[00:05:36] New patchset: Krinkle; "Make it possible to disable Special:Activeusers" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/60394 [00:05:55] New review: Tim Starling; "The output-side message is also shown with --verbose:" [mediawiki/extensions/Score] (master) - https://gerrit.wikimedia.org/r/61181 [00:10:51] New review: Kaldari; "(1 comment)" [mediawiki/extensions/Echo] (master) C: -1; - https://gerrit.wikimedia.org/r/61819 [00:21:27] New patchset: Tim Starling; "Fixed behavior of Title::invalidateCache." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/59848 [00:21:40] New review: Tim Starling; "Looks great, thanks for that!" [mediawiki/core] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/59848 [00:21:41] Change merged: Tim Starling; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/59848 [00:41:06] that doesn't sound like Tim [00:42:41] oh noes, someone has hacked my gerrit account! [00:43:30] or maybe not, since whoever it is is also writing characteristically nasty comments on https://gerrit.wikimedia.org/r/#/c/62233/ ;) [00:49:04] New review: Eloquence; "Please amend the commit message to clarify under what exact circumstances we're suppressing talk pag..." [mediawiki/extensions/Echo] (master) - https://gerrit.wikimedia.org/r/61819 [00:51:15] TimStarling: Did you intentionally bypass Jenkins on that commit? [00:51:23] Looks like it caused a fatal error in the unit tests [00:51:45] https://gerrit.wikimedia.org/r/59848 [00:52:02] Though jenkins-bot was unable to CR-2, see the comment it left for details [00:52:07] PHP Fatal error: Using $this when not in object context in /srv/ssd/jenkins/workspace/mediawiki-core-phpunit-api/includes/actions/InfoAction.php on line 68 [00:52:46] New patchset: Tim Starling; "Revert "Fixed behavior of Title::invalidateCache."" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62405 [00:52:58] The first patch set didn't have that [00:53:00] Strange [00:53:48] rebasing or merging can cause test failures, it happens sometimes [00:55:18] i'm going with gerrit account hacked [00:55:37] lol [00:56:06] as for bypassing jenkins, I took your advice on that [00:57:07] TimStarling: I'm not sure I follow [00:57:09] didn't you point out that rebasing then giving +2 as a workflow would cause merge noise in the history? [00:57:24] I thought, good point, I should just click submit [00:57:59] Nice try, but I'm pretty sure that's not what I said :P [00:58:46] if you rebase and immediately submit, it is all nice and clean [00:59:11] You pointed out to me that rebasing keeps git history cleaner, not the other way around [00:59:24] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62405 [00:59:37] I pointed out that doing so adds noise to the Gerrit thread, and doubles the load on Jenkins/Zuul, which is a bit sensitive at this time [01:00:11] rebasing and also submitting directly, adds those 3 events (test, gate, post) at the same time (instead of deferring post to later) [01:01:04] and is bypasses failures only visible when applying the change to master, which is why we have a gate pipeline that simulates a merge and runs tests again, instead of just verifying that there is V+2 and merging it [01:01:25] anyhow, got early, no worries :) [01:01:51] I just use this crap, I don't write it [01:02:04] Krinkle, related: https://review.openstack.org/#/c/28128/ [01:02:14] I happen to care more about straight histories than gerrit email quantity [01:02:22] anyhow, the rebase trick you do I like (I don't use it myself, but I can appreciate it) [01:02:35] just avoid bypassing the gate test, that's all :) [01:02:59] some projects actually run tests in gate that don't run on patchset-creation, so it can potentially bypass more then just a retest on latest master. [01:03:25] not sure how I can avoid that [01:03:43] by not touching the "Verified" field and not pressing the "Submit" button [01:03:54] but then it would take 10 minutes to merge [01:03:59] instead CR+2 and publish [01:04:03] and it often takes less than 10 minutes to review a patch [01:04:14] so the next patch I review would not be based on the one I just gave +2 [01:04:14] It works asynchronous [01:04:18] and we're trying to reduce that time [01:04:18] it would be based on the one before [01:04:45] but I'm not discussing this for real, I assume you were joking, because this is standard policy to my knowledge. [01:05:01] bypassing it structurally is going to be a problem I think [01:05:30] Let's discuss it later or on the mailing list. There's surely some importnat points here. [01:06:01] But bypassing the gate is not a solution. Then we should agree on disabling it alltogether and all bypass it. [01:06:02] cya ltr [01:11:43] New patchset: Tim Starling; "Fixed behavior of Title::invalidateCache." