[03:51:54] Yippee, build fixed! [03:51:54] Project browsertests-en.wikipedia.beta.wmflabs.org-windows-internet_explorer_10 build #356: FIXED in 24 min: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.beta.wmflabs.org-windows-internet_explorer_10/356/ [04:17:58] Yippee, build fixed! [04:17:58] Project browsertests-en.wikipedia.beta.wmflabs.org-windows-internet_explorer_8 build #422: FIXED in 26 min: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.beta.wmflabs.org-windows-internet_explorer_8/422/ [04:37:10] Yippee, build fixed! [04:37:11] Project browsertests-test2.wikipedia.org-linux-chrome build #785: FIXED in 19 min: https://wmf.ci.cloudbees.com/job/browsertests-test2.wikipedia.org-linux-chrome/785/ [05:31:40] Project browsertests-test2.wikipedia.org-windows-internet_explorer_7 build #603: FAILURE in 27 min: https://wmf.ci.cloudbees.com/job/browsertests-test2.wikipedia.org-windows-internet_explorer_7/603/ [06:45:45] Yippee, build fixed! [06:45:45] Project MobileFrontend-test2.m.wikipedia.org-linux-firefox build #142: FIXED in 18 min: https://wmf.ci.cloudbees.com/job/MobileFrontend-test2.m.wikipedia.org-linux-firefox/142/ [09:19:21] hi zeljkof [09:25:39] hi zeljkof [09:25:53] hi mayankmadan_ [09:26:40] zeljkof, any help with https://gerrit.wikimedia.org/r/#/c/101485/ [09:26:47] mayankmadan_: let me see... [09:28:33] mayankmadan_: did you have a question about the commit? [09:29:30] zeljkof, how can i use regexp in step statements [09:29:53] mayankmadan_: easily :) [09:30:03] :) [09:30:08] where? can you point me to the code that you want to change? [09:30:38] https://gerrit.wikimedia.org/r/#/c/101485/5/features/step_definitions/custom_headers_steps.rb [09:30:42] zeljkof, ^ [09:30:55] looking [09:31:37] ok [09:31:47] are you familiar with regular expressions? [09:32:02] not much [09:33:27] mayankmadan_: ok, so good news is that zero team (yurik) has replied, so now you know what needs to be done, right? [09:33:53] yep [09:34:08] I also do not have a lot of meetings today, so I can help you get the task finished [09:34:19] I should be online for the next 8 or so hours [09:34:27] great [09:34:29] back to regexp-es [09:34:38] did you read this? [09:34:38] http://cukes.info/step-definitions.html [09:34:44] and this? [09:34:49] https://en.wikipedia.org/wiki/Regular_expression [09:35:16] i didnt understand the first link [09:35:20] :) [09:35:22] ok, nevermind [09:35:44] so, in short, regexp is a way to get a piece of text from a string [09:36:08] go to this page [09:36:09] http://rubular.com/ [09:36:38] it says all you need to know, and you can test the code there [09:37:06] thats great [09:37:07] so, back to the commit [09:37:07] https://gerrit.wikimedia.org/r/#/c/101485/5/features/step_definitions/custom_headers_steps.rb [09:37:34] do you know what you need to do now? [09:37:39] or are you still stuck? [09:37:42] i think so [09:37:53] mayankmadan_: ok, then try and ping me if you get stuck [09:38:08] I will be online for some time now [09:38:22] zeljkof, what does (.+) do? [09:38:49] it finds one or more characters [09:38:57] you will see that one a lot in our step files [09:39:09] so in this string "one two three" [09:39:56] zeljkof, i still dont understand whats (.+) for [09:40:01] mayankmadan: ok [09:40:04] in this string [09:40:08] "one two three" [09:40:17] you can get "two" out like this [09:41:13] "/one .+ three/" (without quotes, had to put them in for irc) [09:41:21] try it on rubular.com [09:42:19] in cucumber, regular expressions are used so one step in a step file could be executed for several steps in feature file [09:42:34] let's say you have these steps [09:42:43] given I log in as zeljko [09:42:45] and [09:42:49] and another step [09:42:58] given I log in as mayank [09:43:15] you would have to create two steps in the step file, right? [09:43:38] but if the implementation of the step looks like this: [09:43:59] on(LoginPage).login_as(user) [09:44:22] you can reuse the same step implementation to login both as me and you [09:44:23] like this [09:45:29] Given(/^I log in as (.+)$/) do |username| [09:45:46] do you understand how to use regexp now? [09:45:52] i do [09:46:19] but arent the step step statements going to be hard coded [09:46:31] what do you mean? [09:46:34] you will have to change the statement for different steps? [09:46:45] I do not understand [09:46:53] can you post the code in the pad? [09:47:27] zeljkof, is there going to be only one scenario for all the tests? [09:47:46] mayankmadan: I still do not understand, can you show me the code? [09:48:48] zeljkof, i havent made any changes yet [09:49:09] just make the change, push the code and we can discuss it in gerrit [09:49:23] it is hard for me to answer the question if I do not see the code [09:49:55] have a look at the feature file [09:50:29] zeljkof, does it need to have only 1 scenario? [09:50:38] what? [09:50:49] what needs to have only 1 scenario? [09:51:05] the whole test :) [09:51:20] mayankmadan: for this task to be finished? [09:51:27] or in general? [09:51:33] or in general [09:51:39] *in general [09:51:46] ah [09:51:52] its difficult to explain [09:51:56] let me just make the changes [09:52:06] I think I know what you mean [09:52:16] no, the feature still has to have two scenarios [09:52:32] zeljkof, then what is the use of regexps [09:52:38] but yurik said you can reuse steps [09:52:45] let me show you what I mean [09:52:55] I will comment in gerrit [09:54:21] mayankmadan: ok, the feature file looks good, you do not have to change anything there [09:55:28] steps file does need to change, but a little [09:59:18] zeljkof, now i know why the change is required [09:59:27] ok, great [09:59:32] I am just finishing my comment [10:00:18] by using regexps there just needs to be definition for 1 step that can be used for every scenario, right? [10:00:29] (03CR) 10Zfilipin: [C: 04-1] Adding a test to check if supplying custom headers works (031 comment) [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 (owner: 10Mayankmadan) [10:00:39] mayankmadan: take a look at my comment [10:01:19] zeljkof, i was right too :) [10:01:44] mayankmadan: great [10:01:57] take a look at the steps file and make the change [10:20:12] (03CR) 10Yurik: "Thanks Zfilipin, that's