[00:04:49] anything else before I go to bed? [00:07:34] good night, everyone [00:07:42] MaxSem: nite nite [00:08:00] MaxSem: Спокойной ночи! [00:32:57] JellyBean is out for the Nexus S ... maybe i should finally upgrade [00:33:06] toobad ota still doesn't work [00:33:19] preilly: did you have to unlock the bootloader when you upgraded your nexus s ? [00:33:24] to ICS [00:34:01] tfinc: nope [00:39:26] unlocking boot loader wipes the device ... lame ... will have to do tonight [00:52:19] preilly: can you take a look at https://github.com/wikimedia/WLMMobile/pull/64/files ? [01:09:22] hey jdlrobson [01:11:19] hey yuvipanda ! [01:12:09] i seem to be irregularly available at odd hours only. sigh. [01:12:16] jdlrobson: re #62 [01:12:23] +1 on messages [01:12:40] +1 on link (I mentioned that in my pull req message too) [01:12:46] but, I don't think the last one makes sense [01:13:04] no worries yuvipanda good to see you :) [01:13:14] yuvipanda: yo [01:13:19] will update (1) and (2) during the course of the day and update the pull req~ [01:13:24] tfinc: yo! [01:13:27] yuvipanda: did they find the phone ? [01:13:41] tfinc: nope! :( [01:13:52] jdlrobson: also re: 64, shouldn't we not do requests at higher zooms? [01:13:58] asking for all of europe might not be a good idea [01:14:07] tfinc: I called on friday, will call again today [01:14:17] yuvipanda: why not last one? [01:14:21] also what about my latest comment [01:14:24] yuvipanda: i'm getting caught back up after vacation. which WLM tasks are you tackling now ? [01:14:36] yuvipanda is looking at my uploads as i understand it [01:14:40] tfinc: my uploads [01:14:51] great [01:15:29] jdlrobson: which latest comment? [01:15:33] the switzerland one [01:15:34] ? [01:16:45] jdlrobson: also, re: getting them serverside, you'll need to get pages from commons uploaded per user, with a particular category, and parse out the monument details from that [01:16:53] yup yuvipanda [01:17:05] yuvipanda: i'm guessing it will be complicated but it seems like the right thing [01:17:10] the other thing to do is to stop using a tab [01:17:20] I think that's better UX in the current situation [01:17:25] as the tabs suggest the two lists are related [01:17:31] +1 [01:17:34] a button in top right for example [01:17:40] they edon't seem related [01:17:51] jdlrobson: I just merged your change [01:17:53] killing the tab and moving it to a different page makes sense to me [01:17:56] i'm looking at back button behavior. I think i'm going to retire for the day as my brain is exploding [01:18:08] heh [01:18:27] jdlrobson: i think locally caching is fine for this version, and we can parse stuff out of commons next time [01:18:47] tfinc: i'll call them again today and reply on the thread [01:20:23] k [01:20:39] yuvipanda: I'm thinking for back behavior we should only keep a type of page once in the history [01:20:49] +1 [01:21:16] I need to play around with it some more though to see if that's the right thing to do ™ [01:21:52] and I will do that on the bus ride home [01:21:57] see you guys later :D [01:22:02] yuvipanda, jdlrobson: I set up IRC notifications on https://github.com/wikimedia/WLMMobile [01:22:08] \o/ [01:22:34] \o/ [01:22:40] jdlrobson: seeya! [01:22:45] preilly: thanks! [01:23:37] preilly: if such thing existed I'd post a picture of a cheerleader with pom poms with the caption 'gooooo preilly!' right now [01:23:41] I don't know why I would.. I just would [01:23:44] and such thing doesn't exist [01:23:51] damn you internet. damn you. [01:23:54] things that do not exist on the internet? wow [01:38:11] http://imgur.com/CnUZw [01:45:01] anything not on the internet already can me made on the internet. [15:25:12] hi MaxSem. For the mobileview section text, what html standards are being