[12:04:49] wikipedia logs are stored in an sql database? [12:10:50] Talk to drdee when he's around, I don't think so diegolo. [12:15:46] milimetric: hi [12:15:53] milimetric: I got side-tracked on limn [12:15:53] morning! [12:15:56] no prob [12:16:06] milimetric: I need to get back and finish that deb, I know there were some problems on coco [12:16:10] do you think the debianization is stable enough to put it in a script [12:16:21] what problems? [12:16:49] milimetric: :) I don't think so too https://www.mediawiki.org/wiki/EventLogging [12:16:53] milimetric: last time I remember that coco was in the package but the exec server/server.co didn't work or something like that [12:17:12] it did work! :) [12:17:19] hang on, I'll show you [12:17:20] milimetric: it did ? so the package is all good ? [12:17:51] milimetric: show me please [12:17:58] this was the puppet change that Andrew made: https://gerrit.wikimedia.org/r/#/c/49710/2..3/modules/limn/templates/limn.init.erb [12:18:50] milimetric: oh, but now we have duplicate functionality. This is in both limn.upstart and limn.init.erb [12:18:57] milimetric: how do these two interact with each-other ? [12:19:33] um, not sure, but it's ok, let's focus on the Deb [12:19:59] milimetric: we're focusing on the deb :) [12:20:01] milimetric: https://github.com/wikimedia/limn/blob/master/debian/limn.upstart [12:20:02] also, about what I was saying earlier - it's nonsense because coke is in node_modules [12:20:20] milimetric: https://gerrit.wikimedia.org/r/#/c/49710/3/modules/limn/templates/limn.init.erb [12:20:31] milimetric: these two files contain almost the same things ^^ [12:21:07] hm.... [12:21:24] milimetric: can you talk to Andrew about this please when he's around ? [12:21:30] sure, will do [12:21:38] milimetric: ok, now for your question [12:21:52] milimetric: if the debianization is stable enough to put it in a script ? [12:22:12] but I think that's ok, I think we just delete limn.upstart and the puppet module creates it dynamically [12:22:26] milimetric: you want to automate debianization right ? [12:22:39] yeah, it seems it'd be ok [12:22:59] so I think lowering the node version to > 0.8.1 [12:23:12] and npm to > 0.1.40 [12:23:20] so I don't have to do -uc -us or whatever [12:23:35] but wait why was I doing -d? [12:23:45] milimetric: you can control the nodejs dep version here https://github.com/wikimedia/limn/blob/develop/debian/control#L15 [12:24:12] milimetric: you can control the npm dep version here https://github.com/wikimedia/limn/blob/develop/debian/control#L5 [12:24:55] milimetric: -us -uc isn't related to dep versions. You do -us -uc to get an unsigned package. You will need to generate yourself some GPG keys to get rid of -us -uc [12:25:23] hm, ok, how do I do that? [12:25:36] is it something that most people would do or could it be in the script? [12:25:50] (you don't have to ping my name btw, we're the only ones around :)) [12:26:04] milimetric: https://help.ubuntu.com/community/GnuPrivacyGuardHowto#Using_GnuPG_to_generate_a_key [12:26:17] hehe [12:26:18] ok [12:26:57] well, you want to automate debian package creation so you can get Travis to do it right ? [12:27:39] You can tell Travis to -us -uc. I mean, I think you want to do this so you can let Travis tell you if there were errors in the debianization [12:28:05] right ? [12:29:11] yes [12:29:32] not really so Travis can do it [12:29:37] more so I don't have to [12:29:47] and eventually so Jenkins can do it [12:29:54] Travis can do npm stuff [12:30:26] well after Travis makes the .deb, how do you get it from Travis ? [12:30:36] because Travis has ephemeral containers in which it does the builds [12:30:46] Oh, and the last thing is, can we put the generated limn*.deb, limn*.changes files in the current directory? [12:30:48] and AFAIK Travis does not expose the filesystem ... :| [12:31:09] milimetric: you want to put .debs in the git repo ? [12:31:11] you could like ftp it up somewhere, or use git to commit it [12:31:51] no, so the main idea is to make debianization easy for us. And we'll manually build it for now. [12:32:43] if you find an ftp on wikimedia with public access then I'll take care of it uploading debs to it [12:32:46] deal ? [12:33:36] well, no, you shouldn't have to do that. [12:33:50] it'll be pretty rare