[00:03:44] New patchset: Ori.livneh; "Drop CustomUserSignup (appears to be unused)" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/56553 [00:29:49] Change merged: Reedy; [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/56553 [00:31:22] New patchset: Zfilipin; "Universal Language Selector Accept-Language feature tests now work on Sauce Labs too" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/56557 [00:40:20] New review: Zfilipin; "Jenkins jobs pass: https://wmf.ci.cloudbees.com/job/_debug-browsertests-template/lastCompletedBuild/..." [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/56557 [00:42:00] New review: Zfilipin; "I am not sure why gerrit broke all sauce labs links. Trying again with one link: https://saucelabs.c..." [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/56557 [00:44:31] New review: Zfilipin; "I give up with sauce links for now. I can send them via e-mail if needed." [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/56557 [00:51:41] New review: Zfilipin; "Running all tests did not find any problems: https://wmf.ci.cloudbees.com/job/_debug-browsertests-te..." [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/56557 [01:14:58] New patchset: Mwalker; "CentralNotice deploys off 'wmf_deploy' branch" [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/56560 [01:15:55] Reedy: mind code reviewing the patch above? [01:17:15] Change merged: Reedy; [mediawiki/tools/release] (master) - https://gerrit.wikimedia.org/r/56560 [03:09:22] New patchset: Krinkle; "Add mwext-VisualEditor-doc-test to gate-and-submit." [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/56566 [03:09:35] Change merged: Krinkle; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/56566 [03:42:27] New patchset: Mwjames; "Add Scribunto as SMW testextensions dependency" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/56570 [03:54:16] James_F, you around? [03:55:22] or robla , you around? [03:55:26] qgil: Yes! Just catching up on e-mails. [03:55:47] qgil: robla's account is idle 3 hours now... [03:57:11] qgil: what's up? [03:58:13] robla, nothing, a GSOC submission glitch but james is helping. Thanks! [04:02:09] Change merged: Krinkle; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/56511 [04:03:17] hi hexmode are you following https://bugzilla.wikimedia.org/show_bug.cgi?id=27001#c24 ? [04:03:43] hexmode, it was a nice surprise to find your pull requests at https://github.com/mozilla/mediawiki-bugzilla/pulls [04:03:57] not so nice that they are standing there, though [04:12:31] Ok it's official: we have applied to GSOC: https://www.mediawiki.org/wiki/Summer_of_Code_2013 [04:12:58] thanks James_F for the little help! [04:14:58] Yay. [04:15:15] (I think.) [04:16:17] even without gsoc it has been useful to sort out through http://www.mediawiki.org/wiki/Mentorship_programs/Possible_projects [04:25:57] ah, I love transclusion, especisally when it's late, you are tired and you want to finish a page: [04:25:58] https://www.mediawiki.org/wiki/Outreach_Program_for_Women [09:51:15] New patchset: Pastakhov; "Add jobs for mw/ext/Foxway" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/56587 [10:10:53] hello [10:14:54] hashar: hi [11:04:19] hashar: do you have a minute? [11:04:37] zeljkof: hello :-] [11:04:51] hashar: hi :) [11:05:01] zeljkof: I am looking at the bug you mentioned regarding the mobile interface on beta [11:05:17] great, that is what I wanted to ask, if you saw it [11:05:48] Michelle said it is related to another bug that they recently submitted, I will try to find it [11:06:29] zeljkof: also I think the apaches were stalled with an old copies of some files [11:06:35] zeljkof: you might want to try again [11:06:40] there might be some cache issue too :( [11:07:53] hashar: will try immediately [11:09:08] I did notice yesterday that the bug was not reproducible every time, but 90%+ of the time it was [11:09:28] hashar: looks like it works now [11:09:32] \O/ [11:09:54] cool, fixed :) [11:10:01] I will amend the bug report [11:10:04] https://bugzilla.wikimedia.org/show_bug.cgi?id=46649 [11:10:12] I will run the tests now and let you know something else breaks [11:11:31] I think jenkins jobs are set up to report failures here, so you will see wmf-selenium-bot complaining in 5-10 minutes if something is broken :) [11:11:59] ah reminds me I need to setup IRC notification for the Jenkins beta jobs [11:23:36] New patchset: Zfilipin; "Gems needed for tests to run on Windows" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/56590 [11:34:56] Project _debug-MobileFrontend-template build #3: NOW UNSTABLE in 3 min 31 sec: https://wmf.ci.cloudbees.com/job/_debug-MobileFrontend-template/3/ [11:36:14] hashar: the tests still fail [11:36:16] https://wmf.ci.cloudbees.com/job/_debug-MobileFrontend-template/lastCompletedBuild/testReport/(root)/Manage%20Watchlist/Login_link_leads_to_login_page/ [11:36:23] :(( [11:36:32] https://saucelabs.com/jobs/432ea7d5e5fc49ccb9d2ec0cc55d5ee0 [11:36:52] still redirects (sometimes) to meta [11:38:02] i luuuve sauce labs [11:38:06] * hashar clicks |> play [11:38:20] hashar: sauce is the best thing out there, after beer :) [11:39:05] yeaahh redirect [11:39:22] you know what is the problem? [11:39:33] if I knew about it, it would already be solved -:] [11:39:50] I thought the redirect was new information :) [11:40:34] ok, back to fun with ie8 until you fix this [11:41:31] soo [11:41:36] at some point we hit http://commons.wikimedia.beta.wmflabs.org/wiki/Special:LoginHandshake?useformat=mobile [11:41:47] which redirects to the missing wiki page [11:41:48] I saw that [11:41:59] not sure why that happens [11:42:10] the problem happens when login link is clicked [11:43:39] this link [11:43:41] Login [11:44:34] yeah [11:45:19] which redirects to http://commons.wikimedia.beta.wmflabs.org/wiki/Special:LoginHandshake?useformat=mobile [11:46:06] which in turn does another 302 : Location: http://meta.wikimedia.org/wiki/Missing_wiki [11:46:08] I have noticed the redirect to commons even on production (sometimes) [11:46:23] so the problem is in mobile fronted extension? [11:46:30] the link is broken? [11:47:37] ahhhh [11:47:38] got it [11:47:50] http://commons.wikimedia.beta.wmflabs.org/ <-- missing :-] [11:48:56] great [11:48:59] :) [11:49:06] let me know when I can run the tests again [11:49:08] I must have removed it sorry [11:50:18] !g I1e785f869f6eede2293833b2ed1f595bfbf5cdf5 [11:50:18] https://gerrit.wikimedia.org/r/#q,I1e785f869f6eede2293833b2ed1f595bfbf5cdf5,n,z [11:51:13] great, if that is all [11:51:24] I mean, if that is the problem [11:55:54] ahh jenkins does the deploy automatically [11:55:57] I luve it [11:56:02] https://integration.wikimedia.org/ci/view/Beta/job/beta-mediawiki-config-update/ :-D [11:56:10] there is a jenkins slave on the beta cluster [11:57:04] is beta cluster updated now? can I run the tests? [11:57:30] na bugged [11:57:37] did a typo https://gerrit.wikimedia.org/r/56594 [11:57:44] our tests sucks [11:59:26] http://commons.wikimedia.beta.wmflabs.org/ [11:59:27] \O/ [11:59:39] zeljkof: can you rerun that specific job that was failing? [11:59:45] will do [12:02:51] and the bug is https://bugzilla.wikimedia.org/show_bug.cgi?id=46649 :-] [12:09:43] hashar: thanks :) works fine now [12:10:27] \O/ [12:11:26] so happy [12:11:35] I have to debug some ie8 problems now, but I will go back to testing on beta cluster soon [12:12:08] we should drop IE8 support :D [12:23:20] New patchset: Hashar; "(bug 45084) beta: matrix job to update all databases" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/55949 [12:23:50] New review: Hashar; "added commonswiki database. Job updated manually." [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/55949 [12:29:16] hashar: I am all for dropping IE support :) [13:09:22] New review: Hashar; "deployed" [integration/jenkins-job-builder-config] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/56570 [13:09:25] New patchset: Hashar; "Add Scribunto as SMW testextensions dependency" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/56570 [13:09:33] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/56570 [13:16:23] New patchset: Hashar; "operations-debs-python-voluptuous-debbuild non voting" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/56597 [13:16:48] New patchset: Hashar; "operations-debs-python-voluptuous-debbuild non voting" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/56597 [13:17:07] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/56597 [14:10:14] New patchset: Zfilipin; "Pairing with Runa, work in progress" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/55266 [14:28:09] New patchset: Zfilipin; "Pairing with Runa, work in progress" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/55266 [14:52:01] * jeremyb_ tries again [14:52:05] hi eyoung! [14:52:19] eyoung: /j #wikimania,#wikimania-2013 [14:52:21] :-) [16:42:02] New patchset: Hashar; "operations-puppet-doc for docs.wikimedia.org/puppet" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53753 [16:42:19] New review: Hashar; "rebased" [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53753 [16:46:14] New patchset: Hashar; "generate puppet doc using Jenkins" [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/56608 [16:47:21] Change merged: Hashar; [integration/zuul-config] (master) - https://gerrit.wikimedia.org/r/56608 [16:47:38] New review: Hashar; "Zuul triggers : https://gerrit.wikimedia.org/r/#/c/56608/" [integration/jenkins-job-builder-config] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/53753 [16:47:39] Change merged: Hashar; [integration/jenkins-job-builder-config] (master) - https://gerrit.wikimedia.org/r/53753 [16:55:37] New patchset: Zfilipin; "Enabling tests that were disabled because of a know MediaWiki bugs" [qa/browsertests] (master) - https://gerrit.wikimedia.org/r/56609 [17:20:33] <^demon> Bringing down gerrit for an upgrade for a bit. Should take less than 30 minutes. Please don't panic. [17:23:14] You brokeded all my repo updating :( [17:28:18] I lost my gerrit session [17:28:18] :( [17:28:33] https://gerrit.wikimedia.org/r/login/q/status:open,n,z is missing a style :-] [17:30:40] <^demon> Seriously. [17:30:46] <^demon> *facepalm* [17:31:02] that is not that much of an issue though [17:34:02] I am out for the weekend! *waves* [17:34:22] bye! [17:50:20] Ten minutes till LevelUp office hour [18:00:28] I'm here to talk about LevelUp https://www.mediawiki.org/wiki/Mentorship_programs/LevelUp a little bit and chat with you about how the last quarter went & how the next one might go I'm Sumana Harihareswara and https://www.mediawiki.org/wiki/User:Sharihareswara_%28WMF%29 is more about me [18:00:38] Hmm, that came out a bit disjointed [18:01:10] anyone here for the office hour? [18:01:10] * YuviPanda adds a few JOINs [18:01:14] hello sumanah [18:01:20] hi sumanah [18:01:27] hi anomie & YuviPanda [18:01:37] so, first off, let's talk about how the first quarter went [18:01:48] anomie: you were teaching lwelling about MediaWiki API stuff, right? [18:01:57] * brion is here \o/ [18:02:02] according to https://www.mediawiki.org/wiki/Mentorship_programs/LevelUp/Q1_2013 [18:02:09] sumanah- I was, but somehow we didn't find time to actually do much. [18:02:15] sucheta: awjr - hi, here for the office hour? [18:02:19] anomie: ok. Why is that? [18:02:32] * marktraceur waves [18:02:43] hi sumanah, just lurking - up to my eyeballs with other stuff. if you need anything just ping :( [18:02:44] er [18:02:45] :) [18:02:48] there we go [18:02:53] sumanah- I guess because I didn't bug him about it, and he didn't bug me about it [18:03:13] I'm here! :) [18:03:22] awjr: ah, got it. My main question - how did your LevelUp (re C, with MaxSem) go? [18:03:43] sumanah: slowly; i have found very little time for levelup over the last few months [18:03:59] lwelling: anomie: ok. So theoretically, LevelUp was supposed to be something that your manager managed -- it does not scale to have me nag 70 people .... did either of your managers actually check in on it? [18:04:14] * bawolff is also kind of curious how succesful level up has been in terms of getting contributors more engaged [18:04:29] awjr: let me guess -- Tomasz didn't exactly check in on it, it wasn't an item in the sprint planning, etc. [18:04:38] sumanah- Not really. There was vague mention of level up at some point or other, but no real check-in. [18:04:45] * MatmaRex lurks [18:05:12] bawolff: What I'm hearing here so far is that the first quarter of LevelUp has not been very successful except for OPW people, because with OPW people had more of a stake and a process with lots of check-ins [18:05:14] * waldir lurks [18:06:03] sumanah: is LevelUp mainly concentrated on 1-on-1 mentorship, or can general 'teach cool stuff to interested parties' fall under the umbrella? [18:06:25] brion: I am interested in the latter; taking time to increase our capacity is a good thing [18:06:27] Sumanah: in other words the mentees weren't naggy enough? [18:06:29] sumanah: partly correct. we never discuss it in sprint planning, but it's an expectation that it's happening (it is implicitly folded into our velocity/capacity planning for work to be done) im not sure how others on the mobile team have experienced it, but for me my scrummastering has left me with very little time for development (which is one of my responsibilities), so when i've had an afternoon free for hacking, i've been using it to actually ha [18:06:46] awjr: you got cut off at "i've been using it to actually ha" [18:06:56] doh [18:07:02] so when i've had an afternoon free for hacking, i've been using it to actually hack on mobile specific stuff rather than using the time for level up [18:07:02] bawolff: neither the mentees *nor the managers* were naggy enough [18:07:46] sumanah, from the point of view of a pontential mentee, being naggy might not be something a relative newcomer is comfortable with [18:08:10] waldir: Yes; I know we've had past mentees deal with that in OPW & GSoC [18:08:21] the answer in OPW & GSoC is that the org admin (me) can in fact advocate for those mentees [18:08:26] because there are only 9 of them at most [18:08:30] sumanah- I think next time I'm going to schedule a weekly 15 minutes or so, so it's on my calendar to do it. [18:08:41] sumanah: I'm super-+2 on RoanKattouw_away and TrevorParscal's LevelUp business. [18:08:45] anomie: that's a good idea. And in fact 2 or 3 times a week is better [18:08:49] back in my day, you had to be pretty self-motivating to be a mentee as we had no formal mentoring. you just have to bug people in irc until they helped you [18:08:59] marktraceur: what does that mean? sorry, could you elaborate? [18:09:10] Yeah nobody wants to be annoying but new people especially don't want to be annoying when they are new [18:09:18] sumanah: I'm just saying it was a non-OPW LevelUp action that IMHO was very successful. [18:09:23] sumanah: The JavaScript bootcamp [18:09:26] bawolff: that's where a little organization can help yes :D [18:09:31] marktraceur: ok. what specifically did you do there? [18:09:44] oh "I'm super-+2" means not that you merged it, but that you liked it a lot. [18:09:59] brion, I remember the first time I talked to you on IRC some years ago, I was so scared :P [18:10:10] sumanah: We went through basics and then descended into more specific things, including writing jQuery plugins, documentation, and tests. [18:10:51] Given that a large proportion of mentors and mentees are WMF people, it seems to me like I might need to ask the managers for more buy-in [18:10:57] waldir: I still am [18:11:21] bawolff, your recent email to wikitech replying to te gsoc guy was a good example of how to encourage newbies to not be afraid to not understand stuff when trying to read the code and/or docs [18:11:32] *the [18:11:34] I loved that email bawolff [18:11:46] Waldir: thanks :) [18:12:07] valeriej: how was your OPW experience? what would you improve or change? [18:12:12] * sumanah looks at https://www.mediawiki.org/wiki/Mentorship_programs/LevelUp/Q1_2013 [18:13:16] andrewbogott: csteipp: spagewmf mlitn sucheta superm401 ^demon AaronSchulz - how was LevelUp for you in the last quarter? did you achieve your goal? [18:14:00] brion: so, you saw that csteipp did that security-related office hour this month, which was helpful to the people who showed up and was a way to leverage our existing videos & documentation [18:14:02] sumanah: It was great. qgil, andre__, and I had a wrap-up meeting recently, and I think I should have had a more focused project. I'm