used? [15:28:06] HTML5 [15:31:00] ok, I meant more about coding standards for the style you all are using. i.e. will everything always be inside a div, will sound links be wrapped in spans, will img always have width and height parameters. [15:35:54] I ask because I'm trying to make the articles appear on the page as 'pretty' as I can make it while making sure I handle any formatting that may occur. Those two goals are at odds with each other so I'm looking for any style guides you all use when making html text blocks. [15:45:47] imgs always have dimensions [15:46:09] sections are wrapped in divs [15:46:34] the best way to figure out how this stuff works is to take a look at it yourself [15:47:39] yes, both I and an HTML5 expert here are. I just didn't know if you had a style guide that would help expedite the process. [15:48:10] we don't have one [15:48:47] the best person to talk about nuances of HTML is jdlrobson [15:49:09] ok, thank you [16:14:15] notice: /Stage[last]/Ldap::Client::Instance-finish/Exec[check_autofs]: Dependency User[maxsem] has failures: true [16:14:17] hehe [16:33:21] awjr, yo [16:33:26] whatsup MaxSem [16:33:59] i started playing around with your puppet manifest on labs last night, but i think im still not grokking how to test it correctly [16:34:24] yeah, I was going to ask if wlmpuppet is your creation [16:36:00] I just edit the checkout then force a puppet run [16:37:27] yeah i set it up yesterday [16:37:52] I'm playing on mobile-wlm to avoid collisions [16:37:56] ok sounds good [16:38:53] if you want to play around on a more vanilla setup, let me know and i can step away from wlmmobile [16:41:14] MaxSem i havent figured out how to get puppet to apply certain classes to the local instance - have you? [16:44:39] puppet apply will do it but i'd like a dry-run option [16:45:48] node "i-000002bc.pmtpa.wmflabs" { [16:45:48] include standard, [16:45:48] misc::wlm [16:45:48] } [16:45:58] in site.pp [16:46:02] oh well dang [16:46:04] thanks [16:56:34] MaxSem from a dry-run things look pretty good; maybe we should add the update_from_toolserver.sh file to puppet as well [16:57:08] I'd rather store it somewhere else [16:57:29] why is that? [16:57:42] awjr, http://dpaste.org/vU2Tj/ <-- here's how it looks for me [16:57:59] yeah i had ato take out the cwd as well [16:58:19] well, it's mostly related to WLM instead of ops things [16:58:56] we could have it get checked out from somewhere if it doesn't exist, like some git repo [16:59:01] or even svn [17:00:02] but we can worry about that later [17:00:11] i guess it's better to just get the system built first :p [17:05:46] ah grrr, our puppet repo is not commitable from WIndows:P [17:06:31] doh! [17:06:57] what do you think about using apaches::files and adding apc as a package? [17:07:43] it would require a different docroot but looks like we'd get a better apache conf [17:07:59] sounds good [17:11:58] we probably want ganglia also [17:12:57] I think we can build a very basic site first and then whel we have the API running we can add more advanced stuff to it [17:13:08] s/whel/when/ [17:13:10] yeah [17:13:12] totally [17:14:11] we probably don't want cron running as root [17:16:10] but we can fix up users/perms later [17:16:27] #!