that we build a deb anyway [12:34:26] but what do you think, can we get the generated build files to go into ./debian/limn/build or something like that? [12:35:00] then we can do rm -rf ./debian/limn/ as the cleanup [12:37:46] milimetric: we can do that, yes [12:38:11] cool, I think then the Deb.md could just be a script basically [12:41:10] We tried that before with upd-filters. We can make a script for it, but debianization should be automated for the most part, but manually checked. Sometimes it can go wrong and you need to watch out for all the debug messages to make sure you get it right [12:41:26] But we'll do a script for it [12:43:05] how often do you expect to do releases btw ? [12:43:14] each week, bi-weekly, monthly ? [12:44:54] milimetric: ^^ [12:44:57] no idea right now [12:46:31] ok, so after we make the script you wrote about above, can I mark Limn debianization as done in Asana ? [12:46:59] great guide on the gpg key, just made mine [12:47:13] oh yeah, I think you can mark it done right now [12:47:20] oh cool [12:47:21] the script is just a nice-to-have and I can work on it [12:47:28] ok [12:47:39] yeah, I don't have anything to do right now anyway [12:48:19] so where does debianization say where to put the .deb files? [12:51:41] it does not have an --output parameter [12:51:43] unfortunately [12:52:03] but you know for sure that the .deb and all the other stuff is written to the parent directory of limn/ [12:52:20] so if you have limn in /home/dan/limn , you'll find the output in /home/dan/ [12:53:53] for every build you make you want to store the .dsc , .tar.gz and the .deb [12:54:17] and you find all of those in /home/dan/ (given that you have the paths as in the example above ^^) [14:11:47] yo [14:12:59] diegolo: we have many types of logs :) traffic logs are stored in flat files on hdfs, event logging is stored in mysql [14:13:36] drdee ;) [14:15:20] https://github.com/MrMEEE/bumblebee-Old-and-abbandoned/commit/a047be [14:15:37] probably one of the biggest bugs in history :) [14:19:01] also making the top bugs list in history is probably this one as well http://www.youtube.com/watch?v=uW3keAPJqU8#t=5m29s [14:22:18] nice links! [14:22:31] :) [14:50:22] hey sorry I was gone for a while. I'm back [14:50:44] welcome back :) [14:51:40] oh average_drifter, did you want me to delete the debianization branch? [14:52:07] milimetric: already done [14:52:13] oh, for limn ! [14:52:15] hm, it's still up on wikimedia [14:52:16] yea [14:52:17] yes please [14:52:18] k [14:52:43] k, done [14:53:09] that bumblebee thing is absolutely histerical [14:53:34] I showed my wife, she liked the cat pictures :) [14:54:25] milimetric: haha, nice :) [15:31:44] k average_drifter, I've pushed the debianize script [15:31:47] it's in ./bin/debianize [15:32:14] it seems to work ok for me, I'm sure it can be a lot fancier but we can do that later [15:32:32] Thank You Very Much!!! [15:34:00] (for seeing this debianization stuff to completion - it's very useful) [15:55:46] average_drifter wanna dive into the 500M bump? [15:59:51] morning kraigparkinson [16:01:54] morning everyone [16:09:38] kraigparkinson: hi :) [16:10:10] drdee: yes, after the meeting with Kraig [16:10:56] on some calls this morning for a bit, probably wont surface beyond text until the sprint planning meeting. [16:11:10] er, scrum. :) [16:12:02] kraigparkinson: in 1-2h ? [16:12:32] ok, then ping me please when you're free [16:16:24] drdee: ok, we can talk now [16:18:06] drdee: https://plus.google.com/hangouts/_/3cab98f6fb984151cd7a3d29639a00f1e750669c [16:19:14] brb, hangout link still open [16:30:26] brb phonecall [16:35:12] back [16:35:32] milimetric: do you have standup today ? [16:48:09] http://corner.mimuw.edu.pl/?p=354 [16:50:46] [travis-ci] master/8235806 (#48 by Diederik van Liere): The build has errored. http://travis-ci.org/wikimedia/kraken/builds/4983914 [17:55:35] average_drifter - sorry didn't see - yes we have standup [17:58:07] milimetric: is that right now ? [17:58:10] having trouble connecting to the hangout (again) [17:58:14] yes [17:58:30] kraig, are you on wifi? [17:58:35] try hardwire [17:59:06] ohai [17:59:13] and if you have multiple google accounts, and you don't wanna mess with the parameters like this: ?authuser=1, then try just signing out of your