glad I was able to participate in things like mobile testing and browser automation testing events. [18:14:25] more of that -- online teaching events where people actually show up and learn something -- would be good [18:14:25] sumanah: i missed it but it sounded good from what i heard [18:14:28] <^demon> Not really. I worked with Andrew Otto on advanced git techniques, but our schedules didn't really line up for us to revisit it. [18:14:29] would love to see more of those [18:14:43] <^demon> The one session we had was good, but it just didn't work out to do more. [18:14:49] ^demon: what would have helped you get your schedules in sync? [18:14:58] <^demon> Being less busy :) [18:15:33] On the subject of opw - is there going to be some sort of official write up on how succesful each one was, lessons learned, etc [18:15:35] ^demon: well, we will always have a lot of things other people, and we, want us to do :-) [18:15:38] ^demon: just fix all the bugs already [18:15:53] <^demon> Also, something occurred to me that I'm curious if others have experienced. Not all people make good teachers. [18:16:10] bawolff: qgil is still doing his final interviews with the participants, and I think most of the participants are writing blog posts -- some of them are on Planet Wikimedia (English) and most are also on http://planeteria.org/wfs [18:16:29] <^demon> And I'm curious if we could find non-mentorship things that people could still do that falls under community-fostering. [18:16:30] ^demon: for sure. So for those people there is https://www.mediawiki.org/wiki/Mentorship_programs/LevelUp#Opting_out [18:16:43] bawolff, sumanah I plan to write a blog post summarizing OPW, yes [18:16:55] * bawolff has read the ones on planet [18:16:57] spoeaking of which, have you seen https://www.mediawiki.org/wiki/Outreach_Program_for_Women sumanah ? [18:17:05] <^demon> sumanah: Ah, I guess that wasn't entirely clear to me in the beginning. [18:17:08] sumanah: mlitn and I talked a little bit, and reviewed a few patches. I'll let mlitn say if he feels more comfortable with AF now than before. I thought it was ok. [18:17:18] brion: so, what are the specific topics where you think there's a lack of knowledge that's blocking people from contributing higher-quality work, and where you think there are docs/videos that people haven't watched but should? [18:18:08] i'd love to do or support a talk on how to make extensions MobileFrontend-friendly at some point [18:18:21] "mobile-first design"? :) [18:18:21] or how to properly support high-resolution screens [18:18:28] at least mobile-able design :D [18:18:34] brion: do we already have any existing docs or videos we can point people to, on either of those topics? [18:18:42] csteipp: what would have made the experience better for you? [18:18:54] not really. i have a presentation on retina displays partially put together, i'll have to finish it up sometime and update it [18:19:10] no good docs on mobile integration yet, that's still new [18:19:37] ok. Would having a deadline (as in, "we're doing a presentation on $date so the docs have to be ready by then") help you, brion? [18:20:07] (i was going to do a retina display presentation for mobile, but it applies to some desktops these days too). giving me a schedule would be awesome incentive :) [18:20:46] brion: ok. What's a reasonable date? [18:20:58] mlitn: ping, would love to know your end of the experience [18:21:04] sumanah: probably a little more bugging by management would have helped. I just didn't think about it most weeks. Also, we had a little bit of a slow start I think because I was confused on the matching. [18:21:22] i could do sometime in april or early may, or if that's blocked out i could do after amsterdam [18:21:40] gimme at least a week to finish the press :) [18:21:42] *preso [18:21:43] that bit is my fault (the slow start) -- and ^demon I didn't advertise the opt-out as much as I'm sure some would like. P) [18:21:50] wait, what emoticon is that? disregard the emoticon [18:21:56] (monocle) [18:22:17] brion: I hesitate to assign you a task without getting an ok from Tomasz [18:22:53] sumanah: yeah check in with him when he gets back. beginning of april i'm mentoring a new guy, so later in the month might be better [18:23:06] okay [18:23:51] OPW interns who blogged about their work include http://mikipedian.blogspot.com/ http://mitevam.tumblr.com/ https://terrrydactyl.wordpress.com/ and others on http://planeteria.org/wfs/ [18:23:54] qgil: looking now [18:24:46] I loved mitevam's blog posts. not sure I've looked at the others [18:25:34] I should check on whether her childbirth went okay [18:25:36] * YuviPanda loved valeriej's 'first bug' blog post [18:25:53] YuviPanda: :D [18:26:23] bawolff: you run some metrics stuff -- how is the MediaWiki (core + all extensions, or core + WMF-deployed extensions) review backlog right now? and how was it in early January? [18:27:04] Umm my metrics are rather crappy and don't really record that [18:27:28] ah okay [18:28:02] Whomever runs that bot (susan? ) might know [18:28:56] lemme look at https://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/2013-03-11/Technology_report [18:29:00] The metric thing I did was a bit of a one off thing I did that I never really got back to [18:29:21] sumanah: might be Jarry1250 [18:29:40] YuviPanda: bawolff also did https://toolserver.org/~bawolff/gerrit-stats.htm [18:31:49] sumanah: spagewmf's project got off to a good start but then petered out. Either because he was way too busy or because the project I suggested didn't really fit with what he was aiming for. [18:32:10] Did lots of people have trouble with levelup competing for time with 'normal' work? And how did folks address that? [18:33:55] andrewbogott: It does look as though it's been difficult for people to properly integrate "time for this task" in with other pre-existing work, or even deprioritize other work [18:34:26] i think it's easier to prep a presentation for a given time than it is to commit to long-term mentoring, at least for some folks [18:35:02] I thought that 1-to-1 mentorships would help with this because if you have a specific person in mind, whose capabilities you can specifically and substantially help, then you made a big difference and one that is a bigger motivator [18:35:14] brion: I think my theory must have been wrong [18:35:32] and you are right [18:35:35] for some folks, as you say [18:35:42] Although I was never subject to 20% (on account of being ops) it struck me as nice that it had a specific hourly budget… seemed like you'd have license to say "can't do that, gotta put in my 20%" I don't know that it worked that way though [18:36:19] I certainly found the mentoring bits pleasant and interesting and would be happy to do more of that. [18:36:43] so maybe for the next quarter I need to structure it as: you have 3 choices. [18:36:44] mentorship [18:36:59] presentations & similar chunks of documentation [18:37:26] sumanah, I personally got a lot of value from 1-on-1 interaction in the two hackathons I participated in, even though they've been from different people in different occasions [18:37:27] and code review for stuff people *outside your team/immediate project* [18:37:45] in fact, it blew my progress in non-mentored periods out of the water [18:38:29] it=mentored coding [18:39:21] sumanah, Well, I am happy with the progress. There were times, during this period of three months when I was not, maybe. But of all the things I have achieved - the fact that it made me a part of this community and helped me to get a real-time coding experience is much satisfying. I did learn a lot of things. [18:39:52] waldir: you had a question that people here might be able to answer better than me, about duration and how much time it takes to attain a certain level of skill [18:39:55] why don't you go ahead and ask [18:40:28] sucheta: I'm glad to hear it was so great! in retrospect, what would you change? what would you do differently? [18:40:29] sumanah, And I believe I have been shamelessly annoying at times. But that's OK. It helps. At times. :) [18:41:24] I wasn't sure about commiting myself to LevelUp because te area I was interested in (skins) seemes to involve quite complex code and I wanted to know wether anyone