exactly what i meant in my comment on line 19 :)" [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 (owner: 10Mayankmadan) [10:24:20] (03PS6) 10Mayankmadan: Adding a test to check if supplying custom headers works [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 [10:24:29] zeljkof, the code i pushed doesnt work [10:24:41] can you check it? [10:27:10] mayankmadan: let me see [10:28:57] (03CR) 10Zfilipin: "Yurik : Mayankmadan and I had a long chat in #wikimedia-qa and I have realized he did not understood you, so I have decided to be as expli" [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 (owner: 10Mayankmadan) [10:29:38] mayankmadan: what is the error message? [10:30:26] zeljkof, pad [10:30:58] ok, looking [10:31:12] by the way, yurik is at #wikimedia-mobile, if you have questions for him [10:31:41] mayankmadan: strange, the error message does not say what the problem is [10:31:55] try this [10:32:04] $ bundle exec cucumber -f pretty features/custom_headers.feature [10:32:13] that will output more information [10:35:21] zeljkof, pad [10:36:20] (03CR) 10Yurik: [C: 04-1] "Inline comments. And Zfilipin, thanks for explaining!" (032 comments) [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 (owner: 10Mayankmadan) [10:37:05] mayankmadan: ok, this is the problem [10:37:06] undefined method `close' for nil:NilClass (NoMethodError) [10:37:37] you are closing the browser with @browser.close, but @browser is not initialized [10:39:43] zeljkof, it should be initialized [10:39:51] mayankmadan: but it is not :) [10:39:56] given step is executing first :) [10:40:11] mayankmadan: did you see comment that yurik made? [10:40:16] yep [10:40:26] implementing that change could help [10:40:28] ill do that after fixing this problem [10:40:38] and I will make another comment now [10:40:40] zeljkof, how? [10:41:21] i get it [10:42:10] (03CR) 10Zfilipin: [C: 04-1] Adding a test to check if supplying custom headers works (031 comment) [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 (owner: 10Mayankmadan) [10:42:15] mayankmadan: see my comment [10:42:37] @browser has to be defined in the step file, not in a separate method, I think that is the problem [10:43:30] (03CR) 10Mayankmadan: Adding a test to check if supplying custom headers works (031 comment) [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 (owner: 10Mayankmadan) [10:43:43] zeljkof, Yuri wanted it the other way [10:43:53] i had a long chat with him yesterday [10:44:01] or day before yesterday [10:44:53] trust me, just make the change :) [10:52:58] (03CR) 10Yurik: Adding a test to check if supplying custom headers works (031 comment) [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 (owner: 10Mayankmadan) [10:53:44] mayankmadan, hi, just replied to the patch [10:53:49] there are a few comments, for both files [10:53:52] welcome yurik :) [10:54:00] thx :) [10:54:12] URL is fine, probably even better than constructing it on the fly [10:55:24] as long as there are no code dups, its good - DRY is my god :) [10:55:25] https://en.wikipedia.org/wiki/Don%27t_repeat_yourself [10:55:54] yurik: would you consider moving the test to zero repository? [10:56:01] or is there such a thing? [10:56:30] this? https://gerrit.wikimedia.org/r/#/admin/projects/mediawiki/extensions/ZeroRatedMobileAccess [10:56:49] zeljkof, we have a /maintenance dir in zero, but we should probably have a separate repo... as it gets synced to prod servers... [10:57:12] than again, everyone has their code mixed in, guess it should be configured at the apache level [10:57:29] yes please, we need a tests dir :) [10:57:46] tests/cucumber... [10:57:47] is ZeroRatedMobileAccess the correct repo? [10:57:50] yep [10:57:57] it is usually tests/browser [10:58:08] yurik: https://github.com/wikimedia/mediawiki-selenium [10:58:18] even better link https://github.com/wikimedia/mediawiki-selenium#links [10:58:19] sure thing :) ok, i'm delirious now, 6am, need to get some sleep before big day tomorrow :) [10:58:33] zeljkof, you have my 100% support in all such matters :) [10:58:43] yurik: ok, when the code is ready, I will make the commit and add you as a reviewer [10:58:50] thx :) [10:58:58] make sure it's DRY :))) [10:59:02] zeljkof, still not working [10:59:14] yurik: it's DRY way or highway :) [10:59:19] exactly! [10:59:27] something i keep trying to explain to my coworker ;) [10:59:51] btw, do you guys have the whole thing set up as vagrant role? [10:59:57] zeljkof, i pasted it on the pad [11:00:02] yurik: of course, browsertests role [11:00:15] does it require mac? [11:00:23] yurik: of course not :) [11:00:32] but not sure if vagrant work on windows [11:00:34] because i remember something like that requiring host to be mac [11:00:38] sure does [11:00:41] all my work is in it [11:00:53] yurik: I do not think it is needed any more, it might be in the past [11:01:09] yurik: then it should just work :) [11:01:20] hmmm... there are some issues driving host's browser from VM [11:01:21] it might need some tweaks [11:01:39] yurik: in this case we are driving headless browser (phantomjs) so that should not be the problem :) [11:01:47] awesome :) [11:02:01] but is the phantomjs still on the host? [11:02:09] i would prefer to have tho whole thing on VM [11:02:13] this way - no dependencies [11:02:38] yurik: the role does not install phantomjs in the vm yet, but that should be trivial to do [11:02:41] VM->host communication is evil [11:02:57] sounds good, thx!!!! [11:03:04] yurik: good night [11:03:10] i'm sooo excited!!!! we are getting real tests! just like the big boys [11:03:18] thx for driving this [11:03:25] yurik: we already have 10 or so repos with selenium tests [11:03:39] moving slowly, but moving [11:03:48] mayankmadan: looking [11:03:54] zeljkof, its working now, just needed to restart terminal [11:04:00] mayankmadan: great [11:04:11] zeljkof, no its not [11:04:15] i thought it was [11:04:18] but its not [11:04:19] mayankmadan: is everything up in gerrit? [11:04:24] something tells me i will give up at some point, and do a patch for selenium to support header setting on all browsers [11:04:42] gnight! [11:04:49] good night yurik [11:04:56] zeljkof, not yet [11:04:59] yurik: it might even be supported now, I remember