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62406 [01:31:05] New patchset: Himeshi; "Semantic Forms: Display error message when Special:EditForm accessed by itself" [mediawiki/extensions/SemanticForms] (master) - https://gerrit.wikimedia.org/r/62407 [01:52:12] New patchset: Aaron Schulz; "Refactored Special:ActiveUsers queries to remove filesort/temp table." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62408 [01:57:44] New review: Aaron Schulz; "For the listing query (for non-oversights):" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62408 [02:01:29] New review: Tim Starling; "I reverted the original commit (I38ac616d) due to a post-merge unit test failure. I resubmitted the ..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62406 [02:25:48] New patchset: Ori.livneh; "(Bug 48117) Inline the call to get content model" [mediawiki/extensions/Scribunto] (REL1_20) - https://gerrit.wikimedia.org/r/62409 [02:27:57] New review: Ori.livneh; "Note that this change set targets the REL1_20 branch." [mediawiki/extensions/Scribunto] (REL1_20) - https://gerrit.wikimedia.org/r/62409 [03:24:31] New patchset: Yurik; "Added a comment to run vagrant provision" [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/62410 [03:28:44] New review: Ori.livneh; "Thanks!" [mediawiki/vagrant] (master) C: 2; - https://gerrit.wikimedia.org/r/62410 [03:30:38] yurik: did it work for you? (did you enable an extra class?) [03:31:31] Change merged: jenkins-bot; [mediawiki/vagrant] (master) - https://gerrit.wikimedia.org/r/62410 [03:31:57] * Aaron|home reads https://en.wikipedia.org/wiki/Auto-Tune [03:32:45] ori-l, still running [03:32:51] i gotta get a new laptop :) [03:33:06] ori-l, i'm trying it on mac, but i was hoping to do it on a pc [03:33:19] the browser tests? [03:33:20] in theory, it should run on a pc because the web driver is remotable [03:33:30] yep [03:34:25] ori-l, the goal is to have all the messy ruby and other stuff in VM, and have the minimal set of extra code on the host to expose my local browsers to the VM's control [03:34:29] it's configured to drive a browser running on the guest (as you know, having looked at it). but yes, you're right: you could in principle install the selenium driver on your host's firefox (or for that matter firefox anywhere), and have the tests drive that instead. [03:35:26] ori-l, exactly - i think this is a much better approach as it keeps the VM small, and does not require VM to have UI stuff at all (like FF or anything else) [03:35:39] (or X) [03:36:41] right, that's why i kept it optional. i think we should probably take it a step further and just bundle sauce connect. it's nonfree, but oh well. [03:37:16] there's nothing in the OSS world that is comparable; testswarm is powered by browserstack now [03:38:12] (sauce connect opens a private tunnel for sauce's cluster of selenium-driven browser testing VMs to connect to your instance, so you can test on all major browsers) [03:38:43] it's free as in beer for open-source projects, IIRC [03:39:40] ori-l, sure, that's an option, but i thought we can already do remote testing with selenium with is FOSS ? [03:40:01] to set it up as I described - VM with ruby, host - with the UI [03:40:24] guest shouldn't have any problem communicating with the host [03:41:16] ori-l, just got an error: cd /tmp/vagrant-puppet/manifests && FACTER_virtualbox_version='4.2.10' puppet apply --verbose --modulepath '/etc/puppet/modules:/tmp/vagrant-puppet/modules-0' site.pp --detailed-exitcodes || [ $? -eq 2 ] [03:41:19] well, yeah, but the advantage of vagrant is that apart from some initial bootstrapping you can precisely control every last configuration detail of the machine, so provisioning can be automated and reproducible and exact [03:41:52] and if you're installing software on the host you basically have to hope for the best [03:41:54] ori-l, in theory, yes, but i think with browsers its "good enough" - browsers tend to change overtime anyway :) [03:42:07] what