37:21] **/ [03:37:24] rhgr [03:37:34] https://gerrit.wikimedia.org/r/#/c/53533/ on integration/docroot :/ [03:37:38] doc.wikimedia.org/index.html [03:37:47] the current layout is /srv/org/wikimedia/doc/ [03:38:09] hashar: I did that on purpose to avoid accidentally mixing it [03:38:15] needs to be symlinked [03:38:25] Since puppet can put arbitrary things in /srv/ when asked to do so [03:38:36] that is going to be a mess [03:38:53] I will get rid of the puppet parts eventually to clean that up :-] [03:39:03] hashar: The repo would be set up from puppet, right? It is customary to separate installation and enabling of things [03:39:18] but I first need the contint module to land in production. Faidon been reviewing my change this week. [03:39:39] my idea is that a merge on integration/docroot.git will trigger a jenkins job that will git pull on /srv/ [03:39:45] self deployment :-] [03:39:49] so it'd be like file { "foo.conf" } and apache_site( "foo" ); one for git-clone / git-pull, and one to symlink it where it needs to be be. [03:40:02] hashar: That is still possible [03:40:18] except that it'd git pull in /var/**/docroot [03:40:22] or wherever [03:41:02] I have been asked to use /srv/ :-] [03:41:10] hashar: putting the git repo directly in /srv/ would be a more dangerous. [03:41:11] I guess that is more or less the standard for ops [03:41:19] hashar: For sure [03:41:33] what would be dangerous? [03:41:37] hashar: But there's a difference between overriding /srv with a git repository and putting files there [03:41:45] they will obviously be served from /srv/ [03:42:00] that is the point isn't it ? :-] integration/docroot.git is already public [03:42:09] I'm merely saying we shouldn't turn the whole thing into a git repository as that would be near impossible to puppetise without conflicts. [03:42:25] the idea is to get rid of puppet entirely [03:42:41] to avoid having to wait for ops to review merge our changes. They have really no added value there [03:43:08] I will move any reference to /srv/ in puppet to the integration/doocroot.git repo [03:43:18] No, we move things out of puppet into something else where appropiate, we're not going to have nothing puppetised. gallium is a local hack as it is, making it even harder to maintain is not what you have in mind. [03:43:30] and also need to convert the puppet documentation generator to be a jenkins job instead of a puppet run [03:43:39] Yes [03:43:42] One thing at a time [03:44:02] hashar: That sounds contradictory [03:44:05] the docroot is going out of puppet. for sure [03:44:17] hashar: firs you say you want to put the git repo in /srv, and now you say you want to change references to /srv/ to where docroot.git is? [03:44:20] ops and I are tired about having to review such changes as editing index.html or tweaking some css file [03:44:23] hashar: Yes [03:44:38] so [03:45:17] So, we puppetise only install and enable, not update. 1) git-clone docroot.git 2) symlink to /srv/ [03:45:18] I will put all materials from puppet into integration/docroot.git , that includes the Apache .html header files I am using for nightly builds as well as the landing page at integration.mediawiki.org [03:45:21] update will be from jenkins. [03:45:34] then I get rid of /srv/ as it is and git clone integration/docroot.git to /srv/ [03:45:36] done :-] [03:45:43] no more puppet involved [03:46:03] hashar: Why? How does that make anything easier? it's a one time cost to set up puppet to clone it somewhere and symlink it [03:46:11] I don't mean symlink /srv/ but /srv/org/wikimedia/doc [03:46:13] etc. [03:46:15] cause we do not want to use puppet to deploy that [03:46:30] We don't deploy using puppet but jenkins. [03:46:40] Only the configuration, just like the apache configuration is in puppet [03:46:45] why do you want to use symlink when you can get apache to serve directly from the git working copy? [03:47:22] That's 1 line in puppet that'll "never" change, with the benefit of consistency and easy of understanding for