seeing something about that [11:05:23] mayankmadan: push the code and add the error message to the pad [11:05:34] zeljkof no it is working [11:05:35] zeljkof, i thought devs were against it according to that huge bug discussion [11:05:41] :) [11:05:50] problem is - they mixed setting and getting, which are different [11:05:59] zeljkof, forgot to add a .m in url [11:06:00] and require very different things [11:06:08] yurik: I remember somebody changing mind [11:06:21] something big must have died somewhere [11:06:24] yurik: or realizing the difference [11:07:01] yurik: anyway, we have it working for phantomjs, I can play with other browsers when this is finalized [11:07:20] yurik: that even sounds like a good google code-in task for mayankmadan [11:07:23] :) [11:07:44] mayankmadan, how good are you c# skills? ;) [11:08:05] yurik, i am good with c++ but not with c# :) [11:08:14] yurik: he did not know any ruby a few weeks ago, c# is just another language, right :) [11:08:26] well, yes it is [11:08:39] the more i learn, the better this year gets :) [11:08:42] C# is easy, similar to java, and MUCH easier than ruby [11:09:03] mayankmadan, one thing i promise you - i won't let the code in until i like it :D [11:09:07] yurik: but ruby is so easy :) how can c# be easier? [11:09:30] yurik: oh, he is used to that with us :) [11:09:51] zeljkof, yurik finally all tests are working [11:09:51] c# follows most of the C++ / java / javascript / dart / and lots of other language conventions [11:10:20] mayankmadan: great, push the code to gerrit [11:10:24] * yurik i'l believe it when i see it ... grumble grumble grumble [11:10:28] yurik: yes, ruby has a few unusal things [11:10:49] a few? how about ass-backwards syntax for everything? :) [11:11:10] yurik, what does ass-backwards mean? [11:11:26] yurik: but it is so natural (when you get used to it) :) [11:11:35] its a highly technical expression [11:11:38] like upside-down scrolling on mac (that I hate) [11:11:42] i think that syntax is what kept ruby from becoming mainstream [11:12:04] people are mostly used to the C notation [11:12:13] php is okeish for perl ppl [11:12:24] and at the time there were no good alternatives [11:12:32] hence we are stuck with this language zoo :) [11:12:42] python i think is the one that really nailed alternative style [11:13:18] (03PS7) 10Mayankmadan: Adding a test to check if supplying custom headers works [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 [11:13:28] zeljkof, yurik the code is pushed [11:13:51] to gerrut [11:13:54] *gerrit [11:14:02] mayankmadan, just a convention thing btw - its better to put bug number very first in the comment [11:14:11] this way it shows in all one-line views [11:14:28] i.e. (bug 48218) Adding a .... [11:14:41] all the things you learn :) [11:15:07] yurik: we do not follow that convention [11:15:21] but I will keep it in mind when pushing code to zero repo :) [11:16:18] zeljkof, will the task be considered complete if this is ok? [11:16:25] mayankmadan, "Free wikipedia ..." should also be part of the text [11:16:38] carriers could change the whole thing [11:16:46] ok [11:16:46] minor thing really [11:17:04] so just extend it to after "Banner should be " [11:17:19] me like!!! [11:17:55] generic question - what is the "Scenario" text do? [11:18:03] just a comment? [11:18:04] mayankmadan: as far as I am concerned, yes [11:18:25] yurik: yes, just a name, does not get executed [11:19:24] zeljkof, btw, according to that page i found, I can use any backend scripting i want, not just ruby - they use C# and python and other langs... [11:20:23] http://cukes.info/step-definitions.html [11:20:28] yurik: yes, there is a lot of supported languages [11:20:38] sweet :) [11:20:44] but make sure the language has selenium bindings [11:20:45] (03PS8) 10Mayankmadan: Adding a test to check if supplying custom headers works [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 [11:20:53] c# has both, I think [11:21:05] and python [11:21:32] zeljkof, yurik how about now? [11:21:46] mayankmadan: let me see [11:22:14] i will have to try it myself tomorrow (i really hope installing this phantumjs is easy in the VM) [11:22:19] but looks good to me :) [11:22:38] comment out (or make it parameter-based) saving of the image [11:23:00] yurik: https://github.com/watir/watirbook/blob/master/installation/ubuntu.md#phantomjs [11:23:02] like " (and save image)? " [11:23:28] i was soo hoping for it to be part of the puppet :) [11:23:36] * yurik is lazy, and is still not in bed [11:23:51] yurik: I do not think it is yet, but it should be easy to add [11:24:13] will poke at it tomorrow. Thank you!!!! both of you! [11:24:24] (or was there someone else too/) [11:24:25] ? [11:25:00] yurik: I think just mayankmadan and me are here now :) [11:29:02] (03CR) 10Zfilipin: [C: 04-1] Adding a test to check if supplying custom headers works (031 comment) [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 (owner: 10Mayankmadan) [11:29:20] mayankmadan: one final thing to do and this is done! :) [11:31:06] (03PS9) 10Mayankmadan: Adding a test to check if supplying custom headers works [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 [11:31:20] how about now zeljkof [11:31:31] mayankmadan: let me see [11:31:36] (and run on my machine) [11:41:16] mayankmadan: looks good to me! [11:41:27] I will close the task at google site [11:42:05] mayankmadan: please mark the task as finished [11:42:08] (at google site) [11:44:58] zeljkof, marked [11:45:38] zeljkof, wanted to do a quick documentation task [11:45:44] is there any availaible [11:45:52] *available [11:46:21] (03PS10) 10Zfilipin: Adding a test to check if supplying custom headers works [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 (owner: 10Mayankmadan) [11:46:34] (03CR) 10Zfilipin: [C: 031] WIP: Adding a test to check if supplying custom headers works [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101485 (owner: 10Mayankmadan) [11:47:20] mayankmadan: there is! :) [11:47:20] https://www.google-melange.com/gci/task/view/google/gci2013/5888156557639680 [11:48:05] mayankmadan: looks interesting? [11:48:35] zeljkof, i dont think i can do this one [11:49:07] mayankmadan: why? [11:51:43] zeljkof, i already have a pretty