came before that? [03:42:10] that message, I mean [03:42:32] that's just a generic 'puppet barfed' error; the log prior to that is essential. [03:43:12] ori-l, http://dpaste.com/1111523/ [03:43:47] you didn't reprovision a new machine, right? [03:44:01] ori-l, no, used my old machine [03:44:14] ruby-bundler takes up more RAM than your average 747 flight simulator [03:44:32] ori-l, at some point i would love to have multiple VMs all using the same mw repo [03:44:33] i spent a full day debugging it since it would just crash with a generic error [03:44:55] :) [03:44:58] the programmers apparently thought it was a good idea to catch specific exceptions and rethrow them with a friendly but nonspecific error [03:45:04] anyways, it was an OOM [03:45:12] lovelly [03:45:52] i wasn't sure what to do about it but in the end i thought what the hell, increase the ram size of the VM to 712 (not a power of 2, but I miscalculated :)) and see if anyone complains [03:46:04] you have one of two options: [03:46:25] you can either 'vagrant ssh', cd to /srv/browsertests, and run 'bundle install' [03:46:33] (without the overhead of puppet there's enough memory for it to run successfully) [03:46:51] running... [03:47:06] or, if you want to verify that the puppet manifest works, just use the virtualbox gui and bump up the ram size temporarily [03:47:48] ori-l, would love to do it, but really need to get tests running before deployment on tue [03:48:14] could you tell me how to easily set it up to run tests on that free-as-in-beer platform? [03:48:38] btw, they still refuse to set headers in selenium :((( [03:49:11] i did it at my previous job for something [03:49:14] i'm almost certain that you can [03:49:32] sauce connect is here: https://saucelabs.com/docs/connect [03:49:49] as for how to get the browsertests code working with it, dunno, didn't write that code [03:50:18] i run away from the whole cucumber thing [03:50:34] sigh. thanks :) [03:50:40] will play with it in a bit [03:50:42] what are you trying to do for tuesday, exactly? [03:51:04] get any form of good browser testing in place, or augment the official set? [03:51:13] well, "ideally" i would love to test all functionality of the new zero config stuff [03:51:49] oh, i have something for you, hold on [03:52:16] https://github.com/atdt/karaga [03:52:29] * yurik is being kicked out of a coffee shop :( [03:53:23] ori-l, thx! i will take a look once i get home :) [03:53:31] which coffee shop, btw? [03:53:41] pick me up [03:53:53] oh right, you're in the east village [03:54:23] yep :) [03:54:34] ori-l, why, you know some good ones somewhere? [03:54:42] i'm always on a lookout for good places [03:55:14] * yurik needs https://code.google.com/p/selenium/issues/detail?id=141 [03:55:41] why or why selenium devs refuse set header functionality :((( [03:55:57] ok, off i go, will login in a bit [03:56:02] ori-l, thx!!! [03:56:33] veselka is 24/7 [03:56:41] one of the few places, and right nearby [03:56:42] ohhh noo [03:56:47] i know, i don't love it either [03:57:01] 24hr starbucks on 60th and bway is too far :P [03:57:11] i like that its 24x7, but its the worst possible place cozyness-wise [03:57:17] and i don't think they have wifi [03:57:28] nor do they have cute NYU students [03:57:48] well, maybe yes, but still not cozy at all [03:58:11] got a bobst library card? [03:58:39] nope, but i wouldn't want to go to the lib either [03:58:49] Change merged: Yaron Koren; [mediawiki/extensions/SemanticForms] (master) - https://gerrit.wikimedia.org/r/62407 [03:58:50] cofe shop is my think [03:58:52] thing [03:58:53] btw, it failed [03:59:13] increase the ram by 200 megs [03:59:41] Installing gherkin (2.12.0) with native extensions /usr/lib/ruby/1.9.1/rubygems/installer.rb:552:in `rescue in block in build_extensions': ERROR: Failed to build gem native extension. (Gem::Installer::ExtensionBuildError) [03:59:44] gross, i know. like i said, i didn't write it. [03:59:47] yes. [03:59:52] will do at home [03:59:55] i know that error message well [03:59:58] thx! [04:00:00] it's rethrowing an OOM [04:00:06] yurik, cute NYU students in a coffee shop on a sunday night? Must be finals time. [04:00:15] exactly! [04:00:19] ;) [04:00:26] off i go [04:00:39] you're in nyc? [04:00:57] yurik is, i am only spiritually [04:01:01] moved away a year ago [04:01:06] Yes, you're in the west coast. [04:01:22] I