would have an estimate of how much time/effort would need to be put in to get a reasonable shot at +2 [18:41:33] sumanah: i didn't follow all of this but i saw you list some blog URLs above.. can add to planet if they are not [18:41:46] *the / *seemed [18:41:57] *whether [18:42:12] I am hearing the repeated fear of being annoying, and I hope MediaWiki can become a place that calms those fears -- that helps people learn our norms (assume good faith, leave things better than you found them, be willing to learn and negotiate) and know that as long as they follow those they are not annoying [18:42:33] skin stuff is somewhat complicated, yeah [18:43:13] csteipp: https://gerrit.wikimedia.org/r/#/c/56526/ microcommit [18:43:23] sumanah, And I have shifted to a different time-zone. Not sure which one I exactly belong to. But yes, letting the Earth spin once before becoming impatient - works. And you do get plenty in return! [18:43:25] AaronSchulz: how was your experience mentoring thaiphan? [18:43:33] ^demon: and https://gerrit.wikimedia.org/r/#/c/56522/ [18:43:37] sucheta: that's qgil's suggestion which I endorse :) [18:43:49] ^demon: really want that in wmf13 [18:44:12] binasher: actually should I backport that too? [18:44:18] * AaronSchulz is impatient [18:44:33] <^demon> I was looking at that earlier. [18:44:38] ok, we're getting to the last 15 min of the office hour. sucheta brion ^demon anomie lwelling andrewbogott spagewmf mlitn awjr MaxSem valeriej and others here -- what would you like to learn, or teach, or make, in the next few months, that would improve our community's capacity? [18:45:01] AaronSchulz: i would love to get that backported [18:45:05] AaronSchulz: and the redisjobqueue :) [18:45:07] bbiab [18:45:13] <^demon> I think finding a way to pair unreviewed code with people to review it would be a good thing to investigate. [18:45:16] https://www.mediawiki.org/wiki/Mentorship_programs/LevelUp/Q1_2013 & https://www.mediawiki.org/wiki/Mentorship_programs/LevelUp/Q2_2013 [18:45:19] <^demon> I'm open to any ideas there. [18:45:19] AaronSchulz: even though includes/job/JobQueue.php still has the old name? [18:45:31] csteipp: it has both :) [18:45:32] ^demon: you remember Merlijn's tool there? [18:45:43] <^demon> Yes, but I don't think it's sufficient. [18:46:06] <^demon> It operates on a purely technical level...match this file, repo, etc. [18:46:14] ^demon: http://wiki.growstuff.org/index.php/Pear "So you want to code on Growstuff? We always pair program so in order to start work, you're going to have to pair up with someone. But how do you find someone who's available to pair with you? And what if they're in a different timezone? How do you tell who's available and when?" [18:46:19] <^demon> I'm thinking of ways to find things that people might be "interested" in. [18:46:42] ^demon, the Maintainers page has some info on that, I believe [18:47:05] ^demon: "You can now use the list of available coders to find someone near you who's interested in coding. Or work with our coaches who can help you contact people and make connections, and get you off to a good start." maybe a Pear installation + the Maintainers list? [18:47:13] <^demon> Yes. But if I'm $userFoo, and I commit something, I don't necessarily know to look there. [18:47:25] <^demon> And if it's $someNewFeature, I might not see anyone relevant either. [18:47:29] Sumanah: some stats - feb 22 there were 286 not yet merged patches in mediawiki/core . Now there are 313 [18:47:32] <^demon> Anyway, just cookie licking. [18:47:35] sumanah: his S3 thing was merged into Tyler's patch [18:47:45] According to Susan's tool [18:47:50] ^demon: I don't think you're cookie-licking -- you're not saying "I'm gonna do this thing" and then never doing it [18:48:21] mutante: I figure that I've already suggested to those bloggers that they ask for addition to the planet - if they don't want to be on it, that is their choice [18:48:27] but thank you for the offer [18:48:46] AaronSchulz: ok, and what would you be interested in reviewing, mentoring, or presenting/writing about next? [18:48:52] How do you get added to planet? Aside from submitting a gerrit patch yourself, that is. [18:48:58] Also hi and stuff. [18:49:00] <^demon> AaronSchulz: I think it's all fine. The comments are all nitpicked. [18:49:05] Hi Isarra! [18:49:05] <^demon> You basically just copy+pasted it all. [18:49:22] Isarra: http://meta.wikimedia.org/wiki/Planet_Wikimedia#How_do_I_get_in.3F [18:49:28] * sumanah wishes she could send https://www.mediawiki.org/wiki/Git/Code_review/Getting_reviews to more people as soon as they get developer access or submit their first changeset. ^demon is there a way to do that? [18:49:31] sumanah, I think I am going to concentrate on EtherEditor for some time now. Also, I had got a chance to work with Language Engineering for a little bit of tie in February. I would like to work more on jQuery.ime and more, maybe. It is pretty cool. [18:49:50] oh that's cool! [18:49:52] sumanah: well it's still not merged into master, and the Azure thing is still in limbo (needs to be moved to an extension) [18:50:00] Isarra: i think there's a page on meta you can submit a request in [18:50:05] valeriej: So... make a patch? >.> [18:50:07] I'd like to be done with the s3 thing before looking at something else [18:50:09] Isarra: alternatively you could tell me and I can submit it for you [18:50:10] <^demon> Well, getting access would imply adding something to wikitech, which might be possible. [18:50:13] Yeah, that page looks bloody confusing. [18:50:30] <^demon> Doing something on first commit's probably harder. [18:50:49] Isarra: Hah, you can add your blog feed to Requests for Inclusion, and it will get added... eventually. [18:50:55] AaronSchulz: ok, is that what you'd like to work on in Q2? you can add that in https://www.mediawiki.org/wiki/Mentorship_programs/LevelUp/Q2_2013 [18:50:57] that means getting thai/tyler to fix any remaining issues and finish CR [18:51:06] Isarra: But yes, it's not obvious to do that. [18:51:40] <^demon> AaronSchulz: So, wfIncrStats() is basically a no-op except on cli now? [18:51:54] valeriej: I just added that page to my watchlist :) I'll help triage / add requests from now :) [18:51:59] in terms of i/o, yes [18:52:00] <^demon> Oh, delayed. [18:52:02] <^demon> I see it now. [18:52:06] until flush [18:52:20] YuviPanda: Awesome! [18:52:36] <^demon> Should probably handle __destruct() and have it call flush(), just in case? [18:53:01] oh how nice, marktraceur and I got thanked in http://anteaya.info/blog/2013/03/29/suggestions-for-gnome-opw-applicants/ [18:54:14] ^demon: I had that before and removed it, since the profiler doesn't have it either [18:54:24] probably wouldn't hurt...probably [18:54:40] sucheta: what do you think you'd like your goal for https://www.mediawiki.org/wiki/Mentorship_programs/LevelUp/Q2_2013 to be? [18:55:04] ^demon: meh, I can add it back [18:55:10] Sumanah, the main gaps I have are: [18:55:10] 1. infrastructure, what the production cluster looks like, does and how it is managed. [18:55:17] sumanah: "samanah" :) [18:55:23] yeah, that's how it goes [18:55:24] <^demon> I'm just thinking if something called incr() after wfLogProfilingData() had been called. [18:55:32] <^demon> In practice, it's probably not necessary. [18:55:33] * sumanah listens to lwelling  [18:56:09] ^demon: I was worried if other components needed by flush might be closed when __destruct() calls [18:56:10] AaronSchulz: for Q2, what is the goal? merge into master? [18:56:19] though that's probably not likely either...at least not atm [18:56:31] 2. Performance testing. There have been a few threads recently about profiling and testing on production data. At least the prod data one implied there was an established way to do this on prod slaves. [18:56:47] <^demon> Heh, this is also true. [18:57:15] <^demon> That's actually probably more likely than my paranoia. [18:57:33] lwelling: both of those are generally underknown areas within Wikimedia's technical community, for sure [18:57:48] in my opinion it would be great to have someone like you leading the way as an investigator and educator [18:58:05] ^demon: though it could always handle errors if they happen [18:58:20] 3. Community 101. How RFCs, beaurocrats, village