other people. [03:47:25] and at no cost or down side [03:47:38] what 1 line are you talking about? [03:47:57] symlink /somewhere/docroot/foo.org to /srv/org/foo [03:48:20] which has no purpose? [03:48:36] Why do you want to point Apaches to /somewhere/docroot/foo.org when it can points directly to /srv/org/foo ? [03:49:12] Apache conf points and will stay pointing to /srv/org/foo [03:49:52] that's conventional and what other people expect, it doesn't matter where apache points to, it's just DocumentRoot in the conf file. The difference is putting out outside /srv/ is "different" and has no benefit, right? [03:50:03] So we agree we want apache pointing to /srv/org/foo, right? [03:51:44] I'd say either put the docroot clone outside /srv and make symlinks in /srv to where docroot is cloned, or change change apache conf to point to where the docroot is cloned [03:52:17] sorry I don't get it [03:52:18] but not turning /srv itself where it is in-place into the git repo as that would potentially cause conflicts and is harder to maintain in the future when we move things around as it is not expected to be a repository [03:53:06] e.g. it should be easy to enable some module in puppet on any server and work, and not suddenly show up in untracked files got the integration/docroot git clone. [03:53:09] That's all. [03:53:24] ok [03:53:31] so if you consider puppet should not put files in there [03:53:37] then the problem is solved :-] [03:53:49] we still have the issue of Jenkins publishing files under /srv/ though :-] [03:54:01] There is no spoon. [03:55:16] For our own convenience it is best to have either the files themselves or a symlink to the files in /srv/, that way we don't have race conditions when updating the jenkins jobs and it has the benefit of other people who look in /srv looking for something can find it. [03:55:58] That leaves three options: Turning /srv into a git repo, Turning individual /srv/org/foo dirs into a git repo or turning /srv into a git repo, clone it elsewhere and symlink to those dirs from /sv/org/foo [03:56:26] The latter is probably easiest in the long term and only slightly more complicated to set up. [03:57:20] New patchset: Mwjames; "Add qunit job template for extensions" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53534 [04:01:55] New review: Krinkle; "(1 comment)" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53534 [04:02:37] New review: Mwjames; "Since I don't really know what I'm doing here but on the advice given by Timo I added:" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53534 [04:04:21] New patchset: Mwjames; "Add qunit job template for extensions" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53534 [04:05:53] New review: Krinkle; "(1 comment)" [integration/jenkins-job-builder-config] (master) C: -1; - https://gerrit.wikimedia.org/r/53534 [04:16:13] New review: Mwjames; "(1 comment)" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53534 [04:32:29] New review: Krinkle; "(1 comment)" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53534 [04:57:57] New patchset: Hashar; "(bug 31518) PHPUnit code coverage for mw/core" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53532 [04:58:30] New review: Hashar; "6am GMT to 3am GMT. This way european people are still sleeping." [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53532 [05:51:19] gerrit-wm: Yeah, right. Still sleeping, uh? [05:51:22] hashar: ^ [05:51:37] maybe it died :] [05:52:04] I was referring to the "european people" [05:52:06] xD [05:52:23] ohhh [05:52:35] well I don't consider you as being european :-] [05:52:38] you are not even SF! [05:52:54] I would say you are somewhere on the west of the pacific ocean [05:53:22] Krinkle: wait what day are we? [05:53:46] New review: Krinkle; "(1 comment)" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53532 [05:54:10] hashar: I just woke up after a short sleep (I'm switching time shifts myself as well) on a nice Thursday morning. [05:54:25] ahh [05:54:30] In the