know. I saw a hebrew name and got all stalkerish. I checked out your user page. [04:01:28] * mooeypoo smiles nicely [04:01:30] Haaai. [04:01:40] heh, hey. it's a public page. [04:02:07] Yeah I know, but while I was typing it I thought it sounded a bit stalkerish. [04:02:27] I'm Moriel, btw. Also he-N [04:02:55] oh, hey. you applied for GSOC or something, right? how is that going? [04:03:12] Indeed I have. I have no idea, I am not sure when the results are up.. I think next week? [04:03:26] I started going into the VE code anyways though. Now it's a challenge that must be met regardless of GSoC [04:03:34] Heh. [04:03:57] The funny thing is that I know about RTL troubles from my experience, but I never really got into the nuances of exactly how programs deal with them... it's rather addictive once you do. [04:04:05] You guys need to add a disclaimer on the GSoC page. [04:04:24] It's a wiki. Add one yourself. ;-) [04:04:32] good point. [04:06:06] "Warning: Joining GSoC or similar projects may result in serious open-source addiction, leading to potential social hazards and sunlight deprivation." [04:06:14] That works? [04:06:24] Well, it's too late for this year. [04:07:04] I like that [04:07:46] One thing I will work on, potentially, though, are the tutorials [04:08:44] Some of them are a bit confusing, especially for newbies like me. Or for windows users [04:10:15] use vagrant! [04:10:32] yurik does -- he's on windows and knows what's good for him [04:10:35] yeah I actually gave up on it :\ [04:10:40] I ended up using VM [04:11:10] I tried it, and it worked, but I didn't find as many help pages for little things like where do I find the code or edit it or like.. how do I deal with the raw git commands in case I need them. [04:11:27] I'm sure it's out there somewhere, but it ended up being more confusing than just using my VM with Ubuntu [04:11:45] Especially since I had to get into 'fixing something' relatively quickly because of OPW requirements. [04:12:23] Also, I have git for windows which isn't too shabby either, but I have no idea how to make it work with git review. VM seemed like the best choice at the time. [04:12:43] mediawiki-vagrant is an ubuntu-based vm [04:12:55] yeah, I know, but the interface/how to do it was a bit confusing for me. [04:13:10] some of the commands didn't work and I found it difficult to sort them out with the existing documentation. [04:13:32] like what? maybe I can improve things [04:14:02] hm, I need to revisit it to see, hang on [04:14:42] ah, I think vagrant ssh didn't work for me [04:14:50] but I'll have to try it again to see exactly what the error was [04:15:07] btw, I was entirely confused about where to put my ssh keys in. The site I asked for dev access from had "SSH keys" [04:15:15] but the actual site to put the keys in is the gerrit one [04:15:22] that took a while to figure out. [04:16:21] well, i don't want to distract you from whatever, but the exact error message would be very helpful to me and i'd appreciate it [04:16:36] let me see if I can replicate [04:18:14] ok, silly question -- I already installed everything and had it running once, but now I don't remember how to run it again .. uhm.. any ideas? without reinstalling? [04:18:43] oh, sorry, I found the instructions. [04:19:37] back, found another good place - Table 12 [04:19:58] satisfies all the above requirements ;) [04:22:17] :o [04:22:21] Good morning people [04:23:56] ori-l, do you know this place? [04:24:00] it has been around for a while [04:24:09] hm, ori-l I'm now getting weird errors that I'm not sure if it's my fault or not. The vagrant VM is running, I'm in the vagrant folder, doing "vagrant up" it asked me to do vagrant init first. I did. Now I'm trying "vagrant up" again and it's saying "The box 'base' could not be found." [04:24:20] but the VM is running and I see MW in the localhost:8080 [04:24:38] so.. i'm confused. Also, where do i see the files? VM opens with a blackened window [04:26:51] i don't know table 22, no [04:27:03] but let's see about this problem mooeypoo is having, hrm [04:27:25] 'I'm in the vagrant folder, doing "vagrant up" it asked me to do vagrant init first' [04:27:28] what exactly did it say? [04:27:33] and: what version of vagrant are you running? [04:28:04] ori hang on I think I'm actually going to the vagrant folder while I should be in the mediawiki-vagrant