pumps, and other more indepth parts of wiki culture work. There are doubtless resources on it, but it's not easy to quickly absorb [18:58:30] that's my list of known unknowns :) [18:59:03] Lwelling: number three is an interesting one [19:00:18] Wikipedia culture is a society in itself and a rather complex one [19:01:00] bawolff I don't know what percentage of staff were a cultural insider before starting, I presume more than 50%, but it's tribal knowledge and history that takes a long time to really absorb [19:01:00] The level of understanding an anthropologist needs is totally a multiyear project, but the level of understanding a technical contributor needs is probably a lot easier [19:01:02] sumanah, I think I should check it with marktraceur, if he would like to add my name as his mentee for the next quarter. And would set a goal together, for EtherEditor. [19:01:14] some guided tours would be feasible [19:01:45] some of us are old enough that we have an outdated idea of how the community works. i could use a refresher ;) [19:01:47] Lwelling: I could certainly imagine [19:02:10] sucheta: I suspect I'll be involved with GSoC / OPW again, so it might take priority. But sumanah doesn't seem to have qualms about pushing me towards multiple projects. [19:02:12] ^demon: I amended it [19:02:22] <^demon> Looking. [19:02:24] I do have qualms marktraceur [19:02:31] yes, I think a tech contributor can get by without much knowledge depending on how much they want to defer design decisions to cultural interpreters, but the more involvement you want the more you need to know your users mindset [19:02:38] sumanah: I mean, you don't need to :) [19:02:40] absolutely, you're very right [19:02:58] And how the community works is probably important to prevent misunderstandings between devs and wikipedia [19:03:11] sucheta: marktraceur - it is also okay for marktraceur to say "I would prefer to continue mentoring Sucheta and NOT mentor a new OPW/GSoC person" [19:03:32] sumanah: I'm happy to be an asynchronous mentor to multiple people, and now I think I've learned that I need to more actively block out time to be a mentor. [19:04:26] Well, "asynchronous" in that I'll be around responding to requests when necessary but maybe not pinging constantly. Maybe mentorship programs should be more focused, and I wouldn't be able to split that focus at all. [19:06:40] marktraceur, from personal experience, short bursts of *synchronous* mentoring (in-person, IRC), interleaved with perods of self-learning, experimentation and RTFM, have worked best for me, but I suppose YMMV :) [19:07:12] (speaking as an occasional mentee) [19:07:18] Different people definitely have different needs [19:07:22] yep [19:07:25] this is why I started the last round with "how do you learn best?" [19:07:41] sumanah, are there logs for that? I missed it :( [19:07:58] When I asked how you wanted to learn, all of you said you wanted real-life, hands-on work with mentors who could answer your questions. [19:08:05] (by last round did you mean the previous office hours? I did :P) [19:08:16] I mean the last LevelUp round [19:08:25] I did stuff by email, which just took forever [19:08:38] yeah, I figured. I'll lookup the logs from wednesday [19:08:44] csteipp: the block tests seem to be failing now :/ [19:08:46] https://gerrit.wikimedia.org/r/#/c/56522/ [19:08:47] I was hoping this set of IRC office hours would make things better, but so few people have said what they want to learn or teach that I'm gonna have to email anyway [19:08:48] sumanah: I seem to be the only guy workong on Android stuff around, so if anyone else were interested I would happily mentor/help-along [19:09:03] sumanah- I just can't think of anything :( [19:09:04] i'm unsure where exactly to recruit, however - since it is so 'disconnected' from rest of mw development [19:09:08] waldir: I meant synchronous on a small scale, but async in a larger sense - that I would let the mentats call me, rather than me calling them. [19:09:46] marktraceur, that makes sense and actually sounds pretty similar to what I meant. [19:10:04] YuviPanda: I guess email the mobile list? [19:10:08] Right right. [19:10:21] sumanah: I post announcements there often, it usually is met with deafening silence [19:10:25] anomie: well, you could look at what you wrote me the last time, in Jan [19:10:28] (often -> once every week) [19:10:33] YuviPanda: :( that's really sad [19:10:37] it is. [19:10:40] waldir: It comes from my experience helping people in various IRC channels, I guess. Actually my relationship with terrrydactyl over the past quarter was sort of like that, and it went pretty well. [19:10:44] I may need to do a classified ads thread on wikitech-l [19:10:44] No submodule mapping found in .gitmodules for path 'extensions/E3Experiments' [19:10:46] * AaronSchulz sighs [19:10:49] I also forward those announcements (which include a call for participation) on wikitech-l [19:10:56] which is *also* met with complete silence :) [19:11:05] AaronSchulz: do you know if the db on jenkins is deleted and recreated each time? [19:11:19] I think so...it also uses sqlite on a ramdisk [19:11:21] marktraceur :) [19:11:24] * AaronSchulz does that too :) [19:11:26] sumanah: I'm the *only* person working on this - the commits are all 100% mine so far. *Bad* bus factor [19:11:41] oh, bad bad bad [19:11:43] YuviPanda: I'd like to work on Android. [19:11:53] :D [19:11:59] haha! [19:12:04] YuviPanda: Is "Android stuff" a native app? [19:12:08] csteipp: I'm going to revert that for now since it's making all commits fail [19:12:12] valeriej: your bug reports were super-nice during bug day [19:12:13] marktraceur: yes. [19:12:16] That worked out well [19:12:16] AaronSchulz: no problem [19:12:24] YuviPanda: Is it in F-Droid? :) [19:12:36] marktraceur: wer were using HTML5 based apps before, but have moved off them due to technical issues [19:12:41] *nod* [19:12:41] marktraceur: it can be, if you submit it :D [19:12:43] YuviPanda: :D [19:12:44] or if I submit it [19:12:44] He. [19:12:54] YuviPanda: I might dabble in it then. We'll see. [19:12:55] valeriej: you ever played with native Android apps before? [19:13:39] valeriej: what type of stuff would you like to work on? UI/UX on the phone? Automated testing? Features? Performance work? [19:13:45] sumanah: Not in-depth. I've read some articles/blogs on it. I'd love to learn more. [19:13:49] or just general whatever-comes-my-way? :) [19:14:06] YuviPanda: That one, haha. [19:14:12] :D [19:14:17] YuviPanda: is https://www.mediawiki.org/wiki/Mobile/PhoneGap/Tutorial totally obsolete? [19:14:18] valeriej: turn up on #wikimedia-mobile? [19:14:33] (now that we're doing native apps and not phonegap) [19:14:34] YuviPanda: Am there already! :) [19:14:50] sumanah: depends on how you define 'obsolete'. It is technically accurate, just that we don't want to spend any more time on it. [19:15:06] sumanah: 'obsolete' in the sense 'we do not want new people using it', then definitely yes [19:16:02] YuviPanda: will you go ahead and mark it obsolete in some way? I basically think that whatever I say on that page will be wrong [19:16:11] heh [19:16:13] ok [19:16:15] i'm sure there's some template [19:17:52] waldir: http://lists.wikimedia.org/pipermail/wikitech-l/2012-November/064615.html [19:18:02] ok, we've over time, so I'm gonna call this set of office hours over [19:18:19] YuviPanda & valeriej are tentatively paired for next quarter [19:18:57] I need to talk to the 6 WMF directors about moving to a three-option model where everyone can either mentor, do code review on something that's not their main work project and aim to help volunteers, or do presentations [19:19:14] sumanah, not sure why you sent me that link [19:19:33] sumanah, are there logs for that? I missed it :( [19:19:47] that's as much logging as there was, basically, so I wanted you to see it [19:19:57] did you get any answer to your question about how much time it takes to learn skin-related stuff? [19:20:04] sumanah, sorry I was unclear. I meant the log for the levelup office hours two days ago :) [19:20:06] MatmaRex & DanielFriesen seem like people who might be able to answer that [19:20:25] waldir: ah -- sorry [19:20:53] waldir: http://bots.wmflabs.org/~wm-bot/logs/%23wikimedia-dev/20130328.txt