Netherlands [05:54:44] switching away from Hawaiian time to CET for a few days. [05:54:46] it could have been late evening of wednesday [05:54:52] trying to figure out what timezone it could be [05:55:43] honolulu in hawai is wednesday, 8pm ] [05:55:56] Hawaiian is my usual time, it's basically SF time (WMF office) minus 2 hours. That way I wake up a few hours before the stores close in the Netherlands, and I go to bed before morning rush hour [05:56:40] I wish SF was in GMT-8 or GMT-7 [05:56:57] the sun wake up early there, so they can probably shift a bit [05:57:54] Krinkle: https://gerrit.wikimedia.org/r/#/c/53532/4/mediawiki.yaml,unified [05:58:03] Krinkle: your comment does not make sense to me :-] [05:58:09] I have ansicolor there! [05:58:31] hashar: defaults.yaml global.wrappers[] = timestamps [05:58:40] ah yeah [05:58:42] so that one is redundant [05:58:56] that leaves ansicolor, which I'd recommend adding to global defaults as well. [05:58:59] I think the job definition override whatever is in the defaults [05:59:12] it is basically a global default already since we never use "global" but "zuul" [05:59:14] which does have it [05:59:17] aka I had https://gerrit.wikimedia.org/r/#/c/53532/1/mediawiki.yaml,unified (with no timestamps) [05:59:19] and that missed the timestamps [05:59:26] interesting [05:59:32] maybe I should drop the global [05:59:52] no, then you need to include the description manually [05:59:58] and add project-type [06:00:03] and timeout et. [06:00:29] true [06:00:35] so will let it as is for now :-] [06:00:47] k [06:00:48] doh JJB Is active https://review.openstack.org/#/q/status:open+project:openstack-infra/jenkins-job-builder,n,z [06:00:58] marktraceur sent 3 patches in [06:01:16] we even managed to have a duplicate patch :-] [06:03:27] breakfast time for the little familly [06:03:29] * hashar waves [06:05:16] The little family http://www.alicia-logic.com/capsimages/stu_027.jpg [06:32:49] New patchset: Krinkle; "Defaults: Also add ansicolor to global.wrappers" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53723 [06:34:19] New patchset: Krinkle; "Defaults: Also add ansicolor to global.wrappers" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53723 [06:34:37] Change merged: Krinkle; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53723 [06:49:04] New patchset: Krinkle; "doc/configuration: Fix incorrect name "defaults"" [integration/jenkins-job-builder] (master) - https://gerrit.wikimedia.org/r/53724 [06:49:53] Change abandoned: Krinkle; "https://review.openstack.org/24407" [integration/jenkins-job-builder] (master) - https://gerrit.wikimedia.org/r/53724 [07:01:02] New patchset: Krinkle; "PHPUnit code coverage for mediawiki-core" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53532 [08:01:44] Ryan_Lane: is there a general problem ? cannot "git pull" [08:42:07] New patchset: Hashar; "PHPUnit code coverage for mediawiki-core" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53532 [08:42:22] New review: Hashar; "got it generated, good enough for now." [integration/jenkins-job-builder-config] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/53532 [08:42:23] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53532 [10:09:42] New patchset: Hashar; "Add more parsoid tests" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/53700 [10:10:00] New review: Hashar; "You are the best!" [integration/zuul-config] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/53700 [10:10:01] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/53700 [10:10:31] New review: Hashar; "deployed" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/53700 [12:05:57] New patchset: Hashar; "Revert "Update files and add doc.wikimedia.org"" [integration/docroot] (master) - https://gerrit.wikimedia.org/r/53742 [12:06:08] Change merged: Hashar; [integration/docroot] (master) - https://gerrit.wikimedia.org/r/53742 [12:14:38] New patchset: Hashar; "import web files from puppet" [integration/docroot] (master) - https://gerrit.wikimedia.org/r/53743 [12:15:31] Change merged: Hashar; [integration/docroot] (master) - https://gerrit.wikimedia.org/r/53743 [12:38:23] New patchset: Hashar; "postmerge for integration/docroot" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/53746 [12:38:38] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/53746 [12:48:34] New patchset: Hashar; "integration/docroot only for master, and tweak messages" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/53747 [12:48:48] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/53747 [12:51:02] New patchset: Hashar; "integration-docroot-deploy" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53748 [12:53:10] do we have any plan to relicense code under Module: namespace under GPL [12:53:21] it's difficult to reuse code currently [12:55:42] New patchset: Hashar; "integration-docroot-deploy" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53748 [13:00:16] New patchset: Hashar; "integration-docroot-deploy" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53748 [13:01:11] New review: Hashar; "deployed" [integration/jenkins-job-builder-config] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/53748 [13:01:12] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53748 [13:02:03] liangent: yeah some people complained about that already [13:02:18] liangent: I guess you want to ask on wikitech / or ask Tim Starling [13:02:46] liangent: maybe anomie knows whether we are going to change the license of the Module: namespace on mediawiki.org [13:03:12] hashar: mw.org-only is not enough [13:03:25] it still prevents us from reusing 3rd party code on wikipedia [13:04:09] and we need to resolve this asap because technically we have to ask everyone who has edits in the Module namespace first [13:04:25] hashar, liangent- There's been some talk about licensing issues, but I don't know that it's really gone anywhere. sumanah or robla might know. [13:05:25] New patchset: Hashar; "rm old files" [integration/docroot] (master) - https://gerrit.wikimedia.org/r/53751 [13:05:42] Change merged: Hashar; [integration/docroot] (master) - https://gerrit.wikimedia.org/r/53751 [13:16:38] anomie: I talked about it a few weeks ago in our core meeting [13:16:53] I think ryan lane mentioned the licensing issue to me [13:17:35] anomie: liangent: that would be greg-g [13:20:32] The "Navigable TOC" idea in http://usability.wikimedia.org/wiki/File:Babaco_ntoc_click.jpg - is that idea (and its bug reports) dead? [13:22:52] I think we should change the license now [13:23:02] before people has written hundreds of lines in module namespace [13:50:57] New patchset: Hashar; "operations-puppet-doc for docs.wikimedia.org/puppet" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53753 [13:51:33] New review: Hashar; "Require upstream https://review.openstack.org/24427" [integration/jenkins-job-builder-config] (master) C: -2; - https://gerrit.wikimedia.org/r/53753 [14:01:33] https://bugzilla.wikimedia.org - I can't login as VictorPorton (I've changed the password twice with "Forgot Password" but this does not help. Your Bugzilla is buggy! [14:04:52] <^demon> Our Bugzilla is working fine, you're just trying to use your wiki name for Bugzilla. [14:05:16] <^demon> You have to register for BZ, it doesn't use your wiki user/email. [14:06:00] ^demon: I have previously posted a bug in this bugzilla. So my user name is correct: https://r.mail.yandex.net/urls/ZDrEXHsXsESN3qlrtTPNCg,1363270866/bugzilla.wikimedia.org%2Fshow%5Fbug%2Ecgi%3Fid%3D32809 [14:06:25] Sorry: https://bugzilla.wikimedia.org/show_bug.cgi?id=32809 [14:06:40] ... This bug is associated with my user name [14:07:08] <^demon> You login with e-mail addresses though, not your display name. [14:07:41] ^demon: Yes, it works with my email. Txs [14:30:32] can somebody help me with git http://laxstrom.name/help.txt ? [14:31:50] <^demon> What's git log say? [15:14:39] New patchset: Demon; "Add labsconsole extensions to make-wmf-branch" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/53519 [15:14:39] Change merged: Demon; [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/53519 [15:20:48] chrismcmahon, did anybody else get involved in writing scenarios apart from sucheta, rachel99, valeriej and ourselves so far? [15:21:10] chrismcmahon, also can we mark as DONE some of the bugs listed at https://www.mediawiki.org/wiki/QA/Browser_testing/Search_features ? [15:21:27] chrismcmahon, today I will write another one. Can you please suggest one? [15:21:29] qgil_: I think that is all the people [15:22:23] chrismcmahon, I will do the DONE part in the wiki page [15:22:38] qgil_: sorry in a meeting [15:22:41] ah ok [15:25:40] qgil_: a scenario for the 'regression tests' at the bottom of http://www.mediawiki.org/wiki/QA/Browser_testing/Search_features would be useful [15:27:30] chrismcmahon, ok I have a couple of questions but I can perfectly wait until you are done with the meeting. [15:36:10] qgil_: what's the question? [15:37:31] chrismcmahon, "Umlauts and accents are disregarded in search" is the expected behavior? And this has nothing to do with your "A" scenario? [15:39:27] qgil_: the accent test was what rachel99 and I worked on. you can see https://bugzilla.wikimedia.org/show_bug.cgi?id=44238 documented in the backlog http://www.mediawiki.org/wiki/Qa/test_backlog#Search [15:41:26] chrismcmahon, should I still do "Umlauts and accents are disregarded in search" or is that already covered, then? [15:41:42] qgil_: I'd say pick something else [15:41:54] chrismcmahon, ok, good that I asked. :) [15:42:15] chrismcmahon, "Some ligatures match the separate letters." ? [15:42:29] sure [15:42:33] ok, taken [15:42:43] hashar: Thanks for the merge + deploy, you'll need to teach me how soon :) [15:43:15] hashar: And for the backup at review.openstack.org, though I'm sorry we overlapped on that change [15:43:33] marktraceur: not a big deal :-] [15:43:43] overlapping usually happen with small changes hehe [15:46:43] qgil_: how do we change the channel topic back to normal? :) [15:47:06] Type "/topic" and etc. [15:47:29] zeljkof, but I would still leave the URL to https://www.mediawiki.org/wiki/QA/Browser_testing/Search_features [15:47:35] qgil_: thanks, I thought I had to have special permission to do that [15:47:54] qgil_: I would like to remove "streaming now" part :) [15:48:10] zeljkof, JDI :) [15:50:04] New patchset: Hashar; "integration portal ported to twitter bootstrap" [integration/docroot] (master) - https://gerrit.wikimedia.org/r/53757 [15:50:20] marktraceur: if you know about twitter bootstrap .. :-] [15:51:04] New review: Hashar; "Went wild this afternoon and decided to port the integration homepage to twitter bootstrap :-D" [integration/docroot] (master) - https://gerrit.wikimedia.org/r/53757 [15:53:56] hmm [15:54:15] qgil_: looks like channel topic is not complete, it ends in "...| Cha" [15:54:18] and I am out for the rest of the day now :-] [15:55:09] (innocent voice) zeljkof, yeah, I saw that. Probably we (I) hit some limit yesterday. [15:55:31] qgil_: do you know what was there? [15:55:49] qgil_: I vaguely remember something about where channel is logged [15:55:53] (even more innocent voice) zeljkof, nope [15:56:23] qgil_: I am sure somebody will fix it, eventually :) [15:56:51] zeljkof, http://bots.wmflabs.org/~wm-bot/logs/%23wikimedia-dev/?C=M;O=D [15:57:02] that is a long url [15:58:13] qgil_: looks like topic changes are not logged :) [16:14:50] qgil_, chrismcmahon: a few words about the event: http://filipin.eu/increase-the-backlog/ [16:16:13] zeljkof, nice! we can reuse a lot of that material for the next one [16:16:26] qgil_: that was my idea :) [16:16:39] qgil_: I wanted to write down all notes so I can reuse them the next time [16:16:50] +1 [16:20:38] thanks zeljkof [16:21:23] chrismcmahon: I was just reusing a lot of material that you and qgil_ created :) [16:21:50] chrismcmahon: I have to pay a few bills and then moving to making tests green again [16:21:56] ^demon: is replication of gerrit DB waiting