messed up setup [11:52:02] mayankmadan: you can always install a virtual machine and test there :) [11:52:13] and vagrant part is in a virtual machine anyway [11:52:27] ok then [11:53:21] when there is will, there is a way :) [11:53:25] true [11:53:39] zeljkof, will try my best to complete this in a few hours [11:54:09] mayankmadan: I should be online for a few more, and chris and jeff will be online after that [11:54:12] so go ahead [11:54:19] okey doke [11:54:19] claim the task, I can assign it to you [11:54:28] already claimed [11:55:17] and assigned :) good luck [12:56:15] zeljkof, i have a question [12:57:02] you online [12:57:09] ? [12:57:33] mayankmadan: in a meeting, for 30 minutes more or so [13:00:07] feel free to ask, I will reply later [13:35:09] mayankmadan: I am back, if you have questions [13:48:14] (03PS1) 10Hashar: make gate-and-submit a dependent pipeline [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/101839 [13:48:33] (03CR) 10Hashar: [C: 032] make gate-and-submit a dependent pipeline [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/101839 (owner: 10Hashar) [13:48:42] (03Merged) 10jenkins-bot: make gate-and-submit a dependent pipeline [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/101839 (owner: 10Hashar) [13:49:43] zeljkof, what is the size of the box that vagrant downloads [13:51:15] mayankmadan: I do not know, let me check [13:53:41] (03PS1) 10Zfilipin: Use ordered list when listing repositories that use the gem [selenium] - 10https://gerrit.wikimedia.org/r/101840 [13:58:19] (03CR) 10Nikerabbit: [C: 032] Use ordered list when listing repositories that use the gem [selenium] - 10https://gerrit.wikimedia.org/r/101840 (owner: 10Zfilipin) [13:58:32] (03Merged) 10jenkins-bot: Use ordered list when listing repositories that use the gem [selenium] - 10https://gerrit.wikimedia.org/r/101840 (owner: 10Zfilipin) [14:00:57] mayankmadan: my vagrant box is 1.22 GB [14:01:13] damn [14:01:21] let me check what is downloaded [14:01:36] will need to use gparted to expand my ubuntu partition [14:03:52] good luck :) [14:07:14] zeljkof, so how much is downloaded? [14:07:30] mayankmadan: oops, forgot to check [14:07:54] checking [14:08:53] this is downloaded [14:08:53] https://cloud-images.ubuntu.com/vagrant/precise/current/precise-server-cloudimg-amd64-vagrant-disk1.box [14:08:58] 343 MB [14:09:33] https://github.com/wikimedia/mediawiki-vagrant/blob/master/Vagrantfile#L39 [14:09:38] thats a lot:) [14:09:52] it is an entire ubuntu virtual machine :) [14:21:29] (03PS1) 10Hashar: pep8/pyflakes made concurrents [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101852 [14:21:30] (03PS1) 10Hashar: various linting jobs made concurrents [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101853 [14:21:44] (03CR) 10Hashar: [C: 032] various linting jobs made concurrents [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101853 (owner: 10Hashar) [14:21:45] (03CR) 10Hashar: [C: 032] pep8/pyflakes made concurrents [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101852 (owner: 10Hashar) [14:24:16] (03Restored) 10Zfilipin: WIP Test central-domain login, cross-domain login and central login for anons [qa/browsertests] - 10https://gerrit.wikimedia.org/r/74128 (owner: 10Zfilipin) [14:26:45] (03Merged) 10jenkins-bot: pep8/pyflakes made concurrents [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101852 (owner: 10Hashar) [14:27:15] (03Merged) 10jenkins-bot: various linting jobs made concurrents [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101853 (owner: 10Hashar) [14:27:42] (03PS1) 10Hashar: mwext-VisualEditor-doc-test made concurrent [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101854 [14:27:51] (03CR) 10Hashar: [C: 032] mwext-VisualEditor-doc-test made concurrent [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101854 (owner: 10Hashar) [14:28:13] (03Merged) 10jenkins-bot: mwext-VisualEditor-doc-test made concurrent [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101854 (owner: 10Hashar) [15:02:07] (03PS1) 10Hashar: operations-puppet-* made concurrents [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101865 [15:04:53] (03PS1) 10Hashar: mwext linting jobs made concurrents [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101866 [15:06:57] (03CR) 10Hashar: [C: 032] mwext linting jobs made concurrents [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101866 (owner: 10Hashar) [15:06:59] (03CR) 10Hashar: [C: 032] operations-puppet-* made concurrents [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101865 (owner: 10Hashar) [15:07:17] (03Merged) 10jenkins-bot: operations-puppet-* made concurrents [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101865 (owner: 10Hashar) [15:07:33] (03Merged) 10jenkins-bot: mwext linting jobs made concurrents [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101866 (owner: 10Hashar) [15:29:49] zeljkof: contact me when you can [15:31:39] vandrew_, zeljkof is not online [15:32:16] vandrew_, you can ask anyone else [15:32:40] or you can write about your problem here? [15:33:16] mayankmadan_: hey, thanks, I'll ask right away [15:33:35] vandrew_, so what is it? [15:34:10] mayankmadan_: This is a general step not specific to the delete-item-feature and can be used by other tests too, so please move this step to entity_steps.rb I'm a little confused by this [15:34:21] should I remove the step from the file [15:34:35] vandrew_, some more context? [15:34:47] mayankmadan_: https://gerrit.wikimedia.org/r/#/c/100591/8/tests/browser/features/step_definitions/edit_tab_steps.rb gere [15:35:56] vandrew_, is which comment is troubling you? [15:36:20] *which comment is troubling you? [15:36:51] mayankmadan_: wrong file, here https://gerrit.wikimedia.org/r/#/c/100591/8/tests/browser/features/step_definitions/delete_item_steps.rb and is the first commit [15:36:54] comment* [15:44:55] zeljkof: umm, are you here ? [15:45:01] vandrew_: I am [15:46:15] vandrew_: I think I can help here. Every test can use any step in any of the foo_steps.rb files. So if a step already exists, then we don't want to write it a second time, we want to use it over and over. [15:47:00] vandrew_, isnt i am logged in defined in upload_wizard_steps.rb? [15:47:31] vandrew_: in the case of