starting at 01:54:27 [19:21:10] hm [19:21:11] QUestion? [19:21:12] ? [19:21:13] sumanah, awesome, thanks! [19:21:58] great wmf12 has a similar problem [19:22:06] MatmaRex, DanielFriesen "I wasn't sure about commiting myself to LevelUp because the area I was interested in (skins) seemed to involve quite complex code and I wanted to know whether anyone would have an estimate of how much time/effort would need to be put in to get a reasonable shot at +2" [19:22:34] ^demon: https://integration.wikimedia.org/ci/job/mediawiki-core-qunit/1323/console [19:22:52] actually I don't even bother about +2, although that would be the ideal so I could help the community back. But I feel a little lost trying to learn about the skinning system on my own [19:22:53] waldir: well, i learned all i know about skins by rewriting one [19:23:01] waldir: I think their answer will also depend on where you start -- what experience do you already have [19:23:08] daniel is more knowledgeable than me [19:23:10] <^demon> AaronSchulz: Nfc. [19:23:43] MatmaRex, I learn a lot through RTFMing so I would love to get some help fleshing out skin-related docs [19:23:47] The skin system is honestly a mess [19:24:06] waldir: RTFC(ode) is a much better solution here ;) [19:24:11] lol [19:24:23] because the docs are lacking? [19:24:45] !b 1 [19:24:45] https://bugzilla.wikimedia.org/1 [19:24:47] DanielFriesen, I've seen you say that a couple times; what hope is there for it? what needs to be done and by whom? [19:24:51] waldir: yeah, a little [19:25:02] No not the docs... it's practically layers of patchwork. [19:25:09] waldir: most functions in Skin and SkinTemplate only have types of parameters and return value describer [19:25:17] well, maybe not most [19:25:19] but a lot of them [19:25:35] [29-Mar-2013 19:25:21] Catchable fatal error: Argument 1 passed to Wikibase\EntityContentFactory::getTitleForId() must be an instance of Wikibase\EntityId, null given, called in /usr/local/apache/common-local/php-1.21wmf12/extensions/Wikibase/repo/includes/api/SetReference.php on line 85 and defined at /usr/local/apache/common-local/php-1.21wmf12/extensions/Wikibase/repo/includes/content/EntityCon [19:25:36] tentFactory.php on line 140 [19:25:42] MatmaRex, I meant mostly the docs at mediawiki.org :) [19:25:45] also, yeah, there's OutputPage, Skin, SkinTemplate, and who knows what else, and it's all intercnnected [19:25:50] What do you even want to do with the skin system? [19:26:15] Improve skins and fix bugs, I imagine! [19:26:16] waldir: oh, don't even mention them :P although someone was working on some sort of a skinning guide lately [19:26:21] probably it was daniel :) [19:26:32] DanielFriesen, frontend dev has always been a personal interest of mine, so I'd like to be able to contribute in that area [19:27:02] but almost all I've done so far has been common.css / common.js hacks, and very minor core patches [19:27:17] waldir: skin bugfixing? here's 50-some bugs for Vector: https://bugzilla.wikimedia.org/show_bug.cgi?id=44881 [19:27:38] vector scares me [19:27:45] Heh [19:27:59] So, I imagine waldir needs to learn Vector Bravery [19:28:02] waldir: oh, the code is painful [19:28:06] (should be a badge one can earn) [19:28:09] lol [19:28:13] Well it does have some messy hacks on top of it's own [19:28:24] waldir: if you rewrite it a little, you'll be my personal hero ;) [19:28:31] waldir: both the PHP and the CSS [19:28:33] I wanted to eliminate some things like the way it handles attributes [19:28:50] I would take the journey if someone would point me in the right directions when I got lost [19:29:01] and I also need some pointing on where to start [19:29:09] waldir: i'm on irc on afternoons and evenings UTC :) [19:29:11] Originally it practically re-implemented the entire system of tabs in a Vector-only way. [19:30:30] waldir: about 'where to start' - well, what exactly is it you want to do? [19:30:39] just learn about the skins, in general? [19:30:43] MatmaRex, I appreciate the offer! What I would like to know is what is the path towards making the skin system saner, and if there's anything I can do about that, where I should start. [19:31:05] waldir: uh, daniel is better qualified to talk about that [19:31:25] waldir: https://gerrit.wikimedia.org/r/25170 is probably a good start [19:31:50] Ah, interesting, wasn't aware of that one [19:31:53] waldir: there's a lot of legacy stuff in there to keep the old skins working, and they *need* either removal (what this patch does) or rewrite (what i did for cologneblue) [19:32:11] i mean, just look at the diffstat, +3, -1747 ;) [19:32:31] :) [19:32:58] whoa, that thread is daunting! I'll take a read later. [19:33:11] waldir: if you want to just get an idea of how the HTML is generated and where the data comes from, i'd suggest reading the code for Monobook first [19:33:44] that thread is a mess, and personally i wouldn't call it worth reading [19:33:49] hmm [19:34:20] some discussions about keeping cologneblue, about moving nostalgia to an extension, some vetoes, and about how to og forward in the WMF side [19:34:36] maybe I'll try reading daniel's skinning tutorial [19:34:39] ;) of course I did most of the work by moving legacy skins to SkinLegacy in the first place [19:34:50] Yeah, that's a fair place to start. [19:35:02] waldir: Also read up on ResourceLoader [19:35:24] Hello sumanah, my name is DJ, and I am interested in applying for OPW. I don't have a technical background, but the "Feature videos" project seems like something I could do. [19:35:36] sumanah: It says you can mentor that project; if you have time, do you have any ideas on how I could get started? [19:35:40] waldir: ;) if you can master the backend of RL there's a skin bug or two dependent on RL changes I'd like to have made. [19:35:57] hi Exceedingly_! welcome to MediaWiki's community [19:35:59] ah, ok, will do then, and contact you :) [19:36:17] sumanah: Thank you! :) [19:36:21] Exceedingly_: so, have you ever made short documentary videos before? [19:37:17] sumanah: I have not actually made documentary videos, but I do definitely have video editing experience. [19:37:36] Exceedingly_: Can you tell me more specifically about why Wikimedia interests you? [19:37:38] DanielFriesen, I'll keep in touch with both you and MatmaRex if I hit any roadblocks along the way. Would you mind if I send private emails to you instead of sending to wikitech-l? I would like to help improve the docs along the way as I learn, since I still have a newbie's perspective [19:37:46] 48 Fatal error: Class 'StatCounter' not found in /usr/local/apache/common-local/php-1.21wmf12/includes/GlobalFunctions.php on line 1180 [19:37:46] 35 Fatal error: Class 'StatCounter' not found in /usr/local/apache/common-local/php-1.21wmf12/includes/GlobalFunctions.php on line 1245 [19:38:12] Reedy: what server? [19:38:18] AaronSchulz: Not got that far [19:38:21] I just looked at the logs a few min ago [19:38:27] I'm just logging a bug for a Wikibase error too [19:38:35] I did sync-dir...maybe it was intermediate [19:38:48] *intermittent [19:38:52] Will have a look in a few mins [19:39:32] waldir: no, i don't mind, but my skin knowledge is mostly self-taught - when i see a function i don't know, i learn its docs (if any), then its code, and then i know it ;) so i'm often not the guy to ask about the big picture [19:40:17] sumanah: valeriej recommended me to the program. She has seemed to greatly enjoy her experience, and when she told me that people with less technical backgrounds could apply, I found myself interested. [19:40:51] Exceedingly_: That's great! I'm in agreement that OPW is good for this. But why Wikimedia specifically? because of valeriej's experience? [19:41:08] * sumanah looks at https://live.gnome.org/OutreachProgramForWomen/2013/JuneSeptember  [19:41:12] MatmaRex, that's cool, since I tend to have a hard time moving on when I find something small I don't completely understand :P [19:41:12] AaronSchulz: numerous [19:41:21] Look to have stopped 17 minutes or so ago [19:41:32] waldir: Sure [19:41:38] So transitional [19:41:55] Exceedingly_: check out https://wiki.openmrs.org/display/RES/Outreach+Program+for+Women+Summer+2013 -- "Creating profiles of OpenMRS implementations" might be interesting to you as