one.. er, sorry, let me figure out where I installed everything again first. [04:28:38] yes, that would explain a lot :) [04:28:47] ok that explains it, heh, i forgot where I put it. Just found it :p sorry! [04:29:07] alright, now it says VM is already running. Good. [04:29:11] mooeypoo, make sure you get the latest vagrant repo from git [04:29:35] might even be a good idea to clone a new clean repo [04:29:37] yeah i did [04:30:09] ok, the problem is with the ssh. I was trying to use PuTTY [04:30:11] let me try again [04:30:26] run 'vagrant ssh' from the vagrant-mediawiki library [04:30:29] mooeypoo, you can't run vagrant ssh in windows [04:30:31] oh [04:30:33] unless! [04:30:36] it doesn't find an ssh client [04:30:39] yeah unless i add it to the path [04:30:41] you have the msys [04:30:44] but i will use putty [04:31:00] if you install that magic git msys stuff [04:31:08] ah ok here's a problem: it asks me for a password [04:31:09] New review: Parent5446; "Yes, I think the best solution would be to remove this cookie from the core entirely and then put th..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/53195 [04:31:10] you can do vagrant up and vagrant ssh from it [04:31:12] ... i have no idea what it is [04:31:20] I didn't setup any passwords when I started. [04:31:50] and I can't find where to set up the private/public keys [04:32:02] you need putty-agent I think; yurik would know [04:32:24] I can create the keys, but where do I save them so vagrant knows they belong to it [04:32:28] or where do I set them up? [04:32:47] mooeypoo, you could either go through the puttyagent, or through the msys's connection [04:32:51] there's PuTTYGen [04:32:57] i prefer msys because it works easily with everything else [04:33:05] I can make key pairs there... not good? [04:33:09] mooeypoo, what does vagrant ssh say -- it should tell you where the key is [04:33:31] hm ok yeah i see it for the private key. I assume I set up the public one in PuTTY [04:33:33] New review: Parent5446; "(1 comment)" [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/62406 [04:33:34] let me try [04:34:02] mooeypoo: these look like good instructions, too: http://zamboni.readthedocs.org/en/latest/topics/install-zamboni/vagrant-on-windows.html#configure-ssh [04:36:26] bah, 'server refused our key' [04:36:27] meh [04:38:26] ok i'm going to get coffee, brb [04:43:47] mooeypoo, take a look at the instructions i wrote on my meta user page [04:43:51] ori-l, it ran! [04:43:53] the FF [04:43:58] yei! [04:44:47] cool :) [04:46:54] yurik: what exactly are you trying to test? [04:47:18] are you just trying to get decent test coverage of the extension in general, or some specific piece of functionality? [04:48:27] never mind, saw your comment on https://code.google.com/p/selenium/issues/detail?id=141 [04:48:32] but why do you need browser testing for that? [04:48:57] just craft an HTTP request using the standard library of whatever language you want to use and inspect the output [04:49:08] inspect the HTML you get back, I mean [04:49:32] selenium isn't a good fit for that [04:49:53] it's brittle and slow [04:55:39] ori-l, in theory, you are right that if i wanted to test HTML, its good enough. But in the future, I will want to test that the javascript executes properly [04:55:52] New review: Parent5446; "This is still really really inefficient." [mediawiki/core] (master) C: -1; - https://gerrit.wikimedia.org/r/62408 [04:56:06] so for now yes, it will satisfy my needs, but we really need to get rid of the static html [05:01:00] New review: Parent5446; "(1 comment)" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62408 [05:01:51] well, you should separate the tests for these two things: [05:02:20] 1) if visiting the site from foo carrier, does RL module 'ext.zero.fooCarrier' load? [05:02:32] 2) does ext.zero.fooCarrier function correctly? [05:03:20] the former test suite should just be making http requests and checking the response [05:03:34] the latter should just use qunit and not worry about http headers [05:07:12] btw the karaga lib i sent you earlier is resource-loader aware and lets you execute JS against the page [05:08:33] ^ yurik [05:28:03] ori-l, thx, i will take a look. For now i will try a simple python script to check everything [05:28:13] will look in karaga shortly [06:34:13] New patchset: Wikinaut; "(bug 47264) SpecialVersion: Add