delete_item_steps, you have written a step that already exists, "I am on an item page" in the entity_steps.rb file. Make sense? [15:47:50] mayankmadan_: vandrew_ is working in a different repo, not /qa/browsertests [15:48:08] oh yeah [15:48:11] just saw that [15:48:33] chrismcmahon: how do I use these already made steps ? [15:49:22] chrismcmahon: copy paste the content yes ? [15:49:22] vandrew_: simply use "Given I am on an item page" in your .feature file, and the framework will automatically pick up and use the corresponding step from *any* _steps.rb file [15:51:41] chrismcmahon: when i use cucumber delete_item_steps.rb I get pending # express the regexp above with the code you wish you... [15:55:44] vandrew_: and... [15:56:02] chrismcmahon: so I moved the login step the entity_steps.rb now how do I use it in delete_item_steps.rb ? [15:56:47] vandrew_: do you know the tool called "grep"? [15:57:30] chrismcmahon: yes [15:57:49] vandrew_: I think what you are asking is how to use the existing login step in your test. is that true? [15:58:25] chrismcmahon: yes [15:59:57] vandrew_: in your .feature just use exactly the text of the step you want to make use of, so "Given I am logged in" or whatever the text is. There is no need to write a new step for that "Given" statement. [16:01:02] chrismcmahon: oh, thanks so for example I should delete entity_steps.rb from the *_steps.rb file [16:01:15] wrong paste [16:01:18] Given(/^I am on an item page$/) do [16:01:19] zeljkof, how much time does vagrant up usually takes after downloading box? [16:01:30] zeljkof, its been about an hour and a half [16:02:05] vandrew_: that sounds right. (I'm looking in the Wikibase code right now...) [16:02:25] mayankmadan_: depending on your machine [16:02:35] it takes 10 or so minutes on my machine [16:02:46] but I think on an older machine with less ram it took long [16:02:53] 30-60 minutes [16:03:29] 90 minutes is pretty long, but if it still working, then it is not done yet [16:03:30] :( [16:03:41] zeljkof: how do I check if an item page still exists ? [16:03:44] vandrew_: if you do "/Wikibase/tests/browser$ grep -r 'I am on an item page' *" you can see all the .feature files that re-use that one step. [16:03:54] zeljkof, well my machine has an i5 processor and 8gb of ram [16:04:00] how much time should it take on that? [16:04:09] mayankmadan_: hm, strange, that sounds like a good machine [16:04:16] it should take less [16:04:20] damn [16:04:35] but depends on your internet, it might take long to download the image [16:05:02] vandrew_: I have to go now [16:05:13] the rest of the people in the channel should be able to help you [16:23:09] vandrew_, im back [16:23:10] can you post me that screenshot? [16:23:13] of your irc window? [16:23:32] vandrew_, no need [16:23:36] everyone is back [16:24:51] mayankmadan: just a second [16:24:57] (03PS1) 10Hashar: translatewiki jobs made concurrent [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101886 [16:24:58] mayankmadan: oh ok then [16:25:01] vandrew_, no need of the screenshot [16:25:14] (03CR) 10Hashar: [C: 032] translatewiki jobs made concurrent [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101886 (owner: 10Hashar) [16:25:33] (03Merged) 10jenkins-bot: translatewiki jobs made concurrent [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101886 (owner: 10Hashar) [16:25:54] JeffHall, im doing this task https://www.google-melange.com/gci/task/view/google/gci2013/5888156557639680 [16:26:30] chrismcmahonbrb, ^ [16:26:34] mayankmadan: OK, do you have a question about that task? [16:26:41] i have setup virtual machine [16:26:48] done everything it says [16:26:54] in linux section too [16:27:00] but there doesnt seem to be any error [16:27:33] JeffHall, i dont have access to macosx or windows so i did the task with linux [16:27:55] mayankmadan: As far as I know, Linux should work fine [16:28:14] JeffHall, what should i update [16:28:18] everything is updated [16:29:01] mayankmadan: Ah, I see what you mean. Give me a few minutes to review that wiki page... [16:31:22] (03PS1) 10Hashar: integration/zuul.git related jobs made concurrent [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101887 [16:31:47] (03CR) 10Hashar: [C: 032] integration/zuul.git related jobs made concurrent [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101887 (owner: 10Hashar) [16:32:03] (03Merged) 10jenkins-bot: integration/zuul.git related jobs made concurrent [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101887 (owner: 10Hashar) [16:35:33] mayankmadan: Yeah, that page looks pretty good - only thing Linux-related that might need updating is the "Tested on Ubuntu 13.04 32-bit." statement - if you tested with a different version of Linux, it would be good to update that statement [16:36:13] JeffHall, i tested with 64-bit [16:36:51] mayankmadan: Yeah, so if it worked fine, can you add a note to the page? That way, we'll know that the process has been tested on 64-bit linux [16:39:05] JeffHall, i changed that in virtual machine section too [16:39:46] JeffHall, https://www.google-melange.com/gci/task/view/google/gci2013/5888156557639680 [16:39:51] the task should be complete [16:40:05] mayankmadan: Great! Yeah, I believe you're right that we can complete that task [16:40:10] yes [16:40:35] yeah! [16:40:39] 16 tasks done [16:41:04] mayankmadan: OK, task closed, and I'll close the bug as well. Great work! [16:41:21] thank you :) [16:42:27] JeffHall: how do I go to a non existing item page ? [16:42:48] I get uninitialized constant NonExistingItemPage (NameError) [16:43:05] vandrew_, how about DoesNotExistPage page object [16:43:35] vandrew_: Can you give me some more context? [16:44:38] JeffHall: here is the old test https://git.wikimedia.org/blob/mediawiki%2Fextensions%2FWikibase/129f22419945ad414fda6701e91b25af9b6eb973/repo%2Ftests%2Fselenium%2Fitem%2Fnon_existing_item_spec.rb [16:44:58] I have to go to an item page that does not exist [16:46:50] vandrew_: Ah, OK so you can just make up a page name, and when you try to navigate to it, the MediaWiki system should prompt you and ask if you want to create a new page with that name [16:47:53] JeffHall: something like visit_page("badfsfdr") ? [16:48:11] vandrew_: there are examples for