/usr/bin/sh [17:16:34] ! [17:16:35] sudo -u maxsem ... [17:16:46] :P [17:17:41] hmm, puppet appears to hang after i added apaches::files and apaches::service [17:17:58] dry run? [17:18:47] nope, I test with puppetd -tv [17:19:07] hmm apaches::service will try to sync mediawiki [17:19:09] i dont think we want that [17:21:06] im kinda surprised apaches::packages doesn't include apc [17:21:56] the API is pretty small [17:24:14] do you mean you don't think apc is worth it? [17:25:18] I mean it's not first priority [17:25:32] greetings all [17:26:05] things not to do. upgrade your phone to jelly bean at 1am ... spend the next many hours tinkering with it [17:27:40] tinkering or trying to unbrick it?:P [17:27:56] tinkering. the upgrade went perfectly fine. [17:28:06] :D [17:28:27] 2.3.6 -> 2.3.6 (+tethering) -> 4.0.4 -> 4.1.1 [17:28:39] ooo [17:28:46] * yuvipanda should move when he's not on 3g [17:29:03] it was pretty easy to use clockwork mode to install the new release from a zip file [17:29:12] i've given up on every getting any other OTA update [17:29:25] so i just had to unlock the booloader, install clockworkmod, and then install the zip [17:29:28] pretty simple stuff [17:29:45] i'm glad that i waited till jb as ics was pretty slow on my nexus s [17:29:57] awjr: have you been using stock browser or chrome on ics ? [17:30:12] tfinc i generally use chrome on ics [17:30:35] but i play around with dolphin, stock, and just for kicks sometimes mozilla [17:31:00] tfinc how's jb vs ics? [17:31:23] i probably wont know for 6 months since verizon likes to take its sweet time with ota updates [17:31:39] awjr: i never used ics much feature wise so i can't say. speed wise its way faster from my superficial 10min of testing [17:31:47] heh [17:31:53] awjr: be a man and upgrade it by hand [17:31:54] ;) [17:31:56] haha [17:31:59] lol [17:32:07] s/man/nerd [17:32:29] i'm eager to see how jb does with battery life [17:32:32] tfinc i've futzed with phone os's enough to know im perfectly happy to impatiently wait for my ota [17:32:44] brick enough phones ? [17:32:47] yes [17:32:49] and [17:32:59] there's nothing like going to make a phone call and having your phone reboot [17:33:24] jdlrobson: tests fixed. Fixing your other comments now :) [17:33:38] wasn't that a feature of the phone that mdale had ages ago ? [17:33:40] * tfinc ducks [17:34:10] it was also a feature of my ancient galaxy 3 running stock everything [17:34:18] it doesn't reboot - but opening the 'phone' app takes about 15s [17:34:21] tfinc: careful I am in the office so I can actually throw things ;) [17:34:25] lol [17:34:31] mdale are you around all week? [17:34:33] ( like an old open moko phone ) [17:34:42] which is about all its good for [17:34:48] i wanted open moko to be so much more [17:35:04] mdale: it still has some sort of geek cred, at least here. [17:35:09] ;) [17:35:23] mdale: its more fun when i poke fun and were in projectile range [17:35:23] that is possibly because most people haven't actually tried using one [17:35:31] hehe [17:35:31] plus its fun to get you active in this channel [17:36:09] meetings 11-3 today ... joy [17:36:16] anyone need anything from me ? [17:36:24] awjr: what time will we see you tomorrow ? [17:38:08] tfinc: pm? [17:38:54] tfinc: sometime between 930 and 10am probably [17:38:58] awjr: k [17:39:05] i arrive in sf tonight ~1130 [17:39:49] although i had been booked on a flight that left an hour earlier that has already been delayed 2 hours. woulda missed my connection so switched flights, but i have a feeling it'll probably get delayed too [17:40:52] im stoked to be in SF [17:41:50] beat the heat, be reminded that i work with real people instead of IRC nicks... [17:42:02] i guess i'll have to start putting pants on to go to work [17:42:07] but it's a sacrifice im willing to make [17:42:51] awjr: bring your hat and pants if you want to work in the office. we have standards. [17:42:57] awjr: did I remind you I'm by the beach in shorts when working? [17:43:14] which hasn't been enough recently, with jetlag + college. Should move up [17:43:24] yuvipanda: nice [17:43:48] yuvipanda: how are studies going? you done yet? [17:44:06] tfinc: [17:44:08] mobile team dresscode: pants [17:44:24] tfinc: at least till november is full time. [17:44:29] 9 more weeks, as a friend