[02:55:24] Transitioning today's tech talk to real life... Tell me how to work on https://gerrit.wikimedia.org/r/#/c/30637/ [10:09:29] hashar: are you around? [10:09:40] zeljkof: sure :-] [10:09:47] great :) [10:10:11] I am trying to find your dotfile, the one that adds current git branch to bash [10:10:19] but I can not find it [10:10:49] it is in my Awesome Life Improver under uniX project on github [10:10:54] https://github.com/hashar/alix/blob/master/bashrc [10:10:58] not awesome at all though [10:11:12] awesome life improver :) [10:11:16] thanks [10:11:19] if you have a recent git version you should be provided with a __git_ps1 bash function [10:11:45] I think I provide a very lame fallback in case git version is too old [10:11:52] might want to enable bash_completion too [10:12:18] and double check that your git package does populate the completion script somewhere under /etc/ (I think that is /etc/bash-completion.d/ [10:12:47] I think I have the latest git provided by homebrew: git version 1.7.12 [10:12:57] should got it [10:13:08] (btw, 1.8.0 is available in brew) [10:13:18] I am just updating :) [10:13:58] so the brew formula does provide the bash completion file under `brew --prefix`/etc/bash_completion [10:14:07] which is handled line 12 - 16 of my bashrc script [10:17:36] looks like "brew upgrade git" does the update, and not "brew update git", as I thought [10:21:24] I believe "brew update" simply git pull from the repository [10:21:37] and show you which formulae have been update [10:21:42] like apt-get update [10:23:52] ruby has "gem update package", so I thought brew will have the same [10:25:55] man, irc is so much quieter during CET working hours [10:26:40] ori-l: :) [10:27:06] but quiet as mice you guys work on all the cool projects [10:27:35] i'm on to your secret now [10:28:20] ori-l: and early morning is even quieter [10:28:26] that is usually when I do my hack and coding :-] [10:28:41] start being active around 2pm european time when east coasters wake up [10:29:00] zeljkof: will you be there this afternoon to write a quick report about our sprint ? [10:29:10] zeljkof: I would like to send something back to robla [10:29:34] hashar: I should write the report, or you just need me around if you have any questions? [10:29:47] I though we could write it together using etherpad [10:30:09] hashar: I have to go around 3:30 pm CET [10:30:25] but I will be back around 9 pm CET [10:30:39] so any time except 3:30-9 pm CET is fine with me [10:30:50] mind doing that post lunch? [10:30:55] busy hacking right now [10:30:59] and might not connect tonight :/ [10:31:07] sure, just ping me after lunch [10:31:10] nice [10:31:22] I should be around by 1:30 / 2pm [10:31:25] will ping ya [10:31:48] ok [10:32:04] I think "brew upgrade" was a mistake, now it upgrades everything :) [10:32:23] but I guess I had to do it sooner or later [14:19:36] hashar: I have to go soon, are you here? [14:46:11] zeljkof: seems not :( [14:54:14] zeljkof: doh I forgot :( [14:54:21] will attempt to connect tonight so :-] [14:57:07] hashar: are you still here? [14:57:17] Nikerabbit: yes [14:57:21] I have freshly installed mw master with sqlite, and tests fail: [14:57:27] Configuration read from /resources/nike/melange/mediawiki/tests/phpunit/suite.xml [14:57:31] DB connection error: Cannot return last error, no db connection [14:57:33] make: *** [safe] Error 1 [14:59:57] hashar: otherwise it seems to work well, so I'm puzzled [15:01:36] hmm well [15:01:48] do you have a LocalSettings.php with an installed database ? [15:02:01] oh yeah sqlite [15:02:11] enable debug maybe? [15:02:16] not sure what might be happening [15:02:37] Failed to connect to database 0 at [15:02:53] this is my conf: [15:02:54] $wgDBtype = "sqlite"; [15:02:55] $wgDBserver = ""; [15:02:55] $wgDBname = "my_wiki"; [15:02:55] $wgDBuser = ""; [15:02:57] $wgDBpassword = ""; [15:03:15] $wgSQLiteDataDir ? [15:03:31] $wgSQLiteDataDir = "../db"; [15:03:47] does it need absolute path? [15:03:54] no idea [15:04:01] let my try [15:04:08] I have it set to a full path [15:04:54] yup, that is it [15:05:04] whenever the sqlite file is not readable the sqlite db class should throw an exception [15:05:36] interesting case [15:06:27] ..............................I.............................. 1708 / 5121 ( 33%) [15:06:31] ..................make: *** [safe] Segmentation