using pages that do not exist in the /qa/browsertests repo if there are not any right now in /mediawiki/extensions/Wikibase [16:49:03] vandrew_: we generally name pages that do not exist after a random 10-character string generated as a random number at run time. [16:49:03] chrismcmahon: found it in selenium/lib/pages [16:49:14] vandrew_: OK, great. [16:50:51] whats wrong with DoesNotExistPage? [16:51:07] mayankmadan: doesn't exist [16:53:22] (03PS1) 10Cmcmahon: Move page is failing in IE a lot, this should help [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101888 [16:55:13] JeffHall: I like refactoring: https://gerrit.wikimedia.org/r/#/c/101888/ :-) [16:55:40] turning builds back to green, one issue at a time. :-) [16:56:21] (03CR) 10Jhall: [C: 032] Move page is failing in IE a lot, this should help [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101888 (owner: 10Cmcmahon) [16:56:25] chrismcmahon: Makes sense to me! The Visual Editor browser tests are getting so close to being all green I can almost taste it…. [16:57:58] (03Merged) 10jenkins-bot: Move page is failing in IE a lot, this should help [qa/browsertests] - 10https://gerrit.wikimedia.org/r/101888 (owner: 10Cmcmahon) [17:00:12] JeffHall, wanted to do this task http://www.google-melange.com/gci/task/view/google/gci2013/6157002451451904 [17:01:23] mayankmadan: OK, I can assign it to you, but you have to claim it first I think [17:01:39] JeffHall, but can you first explain me the task a bit [17:01:42] what is jenkins [17:02:16] mayankmadan: Jenkins is our continuous integration system: https://wmf.ci.cloudbees.com [17:02:22] what is meant by change jenkins email notification [17:02:28] which email notification? [17:03:21] mayankmadan: When jobs run in Jenkins, they can optionally send e-mail to whatever addresses you specify. The goal of the task is to improve the e-mail template that is used for those notifications [17:04:19] mayankmadan: JeffHall: some background: http://www.google-melange.com/gci/task/view/google/gci2013/5243150582939648 [17:16:26] chrismcmahon: Oddly enough, I can repro this bug with the Java Selenium bindings using Firefox: http://code.google.com/p/chromedriver/issues/detail?id=636 [17:17:32] JeffHall: with Firefox, even. if we could get a repro manually, we'd have a great VisualEditor bug, but I have never seen the bug in manual operation [17:18:23] chrismcmahon: Yeah, it's an odd one. I'll add my Java code to that bug just so that the Chromedriver folks know it's not unique to the Ruby bindings [17:22:13] mayankmadan, want to do more testing for zero? :) [17:22:16] https://www.mediawiki.org/wiki/Wikipedia_Zero/Test_cases [17:23:12] thanks JeffHall. do you ever hang out in the #selenium channel on IRC? [17:23:58] chrismcmahon: No, but I probably should join that channel [17:24:14] JeffHall: I do from time to time. [17:25:22] JeffHall: if you have some really specific Se question it is a good place to go. it's also often the best way to contact maintainers and stuff [17:26:19] JeffHall: e.g. when you add your update to the 636 issue it might be worth pointing out to one of the Chromedriver maintainers that hangs out there [17:28:12] yurikL yes [17:28:17] *yurik: yes [17:28:18] please [17:28:26] more tasks, more tasks, more tasks [17:28:39] yurik, you still online? [17:29:09] mayankmadan, did you look at that link? [17:29:52] yurik, i dont understand [17:29:57] the links are already created [17:30:02] *the tests [17:30:03] mayankmadan, https://www.mediawiki.org/wiki/Wikipedia_Zero/Test_cases [17:30:20] yurik, in that page, the tests are already there [17:30:24] what do i need to do? [17:30:31] mayankmadan, those are wishes :) [17:30:39] they are not implemented [17:30:43] ok [17:31:18] mayankmadan, ask adam on mobile channel - his nick is dr0ptp4kt [17:31:35] yurik, what is meant by zero rated version of wikipedia or m.wikipedia? [17:31:39] yurik, what do i ask him? [17:31:55] he implemented some testing in PHP [17:33:01] yurik, but what should i ask him? [17:33:10] :) [17:34:09] yurik, are you a mentor in GCI ? [17:34:40] mayankmadan, i'm everything ! :D [17:35:15] yurik, everything? [17:35:46] sure - one butt, hundreds of chairs :( [17:36:18] yurik, so you are a mentor :) [17:36:24] can you create a task? [18:16:47] chrismcmahon: I'm seeing Chrome via SauceLabs starting to display a "save password" prompt: https://saucelabs.com/jobs/c2cfec1d3f2c47c99eb0a6cafb031875/0007screenshot.png [18:17:19] chrismcmahon: Not sure that it's actually causing a problem, but I haven't noticed this before - have you ever seen this behavior? [18:25:48] (03CR) 10Subramanya Sastry: "Those tests are in a different repo (will check with hashar where they are) and those tests should be fixed, but it won't be in this zuul " [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/101460 (owner: 10Subramanya Sastry) [18:31:52] JeffHall: hmm, not sure [18:32:21] JeffHall: I *thought* that those never used to show up because Se just handled them magically [18:33:18] chrismcmahon: Apparently there's an "incognito" command-line option for Chrome that should suppress those, so perhaps we need to modify the mediawiki-selenium gem to pass that option? [18:34:35] JeffHall: I'd say do it if and when tests fail because of it [18:35:32] chrismcmahon: OK, that does make sense - I'll keep it on the back burner until I'm sure it's actually causing a test failure! [18:37:58] JeffHall: yikes, are elements really taking that long to show up? https://gerrit.wikimedia.org/r/#/c/101907/1/modules/ve-mw/test/browser/features/step_definitions/general_markup_steps.rb [18:38:12] JeffHall: longer than 5 seconds? [18:39:11] chrismcmahon: Yep, according to the CloudBees test results. When I run the same test locally, the timeout problem never occurs, so I can only assume it's a consequence of SauceLabs sluggishness [18:41:43] chrismcmahon: For what it's worth, we frequently saw the same problem with Chrome run via SauceLabs in my previous job (using Java Selenium bindings). Suggestion from SauceLabs helpdesk was to add more implicit waits to the tests! [18:43:13] JeffHall: so what I'm seeing is that in Firefox the tests fail because of this bug: https://bugzilla.wikimedia.org/show_bug.cgi?id=58406 . in