is calculating [17:44:44] what happens after november. i thought you didn't go half time till january ? [17:44:46] tfinc: also, I found that my Computer Graphics Lab notebook was about 90% plagiarized directly from wikipedia :P [17:45:03] yuvipanda: did they attribute ? [17:45:06] tfinc: after november I get exams, and it gets into a weird unpredictable state with classes at random days [17:45:10] tfinc: nope [17:45:13] those bastards [17:45:26] tfinc: plus they copied from wrong articles, so they didn't even plagiarize properly [17:45:31] they failed at failing, really [17:46:29] I wonder if there's a 'hall of shame' for such things [17:46:35] yuvipanda: you could make one [17:47:19] on my list of increasingly long to-do things. [17:47:30] i'm manning a wikipedia stall this weekend [17:47:48] (along with Arun and Srikanth!) [17:48:05] jdlrobson: in your latest wlm buld are you seeing duplicate listings in the map view for the same place ? [17:48:33] yuvipanda: nice. get people to test the app. [17:48:41] tfinc: WLM or wiki? [17:48:45] WLM [17:48:54] yessir! [17:48:58] I did see that yesterday actually but I never verified if it was the api or the client code - i'll explore again [17:48:59] yuvipanda: we have yet to see any weekly status updates from you. think you'll be able to do them ? [17:49:05] jdlrobson: i'll open a bug [17:49:10] tfinc: last week was lost to jet lag. I can do from this week [17:49:44] tfinc: think there's a way to do them 'running' rather than fixed? An ether pad or a wiki page? [17:51:16] yuvipanda: if you want to do a short sprint with jrobson then feel free to use the etherpad. otherwise put it in bugzilla. etherpad is NOT long term storage and Ryan_Lane will stab anyone who uses it as such [17:51:25] greetings jorm [17:51:34] jdlrobson: bug open https://bugzilla.wikimedia.org/show_bug.cgi?id=38645 [17:52:47] MaxSem: awjr: what's left to do in our minimal puppet setup? i think we can have the wlm backend on a real box today [17:53:24] the latest of MaxSem's changes that i've seen look good to me and run fine on a vanila labs instance i set up [17:53:35] MaxSem have you committed the changes yet? [17:53:49] nope, moving this stuff to a Linux VM to commit [17:54:20] MaxSem: in labs? [17:54:24] MaxSem: or locally? [17:54:30] i like how installing a new os is like getting a brand new machine. /me must stop tinkering with JB [17:54:30] locally [17:56:30] tfinc: I was talking about something like https://www.mediawiki.org/wiki/User:Yuvipanda/Mobile_status_updates for status updates [17:57:08] tfinc: you could watchlist that, and I could mail out snapshots to mobile-tech every weekend if you want [18:00:26] ahh [18:00:54] yuvipanda: that sounds fine [18:01:08] tfinc: do you want me to mail out snapshots every weekend or would just the page be fine? [18:07:36] here we go: https://gerrit.wikimedia.org/r/16530 [18:09:16] F~~L [18:09:17] MaxSem: Syntax error at 'bash'; expected '}' at ./manifests/misc/wlm.pp:16 err: Try 'puppet help parser validate' for usage [18:09:34] yeah [18:09:52] * yuvipanda tries to complete F~~L, fails [18:10:24] MaxSem: you might also try to get an op like binasher, Ryan_Lane, notpeter, or paravoid to take a look at that [18:10:43] sure [18:10:48] lint first [18:11:03] of course :) [18:14:16] MaxSem: are you trying to do a git checkout? [18:15:29] MaxSem: because you could do that like this: http://pastebin.mozilla.org/1713417 [18:15:56] first of all, it's in SVN :) [18:16:20] MaxSem: well then change this one line command => "git clone ${gitrepo_real} .", [18:16:23] and due to this, automatic updates are scary [18:16:30] in the example I sent you [18:17:17] we'll have to check every commit manually first and only then update [18:17:45] we could also set up a deployment