well [19:42:33] Oh... ori-l ^_^ your mediawiki-vagrant has been a very helpful reference to creating a vagrant setup for gareth. [19:42:52] MatmaRex, DanielFriesen in any case I'll have to squeeze this among my other projects, so don't worry about getting bugged too often :D [19:43:07] :) [19:43:11] DanielFriesen: oh, cool, that's nice to hear. [19:43:27] thanks both for the pointers, I'll keep you guys posted. [19:43:29] let me know if you run into any issues [19:43:35] sumanah: Just got a Skype message that Exceedingly_'s chat client is giving her problems. [19:43:52] * YuviPanda should wander out of his android silo and try ori-l's vagrant setup at some point [19:44:19] After a little more bugfixing I might even be able to commit it to a new repo soon. [19:44:26] oh, thanks for the heads-up valeriej. what client is it? [19:45:24] ori-l: No issues I haven't been able to solve myself... but I do have a question. Any reason for choosing the Ubuntu cloud box over the Ubuntu boxes Vagrant distributes? [19:45:32] sumanah: She's using Chatzilla on Firefox. The browser froze. [19:45:44] bleh. [19:46:16] DanielFriesen: let me reverse the question :) Any reason for choosing the Ubuntu boxes Vagrant distributes over the Ubuntu cloud box? [19:46:39] an earlier revision actually did use the one distributed by Vagrant, but I switched when Canonical made their own Vagrant boxes available [19:47:48] if you go to https://downloads.vagrantup.com/ (note HTTPs) you get an SSL error because the heroku instance serving the files is misconfigured [19:48:13] sumanah: i've put a notice on top of https://www.mediawiki.org/wiki/Mobile/PhoneGap/Tutorial [19:48:15] ori-l: I guess it's a "These boxes are distributed by the people that build Vagrant. They're experts in how it should be setup for vagrant." vs "These boxes are distributed by the people that build Ubuntu. They're experts on the OS." [19:48:15] vagrant is really useful and neat, but I think it's better to rely on canonical when it comes to binary data [19:48:16] thanks for pointing that out [19:50:11] DanielFriesen: well, vagrant boxes from canonical are pretty new. The duplication of effort doesn't represent any kind of mutual distrust, I don't think -- Canonical is being a good collaborator. I expect that if Canonical continues to maintain the Vagrant boxes, the Vagrant-made ones will be deprecated [19:51:17] It's also a question of which site you trust not to get compromised and serve boxes with rootkits [19:51:47] (answer: the one with a non-broken SSL setup :) [19:52:20] * Reedy generates ori-l a self signed SSL cert [19:52:31] ori-l: ;) says the one with a http:// box url. [19:52:41] ow [19:52:44] that is a good point [19:52:50] * ori-l points at Reedy [19:52:52] it's his fault [19:52:56] ..somehow [19:53:10] lol [19:53:29] it's always Reedy's fault [19:53:51] DanielFriesen: if in the course of experimenting with this stuff you happen to find that the Vagrant-provided boxes are better configured, let's let Canonical know. IIRC they published the boxes with an explicit request for feedback from users. [19:53:58] sumanah: (Apologies for my browser freaking out on me.) valeriej passed me the link you sent, and I looked, but there's not a lot of information there yet about projects that aren't specifically coding. I intended to check this and other project pages later, when they updated more. But the mediawiki page already had some specifics, and the one I mentioned looked like something I could do. [19:54:38] Exceedingly_, yes, we will be adding non-coding projects soon [19:54:40] DanielFriesen: http://blog.utlemming.org/2013/01/vagrant-cloud-images.html [19:54:57] Well I'm rebuilding my vm after changing the box. [19:55:13] Same guest images error... so no worse there. [19:55:37] qgil: That's good to know; I'll definitely be keeping an eye out for their addition. :) [19:55:58] Exceedingly_, have you checked the Raw projects at http://www.mediawiki.org/wiki/Mentorship_programs/Possible_projects ? [19:56:15] Exceedingly_: Hope your browser starts calming down! :) Unfortunately I think I have to take a break from mentorship in the coming term, and so I strongly recommend you look at OpenMRS's suggestion -- I bet they'd be very interested in your ability to do writing & video journalism [19:57:11] If it all works out... I'll return to trying to get rid of the error you get when trying to pip install mysql-python because the distribute version that comes with ubuntu is too old. [19:57:12] Exceedingly_, the first candidate to make it to the OPW page will be http://www.mediawiki.org/wiki/Mentorship_programs/Possible_projects#Browser_Test_Automation [19:58:23] qgil: I hadn't checked the Raw projects, and am looking them over now. Thank you for the links! [19:58:24] DanielFriesen: https://github.com/mitchellh/vagrant/blob/master/lib/vagrant/downloaders/http.rb#L26 :((( [19:58:51] blech [19:59:01] sumanah: Alright, thank you very much for the guidance! I will definitely look into that. [19:59:47] Exceedingly_: We will be adding more stuff to our page, for sure. [19:59:54] But it won't hurt you to talk to OpenMRS. [20:00:14] I have met Michael Downey -- he's an enthusiastic guy. [20:00:39] ori-l: In the process of turning the gareth setup into a vagrant one I discovered something while rethinking how to launch the gunicorn daemon (the init script I used on AWS didn't work this time). [20:00:41] also Exceedingly_ let me spam you with a few things to read: [20:00:49] ori-l: http://supervisord.org/ [20:01:03] Exceedingly_: http://open-advice.org/ -- what people in open source wish they'd known when they started [20:01:09] DanielFriesen: yeah, Ive used supervisor a lot [20:01:22] sumanah: oh, that looks neat [20:01:44] Exceedingly_: http://openhatch.org/missions/ interactive tutorials in some of the key tools we use in open source [20:03:39] also, Exceedingly_, please go ahead and sign up for an account at en.wikipedia.org and then go to https://en.wikipedia.org/wiki/Special:GettingStarted to see some easy ways to get started editing [20:04:55] sumanah: Again, thank you so much! I'll definitely look more into OpenMRS and these links you've sent. I really appreciate it. :) [20:05:19] Glad to help Exceedingly_! and feel free to hang around here or #mediawiki to get a taste of our community [20:06:22] sumanah: Alright! I probably will hang around for a while longer, and I'll almost definitely be back. [20:06:42] cool! [20:14:30] Project browsertests-linux-chrome build #247: UNSTABLE in 13 min: https://wmf.ci.cloudbees.com/job/browsertests-linux-chrome/247/ [20:16:24] ori-l: I guess I'll copy that comment into mine too [20:17:28] *facepalm [20:18:31] ori-l: Vagrant needs better docs. I didn't know that synced_folder had a create: option. I went and wrote a hack at the top of my Vagrantfile to ensure that the directories existed! [20:27:06] DanielFriesen: yep, agreed [20:27:57] ori-l: Any specific reason you defined a private network? Also what was the forwarded port auto_correct option again? [20:31:34] Oh lol... so that's why gareth is managing to work even though I haven't coded the database setup into it. ;) Gareth's main page doesn't need to make a single SQL call to work yet since I haven't implemented the dashboard. [20:33:16] DanielFriesen: a bridged network set-up has deeper security implications, since it makes your vm potentially accessible from the outside world [20:33:38] Oh... so the default is less secure. [20:34:12] I guess that's enough reason to do the same. [20:49:16] ori-l: Any reason not to use `([10] + 3.times.map { rand(256) }).join('.')` to generate the private IP instead of hardcoding it? [20:51:22] DanielFriesen: well, the address should be in the private address space [20:51:33] Yup it is [20:51:43] oh, right, the 10.x [20:52:30] Take the array [10], append an array of three numbers from 0-255, then join by '.'