Chrome we're getting a weird "No data received" page e.g. https://wmf.ci.cloudbees.com/job/VisualEditor-en.wikipedia.beta.wmflabs.org-linux-chrome/lastCompletedBuild/testReport/(root)/VisualEditor%20general%20text%20markup%20features/VisualEditor_general_markup__outline_example_____Underline____u_Th [18:44:39] chrismcmahon: Yeah, that "No data received" is pretty vexing, since it suggests that the Selenium server running via SauceLabs was briefly unavailable [18:45:09] JeffHall: well, today's build just kicked off, let's see what it does [18:47:37] chrismcmahon: Makes sense - might have just been some temporary environmental issues on the SauceLabs end of things [18:59:03] (03PS2) 10Krinkle: Added Jenkins tests for commits to new Parsoid repo. [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/101460 (owner: 10Subramanya Sastry) [19:00:50] (03CR) 10Krinkle: "Note that you probably can't re-use the same jobs (mwext-Parsoid-*) as the current configuration may only do the root git init once, and t" [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/101460 (owner: 10Subramanya Sastry) [19:01:18] (03PS2) 10Krinkle: Make mwext-Scribunto-jslint voting [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/100712 [19:01:27] (03CR) 10Krinkle: [C: 032] Make mwext-Scribunto-jslint voting [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/100712 (owner: 10Krinkle) [19:01:41] (03Merged) 10jenkins-bot: Make mwext-Scribunto-jslint voting [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/100712 (owner: 10Krinkle) [19:01:58] Yippee, build fixed! [19:01:59] Project browsertests-en.wikipedia.beta.wmflabs.org-windows-internet_explorer_7 build #421: FIXED in 30 min: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.beta.wmflabs.org-windows-internet_explorer_7/421/ [19:08:17] (03CR) 10Subramanya Sastry: "oh crap .. ok, i'll not modify the original scripts in integration/jenkins-job-builder-config then. I'll copy and modify. Can delete old j" [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/101460 (owner: 10Subramanya Sastry) [19:23:28] JeffHall, any new tasks? [19:24:47] Project browsertests-en.wikipedia.beta.wmflabs.org-windows-internet_explorer_8 build #423: FAILURE in 29 min: https://wmf.ci.cloudbees.com/job/browsertests-en.wikipedia.beta.wmflabs.org-windows-internet_explorer_8/423/ [19:25:14] mayankmadan: No, nothing at this point (other than the Jenkins and PhantomJS issues that are unclaimed) [19:26:01] JeffHall, im gonna do phantomjs one [19:26:16] mayankmadan: OK, sounds good! [19:29:04] (03PS1) 10Subramanya Sastry: Added new jenkins jobs file for /mediawiki/services/parsoid repo [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101912 [19:30:23] (03PS3) 10Subramanya Sastry: Added Jenkins tests for commits to new Parsoid repo. [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/101460 [19:31:12] (03CR) 10Subramanya Sastry: "https://gerrit.wikimedia.org/r/#/c/101460/ has changes to zuul-config" [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101912 (owner: 10Subramanya Sastry) [19:32:28] (03CR) 10Subramanya Sastry: "https://gerrit.wikimedia.org/r/#/c/101912/ has new jenkins tests referenced in the config file here." [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/101460 (owner: 10Subramanya Sastry) [20:00:10] JeffHall, will need some help [20:01:01] why exactly do some tests fails in phantomjs [20:01:11] i mean what is the reason behind them failing? [20:05:54] mayankmadan: It actually looks like Željko resolved the bug associated with that GCI task: https://bugzilla.wikimedia.org/show_bug.cgi?id=49813 [20:06:33] JeffHall, I dont understand [20:06:37] mayankmadan: So I'm a little unclear on what the status is - it may be that he resolved the PhantomJS failures and forgot to close out the associated GCI task [20:07:12] JeffHall, there are still errors while i try to execute tests on phantomjs [20:07:25] mayankmadan: What errors are you getting? [20:08:06] JeffHall, Failing Scenarios: [20:08:07] cucumber features/upload_wizard.feature:28 # Scenario: Navigate to Release rights page [20:08:07] cucumber features/upload_wizard.feature:36 # Scenario: Navigate to Describe page [20:08:07] cucumber features/upload_wizard.feature:46 # Scenario: Navigate to Use page [20:08:24] these are just errors on commons.wikipedia [20:10:24] (03CR) 10GWicke: [C: 04-1] "I think the checkout of a given patch won't work as-is. See inline comments." (031 comment) [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101912 (owner: 10Subramanya Sastry) [20:11:18] mayankmadan: What specific error messages are you seeing when those individual scenarios fail? [20:12:28] JeffHall, http://etherpad.wikimedia.org/p/mayankmadan [20:13:22] (03CR) 10GWicke: "Looks good to me. We should probably wait with merging this until the tests are ready." [integration/zuul-config] - 10https://gerrit.wikimedia.org/r/101460 (owner: 10Subramanya Sastry) [20:14:17] JeffHall, im gonna go sleep now [20:14:20] so good night [20:15:48] (03CR) 10Subramanya Sastry: Added new jenkins jobs file for /mediawiki/services/parsoid repo (031 comment) [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101912 (owner: 10Subramanya Sastry) [20:24:55] (03CR) 10GWicke: Added new jenkins jobs file for /mediawiki/services/parsoid repo (031 comment) [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101912 (owner: 10Subramanya Sastry) [20:24:58] Project browsertests-test2.wikipedia.org-windows-internet_explorer_6 build #600: FAILURE in 28 min: https://wmf.ci.cloudbees.com/job/browsertests-test2.wikipedia.org-windows-internet_explorer_6/600/ [20:25:51] (03CR) 10GWicke: [C: 031] "Makes sense now. Antoine should have a look at this too." [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101912 (owner: 10Subramanya Sastry) [20:50:50] Yippee, build fixed! [20:50:50] Project browsertests-test2.wikipedia.org-windows-internet_explorer_7 build #604: FIXED in 33 min: https://wmf.ci.cloudbees.com/job/browsertests-test2.wikipedia.org-windows-internet_explorer_7/604/ [20:54:19] (03CR) 10Hashar: "At first glance that seems fine though some part are outdated and need to be replaced by shell scripts." [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101912 (owner: 10Subramanya