last local update timestamp" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/59373 [07:39:39] New patchset: Tim Starling; "Remove taint support" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62421 [08:06:41] New review: J; "the ogg length is stored in the file properties already." [mediawiki/extensions/Score] (master) - https://gerrit.wikimedia.org/r/61181 [08:31:01] New review: Hashar; "recheck" [mediawiki/extensions/SwiftCloudFiles] (master) - https://gerrit.wikimedia.org/r/54444 [08:31:58] New review: Nemo bis; "Instead of being whacked, could you please backport this to 1.21? *puppy-eyes*" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/58686 [08:33:55] New review: Nemo bis; "@Nikerabbit: what other ways? Docs say there is no proper way to disable special pages, if there is ..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/60394 [08:37:06] New patchset: Physikerwelt; "New rendering option LaTeXML" [mediawiki/extensions/Math] (master) - https://gerrit.wikimedia.org/r/61987 [08:42:03] New review: J; "with the score code from https://test2.wikipedia.org/w/index.php?title=User:JanGerber/score&action=e..." [mediawiki/extensions/Score] (master) - https://gerrit.wikimedia.org/r/62243 [09:03:17] Change abandoned: Hashar; "Abandoning for now, follow up discussion on the bug report https://bugzilla.wikimedia.org/show_bug.c..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/49930 [09:40:07] New review: Nemo bis; "Per Nikerabbit in the original commit and TWN discussion linked there." [mediawiki/core] (master) C: 1; - https://gerrit.wikimedia.org/r/59045 [10:01:17] New review: Aude; "using the revert here won't work because of merge conflicts. needs a fresh commit against current m..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/59045 [10:01:52] Change abandoned: Aude; "merge conflicts too much" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/59045 [10:34:17] New review: Hashar; "On gallium we indeed use a tmpfs for $wgTmpDirectory. That patch would just skip the overhead intro..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/61276 [10:35:17] New review: Hashar; "(note that the above runs have full logs enabled to different files as well as profiling. There is t..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/61276 [10:46:18] New patchset: QChris; "Fix reviewer email notifications for new patch sets" [gerrit] (wmf) - https://gerrit.wikimedia.org/r/62424 [11:02:31] New patchset: Pastakhov; "Refactoring of runtime and adding arrays support (version 0.2.0)" [mediawiki/extensions/Foxway] (master) - https://gerrit.wikimedia.org/r/62217 [11:05:25] Change merged: jenkins-bot; [mediawiki/extensions/Echo] (master) - https://gerrit.wikimedia.org/r/62211 [11:17:48] @seen ^demon [11:18:01] argh, wm-bot doesn't listen here, does it. [11:18:28] what's demon doing, disappearing for a week? [11:27:51] doesn't seem vanished [11:30:50] New patchset: Zfilipin; "Updated Ruby gems" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/62427 [11:39:26] New patchset: Matmarex; "Remove SkinLegacy.php" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62430 [11:45:45] New patchset: Matmarex; "Remove SkinLegacy.php" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62430 [11:48:22] New review: Matmarex; "So, what about this? It's May already." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/58049 [11:49:37] !doc SkinTemplate [11:49:38] An overview of available documentation about MediaWiki can be found at [11:49:43] New review: Nikerabbit; "(1 comment)" [mediawiki/extensions/SemanticForms] (master) - https://gerrit.wikimedia.org/r/62407 [11:49:43] !class SkinTemplate [11:49:43] See https://doc.wikimedia.org/mediawiki-core/master/php/html/classSkinTemplate.html [11:53:13] New review: Matmarex; "Once again, please. I updated the release notes." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62430 [11:54:17] New review: Nikerabbit; "Is the PECL extension of any use?" [mediawiki/core] (master) C: 1; - https://gerrit.wikimedia.org/r/62421 [11:56:39] New review: Nikerabbit; "I usually use SpecialPage_initList hook." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/60394 [12:14:16] New review: PleaseStand; "Nemo bis: No backport needed. REL1_21 already has the special page." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/58686 [12:31:03] New review: Nemo bis; "Ah, sorry then, I must have remembered RC comments incorrectly." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/58686 [12:40:42] New review: Jeroen De Dauw; "(1 comment)" [mediawiki/extensions/SemanticMediaWiki] (master) - https://gerrit.wikimedia.org/r/61179 [12:46:10] New patchset: Amire80; "(bug 47257) RTL proofread action icons are needed" [mediawiki/extensions/Translate] (master) - https://gerrit.wikimedia.org/r/62426 [12:47:49] New review: Amire80; "Looks good to me. I tenamed LTR files to *-ltr.* and updated the CSS files." [mediawiki/extensions/Translate] (master) C: 1; - https://gerrit.wikimedia.org/r/62426 [12:54:51] New patchset: Demon; "Remove pointless vertical alignment" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62431 [13:01:04] New patchset: Demon; "Move DEADLOCK_* constants into DatabaseBase" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62432 [13:03:49] New patchset: Matthias Mullie; "Bug 42057 - Feedback link on article pages" [mediawiki/extensions/ArticleFeedbackv5] (master) - https://gerrit.wikimedia.org/r/35168 [13:03:50] New patchset: Matthias Mullie; "Quickly enable/disable AFTv5" [mediawiki/extensions/ArticleFeedbackv5] (master) - https://gerrit.wikimedia.org/r/61797 [13:07:52] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62431 [13:14:47] New review: Anomie; "Just delete the offending line for 1.20, the variable you're checking wasn't added until 1.21 either." [mediawiki/extensions/Scribunto] (REL1_20) C: -1; - https://gerrit.wikimedia.org/r/62409 [13:18:44] New patchset: Nikerabbit; "Update statsbar when changing group" [mediawiki/extensions/Translate] (master) - https://gerrit.wikimedia.org/r/62345 [13:22:37] New patchset: Demon; "Use __METHOD__ for default $fname params to database functions" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62433 [13:23:35] New patchset: Matthias Mullie; "Quickly enable/disable AFTv5" [mediawiki/extensions/ArticleFeedbackv5] (master) - https://gerrit.wikimedia.org/r/61797 [13:42:38] New review: Nikerabbit; "(1 comment)" [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/62402 [13:43:26] hashar: Hi. It seems some gerrit plugins disappeared from Jenkins over the last weeks (e.g.: gerrit-plugin-replication). Neither ^demon nor I deleted them. Also, they do not show up in the deleted jobs. Any idea where they've gone? (Jenkin's �All configuration history� does not have �Deleted� entries for them) [13:44:23] (For gerrit-plugin-replication, I received the last failure email on 30th April) [13:44:24] qchris: ahgrhg bb [13:44:38] qchris: are jobs showing in the GUI still? [13:44:44] No [13:44:50] for god sake [13:44:54] I need to phase out Jenkins [13:44:57] :-) [13:44:57] and write something else [13:45:11] qchris: I have upgraded Jenkins last week for some security upgrade [13:45:11] So it's ok if I just add them again? [13:45:16] I think it was on May 2nd [13:45:25] the upgrade must have happily deleted the jobs for some reason [13:45:38] and I don't think I have any backup for them :( [13:45:39] Security updates kill our jobs ... so I have to write more secure jobs then :-) [13:46:08] (The Jenkins history has links to old job configurations, but those links do not work) [13:46:10] ahh [13:46:11] gerrit-plugin-replication/ [13:46:15] the directory is there [13:46:24] Oh. All the better :-) [13:47:08] New review: Nikerabbit; "I don't see any difference between before and after this patch." [mediawiki/extensions/Translate] (master) - https://gerrit.wikimedia.org/r/62235 [13:47:13] New patchset: Amire80; "Fixes for empty interlanguage links list" [mediawiki/extensions/UniversalLanguageSelector] (master) - https://gerrit.wikimedia.org/r/62402 [13:54:05] New patchset: Ganeshaditya1; "Bug: 43680 added the MediaWiki logo as the default logo." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62339 [13:55:50] New patchset: Matthias Mullie; "(bug 47733) Word count is wrong, does not recognize non-ASCII characters" [mediawiki/extensions/ArticleFeedbackv5] (master) - https://gerrit.wikimedia.org/r/62435 [13:58:49] New review: Demon; "Duplicate of another change (as mentioned), should be abandoned." [mediawiki/core] (master) C: -2; - https://gerrit.wikimedia.org/r/62339 [14:03:34] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/61739 [14:05:21] who knows about the beta labs project? [14:11:02] Change merged: jenkins-bot; [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62430 [14:12:47] <^demon> Nikerabbit: What's up? [14:14:09] New review: Ganeshaditya1; "@demon do you mean duplicate of 62338 or the one that rjain has submitted. The one that rjain submit..