fault [15:06:39] doh [15:06:40] for fucks sake why is it impossible for me to run tests succesfully?? [15:08:17] hashar: what is difference between make safe and make destructive? both seem to have the same number of tests [15:08:33] look at the Makefile ? [15:09:37] safe exclude tests from Destructive group [15:10:38] interestingly make destructive hasn't exploded yet [15:13:09] Tests: 5022, Assertions: 894291, Failures: 64, Errors: 8, Incomplete: 3, Skipped: 21. [15:13:51] nice [15:15:20] and you might want to exclude the Dump tests which are slow [15:31:19] hashar: I am back, will be here for the next 30 minutes or so [15:32:15] hashar: why do I get: Error: 1 no such table: unittest_user [15:32:23] and after that: Error: 1 table "unittest_melange_searchindex" already exists [15:32:41] zeljkof: a bit too many stuff to do. I guess we can try this evening [15:32:45] zeljkof: else on monday [15:32:56] zeljkof: sorry to have forgotten about that later on :( [15:33:16] Nikerabbit: looks like the table hasn't been deleted in a previous run [15:33:34] hashar: no problem, I will probably be online after 9pm our time, or Monday [15:33:48] hashar: I will be online Saturday morning too [15:33:50] Nikerabbit: maybe we should clean all unittest tables before attempting to clone them [15:34:13] zeljkof: I am doing paperwork / accounting on saturday morning ;-] [15:34:16] hashar: that still doesn't explain why unittest_user was not created :( [15:34:17] so I guess monday [15:34:23] the test has @group Database [15:34:30] zeljkof: will start out something over the week-end so we have a base to improve on monday [15:34:41] hashar: ok, Monday then :) [15:34:55] zeljkof: I think it is better, might not show up this evening :( [16:28:29] New patchset: Hashar; "Add job for testing Validator extension" [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/32835 [16:28:47] New review: Hashar; "rebased, lets deploy it." [integration/jenkins] (master); V: 0 C: 2; - https://gerrit.wikimedia.org/r/32835 [16:28:52] New review: Hashar; "rebased, lets deploy it." [integration/jenkins] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/32835 [16:28:52] Change merged: Hashar; [integration/jenkins] (master) - https://gerrit.wikimedia.org/r/32835 [19:59:43] Does moving an article change its id? [20:00:08] No [20:09:16] <^demon> StevenW, RoanKattouw: Deleting an undeleting does, though. [20:09:30] Gotcha. Thanks! [20:10:52] Yes [20:46:25] Krenair: Is there anything else that needs to be fixed in https://gerrit.wikimedia.org/r/#/c/31329/ ? [21:35:13] Change abandoned: Hashar; "replaced by https://gerrit.wikimedia.org/r/#/c/33580/" [integration/jenkins] (zuul) - https://gerrit.wikimedia.org/r/33539 [21:45:26] ok, I'm going to try going through https://www.mediawiki.org/wiki/Review_queue for a moment [21:46:55] sumanah: Good luck! It's dangerous to go alone, take this. [22:08:13] we need to find someone willing to backport+push https://gerrit.wikimedia.org/r/#/c/33417/ [22:09:16] robla: I got it [22:09:26] RoanKattouw: thanks! [22:19:04] robla: Done [22:19:30] excellent, thanks again! [22:25:50] aww thanks [22:55:35] hi marktraceur - it is The Ten Minute Proposal Discussion [22:55:52] So, I understand that some of your interests are: [22:55:54] Heeeere we go! [22:55:54] mentorship [22:56:02] zealous free software outreach [22:56:07] parsers [22:56:11] OpenHatch [22:56:13] * marktraceur takes issue with "zealous", but OK [22:57:01] Wikimedia [22:57:13] parkour [22:57:17] bridge [22:57:55] I think we're devolving out of what might be useful at a tech conference, but yes, right so far :) [22:58:28] what kinds of talks do you like to hear at a tech conference? [22:59:07] sumanah: I'm usually happier when it's a tutorial. So I've already submitted two coding tutorials and one sort-of-demo to SCaLE. [22:59:33] ok. Who are your audiences, and what do they want? what motivates them? [23:00:18] I especially want to hear places where the audience *is different from you* and *has different, but legitimate motivations* [23:00:21] sumanah: Well, I'd like to help new programmers make cool stuff, so the tutorials are focused on them (mentorship, which you got above), and the demo is of free games, so I guess that falls under Free Software outreach. [23:00:59] sumanah: I'd imagine the difference in the audience is, I have some knowledge that