Sastry) [20:58:55] Yippee, build fixed! [20:58:56] Project browsertests-test2.wikipedia.org-windows-internet_explorer_8 build #624: FIXED in 33 min: https://wmf.ci.cloudbees.com/job/browsertests-test2.wikipedia.org-windows-internet_explorer_8/624/ [20:59:52] better all the time. the other IE failures today were just glitches, not errors in the tests or bugs. [21:08:30] (03PS1) 10Hashar: update parsoid tests to use mw-install-sqlite macro [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101999 [21:10:47] Project Flow-test2-wikipedia-org-linux-chrome build #14: FAILURE in 5 min 34 sec: https://wmf.ci.cloudbees.com/job/Flow-test2-wikipedia-org-linux-chrome/14/ [21:15:08] (03PS2) 10Krinkle: Add 'npm' macro and jobs for oojs [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101461 [21:18:56] (03PS1) 10Hashar: parsoid-php-parsertests fetch mw/core faster [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/102001 [21:20:51] (03PS1) 10Hashar: parsoid-php-parsertests wrappers (times + colors) [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/102002 [21:22:29] (03CR) 10Hashar: [C: 032] update parsoid tests to use mw-install-sqlite macro [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101999 (owner: 10Hashar) [21:22:31] (03CR) 10Hashar: [C: 032] parsoid-php-parsertests wrappers (times + colors) [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/102002 (owner: 10Hashar) [21:22:32] (03CR) 10Hashar: [C: 032] parsoid-php-parsertests fetch mw/core faster [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/102001 (owner: 10Hashar) [21:23:04] (03Merged) 10jenkins-bot: update parsoid tests to use mw-install-sqlite macro [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101999 (owner: 10Hashar) [21:23:34] (03Merged) 10jenkins-bot: parsoid-php-parsertests fetch mw/core faster [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/102001 (owner: 10Hashar) [21:23:35] (03Merged) 10jenkins-bot: parsoid-php-parsertests wrappers (times + colors) [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/102002 (owner: 10Hashar) [21:30:14] Yippee, build fixed! [21:30:14] Project browsertests-test2.wikipedia.org-windows-internet_explorer_9 build #657: FIXED in 39 min: https://wmf.ci.cloudbees.com/job/browsertests-test2.wikipedia.org-windows-internet_explorer_9/657/ [21:31:37] Project Flow-test2.wikipedia.org-linux-firefox build #12: FAILURE in 6 min 30 sec: https://wmf.ci.cloudbees.com/job/Flow-test2.wikipedia.org-linux-firefox/12/ [21:33:49] (03CR) 10Hashar: "Sounds almost good. I will definitely deploy those jobs tomorrow and play tests them, should not cause any troubles." (037 comments) [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101912 (owner: 10Subramanya Sastry) [21:37:11] (03PS3) 10Krinkle: Add 'npm' macro and jobs for oojs [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101461 [21:40:34] (03CR) 10Hashar: "a bit confused with the npm / npm install semantic. Is npm install running tests as well ?" (032 comments) [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101461 (owner: 10Krinkle) [21:41:50] (03PS2) 10Krinkle: Clean up: Remove a few redundant 'scm' properties [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101462 [21:48:47] Krinkle: so yeah most spam is directed here [22:01:35] Just learned from the SauceLabs heldesk that they're currently using a very old version of ChromeDriver for most of the Chrome tests we're running via their platform [22:02:17] That might explain some of the continuing issues we seem to have with jobs that run via CloudBees and use Chrome [22:08:44] JeffHall: that's not very nice. did they say anything about bringing it more up to date? [22:09:29] Well, it's sort of interesting - the exact response was: [22:09:31] The "26.0.1383.0" refers to an older version of chromedriver that we use for our Chrome versions 28 and under. [22:09:31] We use a newer version of chromedriver2 (2.4) for Chrome versions 29 and greater. [22:10:03] huh. that is several versions behind, I think 2.7 is current [22:10:26] And from looking at a recent VE job that ran via SauceLabs, the Chrome browser version was 28, which means they used the pre-2.4 version of ChromeDriver for that test [22:11:54] I think [22:11:54] Željko indicated that we can't control the Chrome browser version used with SauceLabs tests, which means we're running several versions behind there as well [22:13:46] I need to learn more about what we can do to get SauceLabs to at least use a current version of the Chrome browser [22:15:52] JeffHall: yeah, I have not done that research. I have seen discrepancies in Chrome behavior between local and Sauce, but nothing specific to VE [22:17:20] chrismcmahon: Yeah, I don't believe the discrepancies I'm seeing are unique to VE either, but I'm determined to get the VE SauceLabs tests back to green, at least in beta labs! [22:55:49] (03PS1) 10Krinkle: Remove unused 'tools/gruntjs' submodule [integration/jenkins] - 10https://gerrit.wikimedia.org/r/102022 [23:14:41] (03PS4) 10Krinkle: Add 'npm' macro and jobs for oojs [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101461 [23:14:46] (03PS3) 10Krinkle: Clean up: Remove a few redundant 'scm' properties [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101462 [23:15:23] (03PS5) 10Krinkle: Add 'npm' macro and jobs for oojs [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101461 [23:18:02] (03CR) 10Krinkle: "Pushed oojs-core-npm and oojs-ui-npm to Jenkins." [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101461 (owner: 10Krinkle) [23:20:27] (03CR) 10Krinkle: [C: 04-1] "Not yet enabled in Zuul, testing first https://integration.wikimedia.org/ci/job/oojs-core-npm/1/console." [integration/jenkins-job-builder-config] - 10https://gerrit.wikimedia.org/r/101461 (owner: 10Krinkle) [23:25:39] JeffHall: anything you need offhand? I'm about to take off. (I'm looking at potential rental places this week) [23:28:40] chrismcmahon: No, I'm pretty good - just finishing up a new test for switching editing modes in Visual Editor [23:29:56] JeffHall: groovy. see you tomorrow then, I believe I have now gotten all the IE builds back as green as they are going to get, more Flow coverage from me tomorrow, and VE as it develops... [23:30:15] chrismcmahon: Sounds good!