other accounts. [18:01:08] standuuuuppppp [18:01:13] drdee [18:01:14] Thanks, that's what it is. [18:14:33] ottomata: maybe git checkout into a detached head representing the checkpoint right before you want review, git push that into the repo (which means bypassing gerrit for the historical part). then git checkout HEAD, and git review that stack. it should see the delta. [18:21:19] hey drdee [18:21:56] I cannot attend 1h 30 of meetings this morning as I am behind with urgent work, do you prefer I attend the demo or (part of) the sprint planning session? [18:22:10] also milimetric ^ [18:22:20] planning [18:22:25] ok DarTar [18:22:31] alright [18:22:40] I'll join at 11 then [18:22:40] you are welcome anytime! :) [18:27:13] milimetric: I have a lot of feedback on the dashboards, mostly cosmetic changes to improve usability, looking forward to chat [18:31:54] can you guys invite me to the DEMO friday please ? [18:31:56] milimetric: ^^ [18:32:04] milimetric: uhm, stefan.petrea@gmail.com [18:32:37] milimetric: worked [18:59:04] is there a room in the office for the sprint planning or is it really collab on 3? [18:59:27] DarTar: we're still having a demo in isadore on 3, btw [18:59:40] k joining you there [18:59:42] k [19:02:13] i'll attempt to snag the room [19:05:56] milimetric: do we get invited to the sprint planning ? [19:05:59] by e-mail ? [19:06:02] I'm wondering if I'm in the wrong hangout... [19:06:05] you should have been [19:06:10] yeah, everyone's in [19:06:14] ok, link is: https://plus.google.com/hangouts/_/a838646099396288ed3c328afe579a90585d1aab [19:06:19] stefan, I'm inviting you now [19:07:09] milimetric: thanks [19:07:18] My computer's crashing guys - gotta restart [19:07:38] ... [19:07:38] https://plus.google.com/hangouts/_/calendar/ZGFuZHJlZXNjdUB3aWtpbWVkaWEub3Jn.q9pqfp6e4lr3j9rrsmj8r0t350 [19:07:44] for those engaging in limn sprint planning. [19:08:05] kraigparkinson! [19:08:07] drdee! [19:08:15] !! [19:09:13] can't join... [19:09:16] for some reason [19:09:18] tried multiple times [19:09:25] multiple sign in? :) [19:09:29] ?authuser=1 [19:09:34] or whatever... [19:12:27] tried [19:12:33] no luck :( [19:12:54] can't believe this [19:13:06] milimetric: were you able to join ? [19:13:43] wow, this thing is not very :) , it's more like :( [19:14:03] milimetric is in [19:14:10] i can try to invite you? [19:14:13] please do [19:14:54] i just did it, though I'm not sure that will show up for you [19:16:13] erosen: thanks ! [19:16:19] np [19:27:01] does limn or kraken do smoothing of the data? [19:27:10] just wondering about this [19:27:27] wikistats calls this "recalc" I think [19:31:46] DarTar: reminder about email :D [19:36:14] YuviPanda: I'm actually talking about our plans for EL data visualization in the analytics sprint planning session [19:38:01] DarTar: ah, sweet! [19:42:24] YuviPanda: I got sidetracked by some urgent data problems we discovered yesterdays with E3, I'll start the thread today [19:42:34] DarTar: sweet! Thank you :) [19:44:21] btw we're getting a couple of requests for remote attendance of the EventLogging workshop on March 7, if you're interested in attending please sign up and mark that you're remote, we'll stream if we have a critical mass [20:02:41] milimetric: maybe resume a hang out / Skype when you're back? More effective than IRC methinks [20:02:44] brb [20:02:59] DarTar: just saw the EL workshop thing. Will do [20:05:05] great [20:10:22] milimetric: i need to grab lunch :( [20:10:28] i will try to be back asap [20:10:31] no prob [20:10:35] I'm just talking to Dario [20:10:38] DarTar, ready to hangout [20:11:53] kk [20:12:00] oh DarTar-away ^^ [20:12:10] I invited you to the hangout, lemme know if it doesn't reach [20:14:37] milimetric: finding a room, brb [20:14:42] cool [20:15:07] milimetric: ps. i wrote up subtasks off the top of my head for the Viewport stuff. https://app.asana.com/0/701374192205/4002219179712 [20:15:08] brb [20:15:09] ! [20:15:27] cool, thanks! [20:15:30] GO EAT :) [20:16:14] milimetric: is that my personal gmail account? [20:16:33] no, it was @wikimedia.org [20:16:40] but it said you were outside of the org [20:16:47] woot [20:16:52] lol :) [20:17:00] let me try again [20:17:02] i'll type it in manually to make sure [20:32:15] anyone else having gapps