on https://gerrit-review.googlesource.com/#/c/37711/ by any chance? [16:22:22] <^demon> You mean of the actual database? [16:22:53] yes [16:23:02] <^demon> No, it's waiting on someone to do it. [16:23:06] <^demon> There's nothing upstream we need for that. [16:23:09] Ok. :) [16:24:58] zeljkof: I am looking at the aftv5 test now [16:25:55] chrismcmahon and zeljkof - how's everything going? [16:26:14] cheezy: hi there! :) [16:26:31] hi cheezy [16:26:35] hello gents [16:26:44] cheezy: busy day yesterday, did you see this? http://filipin.eu/increase-the-backlog/ [16:26:58] I did - listening to the preso now [16:27:16] heh. I haven't listened to it. hope I didn't suck... [16:27:39] sounds fine to me [16:28:40] chrismcmahon: I listened to it today, it was good [16:28:45] are you guys still in the bay area [16:28:57] chrismcmahon: I was too distracted to listen carefully yesterday :) [16:29:08] cheezy: no, I am back to the old continent [16:29:52] cheezy: zeljkof is in Croatia and I am in Colorado. I'll be in CA for a couple of months starting Apr 1 [16:30:18] I'll be there week of April 22. We'll have to have dinner. [16:31:00] cheezy: I'll be omw to GTAC in NYC Apr 22 I believe [16:31:08] ah [16:31:34] chrismcmahon: are you going to be in Austin in early May? [16:31:44] cheezy: yes! [16:31:55] great [16:44:33] good luck to you guys and happy testing [16:44:47] I have to leave the channel [16:44:53] cheezy: thanks, if you have free time, let us know :) [16:45:08] free time? what's that? [16:45:33] cheezy: I have heard it is something like free beer :) [16:45:42] then I'm in! [16:46:00] although I don't mind paying for my beer as long as it is good beer... [16:46:12] bye [16:46:18] cheezy: bye [16:47:59] For the oldtimers here: The "Navigable TOC" idea in http://usability.wikimedia.org/wiki/File:Babaco_ntoc_click.jpg - is that idea (and its bug reports) dead? [16:51:06] zeljkof: You know, I added the corrected scenario on the backlog wiki yesterday. Was I supposed to write both? As in ours, and the one corrected by you? [16:52:51] sucheta: you should add the scenario you think is better :) [16:52:51] mlitn: I notice that the link to "All comments" in AFTv5 no longer exists on e.g. http://en.wikipedia.beta.wmflabs.org/w/index.php?title=Special:ArticleFeedbackv5/Aftpage&ref=cta#806003a219dee512b39c9ec255ae01b0 is that by design? [16:53:11] sucheta: I gave my suggestion, does not mean it is correct [16:53:36] chrismcmahon: there's a couple of things that did not yet get reviewed, so these should be expected next week [16:53:38] and that's one of them [16:53:39] sucheta: by the way, my notes on the event: http://filipin.eu/increase-the-backlog/ [16:54:29] zeljkof: ^^ that's the source of the test failures for AFT right now [16:55:11] zeljkof: Ah, no :) I meant, the better one only. And checked the notes. Thanks :) [16:55:12] chrismcmahon: the link is missing? [16:55:25] yes [16:57:48] chrismcmahon: that should be easy to fix, right? [16:58:14] zeljkof: can we comment And steps in a .feature file? [16:58:23] valeriej: take a look at my notes from the event yesterday: http://filipin.eu/increase-the-backlog/ [16:58:45] chrismcmahon: we _can_ but if the step is not needed, just delete it [16:59:06] zeljkof: I want to just not execute it for the next week or so [16:59:32] chrismcmahon: delete it anyway, we can restore it from git history [16:59:46] chrismcmahon: I _really_ do not like commented out code :) [17:00:24] zeljkof: OK, checking that test passes now, will do git review [17:00:33] chrismcmahon: great [17:00:58] * valeriej is reading [17:03:34] zeljkof: Looks good! [17:03:44] valeriej: thanks :) [17:07:54] New patchset: Cmcmahon; "remove steps that depend on All comments link which no longer exists for now" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/53772 [17:25:44] Aaron|home: [17:25:44] 3 Fatal error: Undefined class constant 'SERIALIZER_IGBINARY' in /home/wikipedia/common/php-1.21wmf11/includes/clientpool/RedisConnectionPool.php on line 