I think they may not have, so I want to impart that information so that they can better understand something. [23:01:29] That's it? they're alternate-universe Marks who just happen not to know this thing? [23:01:43] We can focus specifically on SCaLE and LibrePlanet -- I know more about SCaLE and so am looking up past LPs [23:02:13] sumanah: Well, no, clearly not, but that's the difference I'm playing to when I came up with the proposals [23:02:23] Who are they and what do they want/ [23:02:24] ? [23:02:58] "new programmers" is broad; can you think of a specific persona maybe? I find that helps me [23:03:16] sumanah: Well, the one group is newish programmers, or programmers looking to learn a new language, who might find a good foothold in the projects I present--both of them being projects with easy ways to add small features [23:03:35] In particular, I'm presenting on our Lua framework and the Etherpad Lite plugin framework. [23:04:12] ....and thinking about adding one proposal for another Lua API that I apparently had deemed less important, not sure why [23:04:31] (I mentioned parkour & bridge because heck, I've keynoted at a tech conference whilst doing stand-up comedy, and others have presented at tech conferences with presentations kind of like your "what FLOSS can learn from parkour and vice versa" post) [23:04:43] Hm, interesting. [23:05:05] I guess I could try reading back through my blog (which is probably down ATM) and try to extract more talks from there [23:05:13] That's a good idea, I think [23:05:22] or interesting email conversation threads you've been in [23:05:31] Agreed. [23:06:11] New programmers looking to learn a new language might also be interested in the Parsoid stuff you've done, as an overview of a real project made with that framework [23:06:41] New programmers also often benefit from case studies - "I did this hard thing and here are the obstacles and lessons therein" [23:07:01] especially if the presenter is entertaining [23:07:29] sumanah: I guess so. I can see myself saying something like "I graduated University. And then the WMF hired me. And then they told me I would be on the team rewriting their Parser" and being increasingly flabbergasted [23:07:33] so, it's been 10 min, that's my advice. More empathy is probably the biggest thing [23:07:42] sumanah: Arigatou :) [23:08:40] oh, and given what I see at http://libreplanet.org/wiki/LibrePlanet2012/Schedule , I think your more zealotry-ish interests and blog posts will be of interest as proposals, as well [23:09:07] wait, when you say "take issue" did you mean that's offensive? [23:09:10] or just that you disagree? [23:09:44] sumanah: Just that I disagree - it implies that I do things that are unreasonable which, given the importance of digital freedom in the digital age, I'm not sure is the case. [23:09:58] I mean, if I were breaking windows or crashing banking servers I'd understand the word [23:10:11] have you ever heard the phrase "moving the goalposts"? :) [23:11:15] sumanah: Seems pretty easy to understand, and really I'm willing to be proven wrong, but I don't think I've done anything so unreasonable to further my cause to merit the "zealot" moniker. [23:11:16] OK, let's see whether I can get https://meta.wikimedia.org/wiki/Wikimedia_Blog/Drafts/Lead_our_development_process_as_a_volunteer_product_manager out the door to the blog post reviewers today [23:11:30] sumanah: And yes, it does strike me as having negative connotations, though not necessarily insulting [23:11:43] Some other time re the zealotry fine-distinctions conversation. [23:11:48] *nod* [23:11:49] Cheers [23:13:07] <^demon> You know, "Gerrit Code Review - A Quick Introduction" is better than any of the awful crappy docs I wrote. [23:13:44] <^demon> Really should've just written (very brief) docs about our process, and linked liberally to the gerrit docs. [23:16:43] ^demon: what are your allegedly crappy docs? [23:16:54] <^demon> Everything on mediawiki.org [23:16:59] <^demon> They're awful. [23:17:05] I think our docs are not bad. They managed to get me do my first commit. [23:17:05] <^demon> (At least anything I wrote) [23:17:18] So they must be fool-proof and what not [23:17:20] <^demon> [[Git/Workflow]] is the worst offender. [23:17:29] <^demon> [[Git/Tutorial]] is the only good one--because I didn't write it. [23:19:05] * sumanah raises hand