problems? [20:33:00] milimetric? [20:33:03] DarTar? [20:36:38] working fine here [20:38:27] my fault. [20:38:34] i changed things through google, not ldap [21:30:50] grrr, got dropped... [21:30:51] coming back... [21:37:52] Gulthor: :-] [21:38:58] Gulthor: I guess limn contact is Dschoon :-D [21:39:09] though he is probably eating right now (he is in SF) [21:39:15] GMT-8 [21:39:25] I'm pseudo-about! [21:39:42] dschoon: hi david :-] [21:39:46] Howdy! [21:39:49] dschoon: Gulthor is my brother [21:39:55] limn's doting parents are Dan Andreescu (milimetric) and myself [21:40:12] working on a project that use some big real time data and who might be interested in limn / d3.js [21:40:13] ;) [21:40:29] Hi Gulthor! [21:45:34] hi Gulthor! [21:45:39] (we were all in a meeting :) [21:45:54] Gulthor, what's your project about? [21:47:45] I am on phone with him, about to hang [21:48:01] coolio. i'll be around for a bit. [21:49:37] Gulthor: bn :-] [21:49:43] Back, we're kind of talkative brothers [21:49:46] leaving out, just wanted to introduce Gulthor around [21:49:55] Good night hashar [21:49:59] dschoon: see you on monday :-] [21:53:41] see you, man [21:53:56] Howdy Gulthor, nice to meet you. [21:54:39] dschoon: I'm working on a financial application (portfolio backtesting framework) ; being a d3.js user, I might be using limn [21:55:29] oh, excellent [21:55:38] i am interested both in your interest in limn, and in your application :) [21:55:41] I'm still in the early development stage, mostly coding the back end, but hashar and I were discussion front end solutions [21:55:53] *discussing [21:56:35] *nod* [21:57:12] well, limn is a young project. it's only been around 6mo or so, and it didn't have anyone dedicated to it until December [21:57:19] but! [21:57:31] that means there's lots of opportunity to get involved and make a big impact [21:57:41] it sounds promising [22:00:29] I'll follow the development closely and provide feedback if I decide to use limn, which is quite likely [22:01:12] that sounds great. [22:01:39] you should definitely send your thoughts and feature requests our way while you're making up your mind [22:02:20] however I'm a self-taught developer who recently got into JavaScript/Node.JS, so it'll take some time before I can be of any help :P [22:02:22] i'll be honest -- b/c limn is young, and driven by our use-cases here, it's not the easiest project to just drop in yet. [22:02:27] no worries :) [22:02:31] i'm self-taught as well [22:03:11] * YuviPanda too! [22:03:22] (going semi-afk, cooking - still reading though) [22:06:36] Gulthor - I' [22:06:44] I'm the other Limn doting parent :) [22:07:24] I'll paste some links of some dashboards and examples we have up right now, and I'm available to talk. I have to leave soon, but I'll PM you my email [22:07:51] Ok, sure! Thanks [22:08:15] These are two dashboards we just did here, along with the github repository where the graph, dashboard, and datasource definitions live: [22:08:17] http://dev-reportcard.wmflabs.org/dashboards/metrics [22:08:20] http://dev-reportcard.wmflabs.org/dashboards/features [22:08:33] https://github.com/wikimedia/limn-editor-engagement [22:09:45] Cool! It looks like I won't have to reinvent the wheel [22:11:06] yeah, the focus right now is to make it easy for anyone to get their own Limn instance up and running. Right now the front-end and back-end are tied together [22:11:50] one general note: the approach is a lot less like Rickshaw.js and a lot more like "The Grammar of Graphics" [22:14:46] milimetric: maybe we should rename the project ggGUIplot [22:15:04] oh boy [22:15:06] g-g-g-g-g-g-.. [22:15:37] :P [22:17:03] I'll look at limn's source when I get rid of that last back end bug I've been fighting against for the last 2 hours ;p [22:32:53] good weekend everyone, take care and see most of you next week :) [22:39:23] laterz [22:53:24] laatas! [23:47:42] [travis-ci] master/725b007 (#49 by Diederik van Liere): The build has errored. http://travis-ci.org/wikimedia/kraken/builds/4992530 [23:49:47] [travis-ci] master/83e9995 (#50 by Diederik van Liere): The build has errored. http://travis-ci.org/wikimedia/kraken/builds/4992614 [23:51:24] [travis-ci] master/32f4abc (#51 by Diederik van Liere): The build has errored. http://travis-ci.org/wikimedia/kraken/builds/4992714