90 [17:25:47] AaronSchulz: ^ [17:27:19] yeah, must be missing from the pecl ext [17:27:22] meh, php it is then [17:27:45] * Aaron|home wonders why he didn't get that on fenari [17:27:57] Reedy: what server? [17:28:35] AaronSchulz: srv193 [17:28:38] yay for testwiki [17:38:47] Change merged: Zfilipin; [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/53772 [17:40:03] chrismcmahon: merged the commit [17:40:42] thanks zeljkof [17:58:00] any git gurus around? [17:58:26] ^demon: git log doesn't seem to say anything that I find weird [17:58:37] Nikerabbit: again same problem? [17:58:41] did you try ^demon> What's git log say? [17:58:49] ah too slow I've been [18:07:50] ^demon: https://gerrit.wikimedia.org/r/#/c/53538/1 [18:14:34] andre__, ^demon your feedback about https://www.mediawiki.org/wiki/User:Qgil/Contributors is welcome, especially on the tags / ontology and the connection between Gerrit / Bugzilla and tasks. [18:14:48] andre__, ^demon not urgent and comments in the discussion page are fine [18:15:11] * andre__ adds to TODO list [18:16:05] Nemo_bis, your feedback about https://www.mediawiki.org/wiki/User:Qgil/Contributors is also welcome! [18:16:49] mlitn: How is today's deployment to AFT5 coming along? Is the old data converting well to the new schema? [18:17:37] it's still going strong, but will take a couple more hours to complete [18:18:03] Platonides I also want to invite you explicitly to comment on https://www.mediawiki.org/wiki/User:Qgil/Contributors [18:18:32] New patchset: Ori.livneh; "Add NavigationTiming extension" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/53781 [18:18:48] FYI for others on this channel: AFT5 is now down on en-wiki and de-wiki for at least 24 hours, depending on how long it takes for the old data to be merged on the new schema. You can follow our progress for this multi-day release on this Etherpad checklist: http://etherpad.wikimedia.org/AFT5-release Fingers crossed! [18:19:37] Change merged: Ori.livneh; [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/53781 [18:19:50] qgil: watchlisted [18:20:05] thanks [18:20:07] qgil: have you checked the link in https://www.mediawiki.org/wiki/Wikitech-l/best_of [18:20:36] and Nikerabbit's thread "is the current system too hard for volunteer devs" (something like that) [18:20:45] mlitn: Cool, thanks! Keep us posted on your progress, so we get a sense of overall timing. Also, check off completed tasks on the Etherpad checklist. Have we forgotten anything important that should be there for the next few deployments? [18:21:14] Nemo_bis, no, thank you. This is why I'm inviting people like you to comment even at this earliest stage :) [18:25:17] fabriceflorin: it's current pace is too slow to pick up production testing of the new features today [18:25:29] I won't be "re-enabling" AFT before tomorrow :) [18:26:14] by tomorrow evening, everything should be set though [18:31:33] mlitn: Thanks for the update. I expected that we would not 're-enable' AFT5 to production today. But please let us know ASAP if you think we'll be down for more than 24 hours. Is it possible that the conversion to the new schema could come to a complete halt while you are sleeping? If so, do we have to start all over again? [18:33:37] no way to tell for sure, but I doubt we'll need to postpone; 24h should do [18:34:04] as soon as I notice we might not make that, I'll let you know, but so far so good :) [18:45:17] qgil: I like the ideas you're putting out there, after a quick skim. I added a few notes in the bits about People. [18:45:41] marktraceur, thank you! [18:51:16] zeljkof: have you ever used the "scroll into view" feature in selenium? [18:51:42] chrismcmahon: I am scrolling like I lived there :) [18:53:21] chrismcmahon: what is the problem? [18:53:25] zeljkof: OK, I think the GuidedTour test is failing in IEs because the gt-gettingstarted-4 element is too low to be "visible". (FF and Chrome can see it, but I think IE can not) [18:53:59] chrismcmahon: sounds familiar [18:54:36]