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62339 [14:15:57] ^demon: I'd like to have ULS on sidebar configuration on beta labs (or is there better place) but have no idea how it works [14:16:34] <^demon> It runs on master, and uses the normal wmf-config files like from production. [14:17:04] <^demon> DB names are the same as in production (enwiki, etc) [14:17:14] ^demon: nice... and those files are in the same or different repo? [14:17:18] <^demon> Same repo. [14:18:57] aaah I see [14:19:01] never noticed those before [14:20:20] http://deployment.wikimedia.beta.wmflabs.org/wiki/Main_Page the config links are not working here [14:21:05] <^demon> Hrm, yeah that seems wrong. [14:21:41] so does the setting changes go live automatically after I merge a commit to those files? [14:21:51] New review: Hashar; "I might have find the root cause that prevented Jenkins from finding the unit tests." [mediawiki/extensions/LabeledSectionTransclusion] (master) - https://gerrit.wikimedia.org/r/38114 [14:22:20] New review: Demon; "I'm not sure what the added benefit is of the info box under the place where we already configure th..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62339 [14:24:51] <^demon> Nikerabbit: That part I'm not 100% sure on. I'd *think* so (since code changes do go live). [14:25:03] <^demon> Easy enough to find out after it merges :) [14:26:36] New patchset: Matthias Mullie; "(bug 37917) Ugly bolding of entire section when one "other filter" is selected" [mediawiki/extensions/ArticleFeedbackv5] (master) - https://gerrit.wikimedia.org/r/62437 [14:29:26] New patchset: Alex Monk; "Fix CharInsert not working in LQT" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62438 [14:35:14] New patchset: Matmarex; "display: inline-block for .errorbox, .warningbox, .successbox" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62439 [14:35:25] New review: Brion VIBBER; "J, is that the output from Lilypond or the output from ImageMagick? Try again with $wgScoreTrim = fa..." [mediawiki/extensions/Score] (master) - https://gerrit.wikimedia.org/r/62243 [14:35:54] Nikerabbit: on labs-l I also read that beta just got half as slow as it was (from 500 to 250 ms or something) after it got rid of the horrible glusterfs... hopefully this should make it more useful for ULS testing too [14:36:20] New review: Ganeshaditya1; "Hmm .. The help box doesn't mention how to set the logo later on, like if the admin chooses to ignor..." [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62339 [14:37:26] Nemo_bis: shouldn't affect much [14:37:47] Nikerabbit: never underestimate the brokenness of labs :) [14:37:55] *beta [14:38:03] Nemo_bis: yes I will break it further ;) [14:38:05] first and only time I tried to test something there I got a fatal upon save [14:38:14] New patchset: Reedy; "Update code to match MWSearch updates, fixing fatal" [mediawiki/extensions/ProofreadPage] (master) - https://gerrit.wikimedia.org/r/62440 [14:40:18] New patchset: Reedy; "Update SearchEngine::searchText() and SearchEngine::searchTitle() return type hints" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62441 [14:40:24] lol [14:40:29] do you guys know what the MediaWiki_I18N class does? [14:40:50] * MatmaRex waits with the punchline until someone replies [14:41:09] MatmaRex: I think it may be used by the skins as wfMessage replacement [14:41:39] Nikerabbit: no, it removes all dollar signs in the values passed to it. [14:41:50] and calls wfMessage, yes. [14:42:00] i'm killing it right now ~ [14:42:11] o_O [14:42:13] * brion blames the distant past [14:42:43] yeha [14:42:44] yeah* [14:42:50] brion: it has your comments on it, back from 2004. [14:42:58] (i git-blamed) [14:43:05] sorry! [14:43:26] i have the vague impression it was related to how the original monobook skin was done with a template library we abandoned [14:43:39] yeah [14:43:50] <^demon> brion: phptal! [14:43:51] it has some funny PHPTAL-related comments on it, too [14:43:54] New patchset: Matmarex; "Kill the MediaWiki_I18N class" [mediawiki/core] (master) - https://gerrit.wikimedia.org/r/62442 [14:43:59] heh [14:44:01] like, quote [14:44:26]