. [20:52:44] ;) aren't ruby one-liners /pretty/ lol [20:53:30] well, a randomly-generated IP is just as likely to create a conflict [20:54:02] (not entirely true, since the distribution is probably skewed toward memorable IPs like 10.11.12.13) [20:54:35] YuviPanda, hello [20:54:42] DanielFriesen: but this way you get something memorable and specific that works in most cases and could be referred to specifically in docs or when troubleshooting, and it's easy enough to change [20:54:44] hey Rahul_21 [20:55:15] ori-l: Well on the positive side... if you have a conflict... you just reload. ;) which you don't get normally. [20:55:20] DanielFriesen: remember too that IP assignment happens every time you boot the machine. if it was not saved somewhere, the value would be recomputed every run, and you'd get conflicts some times but not others [20:55:43] YuviPanda, would you like to review a ptach? [20:55:54] depends on how easy it is :) [20:55:56] link? [20:56:13] I wish vagrant would just allow you to omit the ip and have it pick one that's not in use. [20:56:44] https://gerrit.wikimedia.org/r/#/c/56672/ thats a small change [20:56:47] !g 1 [20:56:48] https://gerrit.wikimedia.org/r/#q,1,n,z [20:56:58] and https://gerrit.wikimedia.org/r/#/c/55517/ thats a lil bigger [20:59:15] Rahul_21: for the first one I do not have enough knowledge of the surrounding code to make that decision [20:59:22] Rahul_21: plus we usually like to keep exposed surface low [20:59:32] Rahul_21: marktraceur might be able to help review https://gerrit.wikimedia.org/r/#/c/56672/ [20:59:40] marktraceur: super small patch that one [20:59:59] YuviPanda, try your hand at the 2nd one :) [21:06:18] Rahul_21: sorry, out of my knowledge-zone too :( [21:07:08] YuviPanda, ok how about this one https://gerrit.wikimedia.org/r/#/c/55518/ ?(last one) every1s given it a +1 [21:09:40] Rahul_21: merged [21:10:50] YuviPanda, hi5!you \m/! [21:11:01] :) [21:11:55] Rahul_21: feel free to close https://bugzilla.wikimedia.org/show_bug.cgi?id=23472 [21:13:53] done [21:14:12] :) [21:15:34] YuviPanda, now its marktraceur 's time [21:15:49] Rahul_21: :) [21:16:14] YuviPanda, you've been affected by "bazooka"? [21:16:39] Rahul_21: ? [21:17:27] YuviPanda, http://gadgets.ndtv.com/internet/news/bazooka-attacks-slowing-down-the-internet-say-experts-347495 [21:17:44] Rahul_21: oh that. it's sortof for the most part a massive over-reaction :) [21:18:38] YuviPanda, since 2 days it took me an hour to get my facebook profile page [21:19:23] :) I haven't been on facebook for a while, so I guess I haven't noticed :P [21:19:34] i'm usually on 3g, so i'm used to high latency internet i guess [21:19:55] YuviPanda, spoiled brat :P [21:20:00] Rahul_21: Why are we making this public? [21:20:32] Rahul_21: stable broadband > 3G [21:20:36] marktraceur: the facebook part? [21:20:43] marktraceur: also i thought you were off on Friday! [21:20:45] awjr: No, the validateName function [21:20:46] I am! [21:20:53] Er, not awjr. [21:20:56] Sorry. [21:20:57] eh? [21:20:58] oh [21:20:58] ah, yes. [21:21:07] everyone mistakes awjr for everyone else ;D [21:21:13] hehehe that's the bagillionth time that's happened today [21:21:22] :D yeah, first me, now marktraceur [21:21:36] * YuviPanda lets marktraceur and Rahul_21 sort it out :) [21:21:52] No, it was a wrong-window thing. [21:22:14] marktraceur, if your free could you review this one line patch https://gerrit.wikimedia.org/r/#/c/56672/? [21:23:07] I'm sitting here looking at it but it doesn't answer my question. [21:23:30] Rahul_21: _Why_ does this need to be public? [21:24:02] marktraceur, the pronoun's causing confusion [21:24:16] Why does this function need to be public? [21:25:13] marktraceur, thats what the bug report says [21:25:41] marktraceur, https://bugzilla.wikimedia.org/show_bug.cgi?id=38221 [21:25:59] Rahul_21: Can you explain some of the use case in the commit message? Also use the Bug: 38221 syntax rather than polluting the first line please. [21:28:16] YuviPanda: I have to go now but if Rahul_21 has any questions please help him. I'll merge the patch if those concerns are met. [21:28:43] marktraceur: will do. Thanks :) [21:29:05] Rahul_21: marktraceur is asking justification of that bug report itself, on *why* he needs those to be public, which extension / code he is using them in, etc [21:29:17] Rahul_21: if you're unaware of those, pinging the original filer might help [21:29:37] Rahul_21: it's generally good practice to make public only as few methods as possible (so that they can be changed in the future easily) [21:30:16] YuviPanda, his name is Christian Neubauer,is he on the irc? [21:30:27] i do not know, Rahul_21 :( [21:30:32] but you can find his email from that bug and mail him [21:30:47] YuviPanda, yes thats the best thing that can be done [21:30:53] yup [21:31:31] and mark mentioned "Bug: 38221 syntax rather than polluting the first line please." ,i did not get that [21:33:33] Rahul_21: see https://www.mediawiki.org/wiki/Gerrit/Commit_message_guidelines [21:34:41] hes a strict disciplinarian it seems :) [21:35:00] don't think of it that way [21:35:20] ok :) [21:35:35] Rahul_21: when you make a method public, it usually means a guarantee that that method is not going to fundamentally change or go away in the next few releases [21:35:52] Rahul_21: so it's hard to fundamentally change it in the future if need arises [21:36:00] so... needs justification before being made public [21:38:22] YuviPanda, When uploading a file using the UploadBase.php class, the getTitle() function [21:38:23] will return the title of the uploaded file or null if the title is invalid for [21:38:23] some reason. However, to figure out why the title is invalid, you have to [21:38:23] access the member variable mTitleError or call the function validateName(). [21:38:23] Both are protected though so they can't be called by external classes. [21:38:23] The easiest fix would be to make validateName() public instead of protected. [21:38:38] you're just copy pasting the bug report :) [21:38:51] what exactly does he mean by 'upload using UploadBase class'? [21:38:58] and where is he going to call validateName from? [21:39:06] which are the 'external classes' he is talking about? [21:39:48] it made sense to me! [21:40:01] it makes sense, I understand it [21:40:03] wait ill check out a few things related to this [21:40:12] i'm simply asking what these 'external classes' are [21:40:13] that is all [21:55:17] YuviPanda, my net is causing troubles! [21:55:35] YuviPanda, have a match tommorow at 6!bye [21:55:38] Rahul_21: did you get my messages? [21:55:41] Rahul_21: ah, have fun :) [21:55:53] YuviPanda, yes i got them [21:56:04] YuviPanda, i sent a mail to the filer! [21:56:07] nice :) [23:08:29] ori-l: Fun... I managed to run into a situation where I need a newer version of puppet than I have [23:09:15] with mediawiki or gareth? [23:09:24] ori-l: Gareth [23:10:34] ori-l: I think I can fix it by using stages. Basically I was adding dependencies so that the various python packages, etc... were installed before any of the Gareth setup was done. [23:10:40] ori-l: And I ran into this puppet bug [23:10:41] http://projects.puppetlabs.com/issues/7422 [23:10:56] Which they fixed in 2.7.20 [23:11:11] But vagrant boxes come with 2.7.11 [23:12:40] Yippie, build fixed! [23:12:40] Project browsertests-linux-chrome build #248: FIXED in 11 min: https://wmf.ci.cloudbees.com/job/browsertests-linux-chrome/248/ [23:14:16] DanielFriesen: fortunately for you, puppet provides twelve different ways to express dependencies [23:15:23] ->, <-, ~>, <~, before:, require:, notify:, subscribe:, and let's not forget run stages [23:33:52] Looks like run stages won't work [23:34:17] Eh... [23:34:19] Ehhh! [23:34:23] nooooo don't use run stages [23:34:29] i was just joking [23:34:40] *facepalm* I can't believe I missed comment 11 [23:35:08] Run stages are conceptually a great idea... but implemented they have too many limits [23:38:07] I can't believe how trivial what I actually had to do was [23:38:57] Modify three require/before entries so that the single item contents are wrapped in an array. [23:41:08] Dummy resources would probably work better than run stages [23:42:31] package { [...]: ..., before => X['package-setup'] } ... exec { 'setup-foo': ..., require => X['package-setup'] } [23:44:12] X (or whatever you call it) does nothing. But you can make all your package setups, etc... you'd conceptually think of putting in a stage run before some X you define. Then everything that sets up application stuff that requires all those system packages be installed could just require it. [23:44:28] No issues with notify like run levels.