branch [18:18:05] which is probably a good idea anyway [18:18:40] awjr, as long as it's out of WMF engineers' solitary control, I'm against automatic updates [18:19:01] MaxSem the deployment branch could be within our control [18:19:30] jdlrobson: updated https://github.com/wikimedia/WLMMobile/pull/62 [18:19:44] like a svn repo in a different place with an external to WLM? [18:19:48] MaxSem: you also want the generic apache manifest [18:19:53] MaxSem: and you need to make a vhost [18:20:27] apache::service? [18:22:35] MaxSem i dont think we'd need/want to use externals - couldn't we get a branch set up for us for WMF-specific deployments in the TS repo? [18:23:02] awjr, it will have the same access permissions as the rest [18:23:17] oh we can't get more fine-grained permissions? [18:24:49] MaxSem: have you run grep and looked for apache_site ? [18:25:28] MaxSem: look at puppet/manifests/misc-servers.pp [18:26:01] MaxSem: see line 531 class misc::etherpad_lite {¬ [18:27:30] jdlrobson: also https://bugzilla.wikimedia.org/show_bug.cgi?id=38646 [18:27:54] yuvipanda: will look into all this after i've sorted out back [18:27:57] proving a pain [18:28:06] jdlrobson: get a better chair :P [18:28:14] jdlrobson: okay! [18:28:35] I shall file a bit more bugs, look at WPMobile and head off [18:28:41] MaxSem: also mode => 0777; [18:30:01] MaxSem: look at puppet/templates/apache/sites/etherpad_lite.wikimedia.org.erb [18:32:20] MaxSem, awjr: Why isn't this in git btw? [18:32:37] preilly: you mean the WLM stuff? [18:32:41] because it's a volunteer project [18:32:42] awjr: yes [18:32:55] what max said [18:32:57] MaxSem: so can't we make a repo for it? [18:33:02] it's a TS project and it uses an appropriate repo [18:33:22] preilly, so we'll have to drag volunteers over, which would be rude [18:33:28] sounds like something that doesn't need to run on wmf hardware ;) [18:33:56] regardless of what happens we need a better way to maange deployments of this software than svn up'ing from trunk [18:34:05] but srsly, i want to have tim and/or chris give this a security focussed code review. i'd prefer it be in a wmf repo that they can commit to [18:34:22] then we'll have to fork [18:34:41] git-svn if you want to [18:34:48] MaxSem who all besides multichill and platoniedes are involved in the project? [18:35:05] see SVN log [18:35:10] oh right [18:35:19] I don't even know some of these people [18:35:24] i've purged most svn usage from my memory [18:36:10] MaxSem: also in regards to https://gerrit.wikimedia.org/r/#/c/16530/3 you shouldn't have a manifests/misc/wlm.pp [18:36:30] MaxSem: that should be inside of the misc-servers.pp file [18:36:41] awjr, right click on folder, TortoiseSVN --> SVN Log:P [18:36:44] MaxSem: awjr: how can i point people to code review this, is there an websvn url? [18:36:54] i believe there is [18:37:31] binasher: https://svn.toolserver.org/svnroot/p_erfgoed/ [18:38:36] thanks [18:38:41] svn log | grep user [18:38:54] or [18:38:55] svn log | egrep 'r[0-9]+ \| user \|' [18:38:56] binasher, a note for reviewers: we will expose only part of it needed by the app, so users will have only api.php and only with $dbMiserMode = true [18:39:26] binasher: https://fisheye.toolserver.org/changelog/erfgoed [18:41:53] hmm, fisheye looks much fastger than the last time I've seen it [18:42:01] now it's actually useable [18:42:36] oh you can kind of do code review with fisheye [18:43:14] preilly, so what's the difference between manifests/misc and manifests/misc-servers.pp ? [18:50:56] MaxSem: binasher will answer that for you... [18:52:20] misc/ is for per-project misc manifests [19:04:59] [WLMMobile] jdlrobson pushed 7 new commits to master: http://git.io/xd_Ukw [19:04:59] [WLMMobile/master] Show completed uploads in My Uploads page - YuviPanda [19:04:59] [WLMMobile/master] Ensure user is logged in before showing them My Uploads - YuviPanda [19:04:59] [WLMMobile/master] Add detail page for completed uploads - YuviPanda [19:04:59] chrismcmahon: have you looked at: https://fisheye.toolserver.org/changelog/erfgoed at all? [19:05:32] yuvipanda: thoughts on moving to latest (and not so stable) leaflet to fix problems with pinch zoom? [19:05:42] jdlrobson: wh00t! [19:06:17] MaxSem: can I get the link to the etherpad about GeoData again [19:06:23] MaxSem: I can't seem to find it [19:06:36] http://etherpad.wikimedia.org/GeoData [19:07:00] MaxSem: thanks [19:07:03] jdlrobson: i'm not entirely comfortable with that, since they're known to break things in master on random platforms intermittently [19:07:15] jdlrobson: bit me in one of my other mobile apps using Leaflet (OSM editor) [19:07:22] so it gets a -0 from me [19:07:43] preilly: since we haven't heard back from brion about the cordova bugs go ahead and just take them on. do you need anything from me to start working on them ? [19:08:01] yuvipanda: other option is to disable pinch zoom altogether on android [19:08:13] tfinc: nope [19:08:18] preilly: I've used Fisheye at other jobs, the Atlassian tools are pretty nice. I liked Greenhopper a lot for project management, it's similar to Mingle that the E2 team is using. [19:08:44] jdlrobson: it's a -0, so if you think you can freeze on a particular commit and feel comfortable with it, go for it :) [19:09:24] jdlrobson: updating to master fixes leaflet issue with weird zoom? [19:09:40] jdlrobson: Axel was unable to login on his phone [19:10:56] multichill: what's axel's username? [19:11:09] yuvipanda: yup [19:11:19] I linked you with the leaflet issue [19:11:31] where? [19:11:40] i seem to have missed it [19:12:07] https://commons.wikimedia.org/wiki/User:Haxpett [19:14:33] multichill: does he get any error messages? [19:14:55] yuvipanda: on bugzilla [19:15:17] ah, right. /me goes to peek [19:15:23] Ihttps://bugzilla.wikimedia.org/show_bug.cgi?id=38348 [19:15:26] ^ yuvipanda [19:16:42] jdlrobson: https://commons.wikimedia.org/wiki/User_talk:Haxpett#WLM_Mobile_login_problems [19:17:20] jdlrobson: cherrypick the commits or move to a particular commit in master early on and stick to that throughout [19:17:32] cherrypicking might not be possible [19:17:43] * yuvipanda hates how leaflet does super-slow releases, and that was considered 'not an issue' on twitter when i poked them a while back [19:19:16] jdlrobson: too sleepy to make my brain decide if the back button repeat detect algorithm is complete and correct. shall do tomorrow [19:19:28] no worries [19:19:32] i know the feeling [19:19:35] screwed with my mind a lot [19:19:58] jdlrobson: I leave the leaflet decision in your hands. -0 from me on moving to master, but that's not a strong objection. Go ahead with it if you think that's fine. [19:23:09] alright, i'm off [19:23:24] i should now go memorize a few pages of definitions and go sleep. [19:23:46] jdlrobson: will continue with the photo object refactor tomorrow, should be able to extend my uploads to support incomplete downloads too [19:23:55] seems like a good idea [19:24:09] gnite! [19:24:15] yuvipanda: good night [19:26:31] New patchset: Jdlrobson; "follow up to 3002efab (bug 38595)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/16539 [19:26:52] Change abandoned: Jdlrobson; "(no reason)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/16539 [19:30:12] New patchset: Jdlrobson; "rename language id follow up to 3002efab (bug 38595)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/16541 [19:43:49] tfinc: http://etherpad.wikimedia.org/WLM-API-to-WMF [19:44:01] outlines the outstanding steps to get the WLM API running on the cluster [19:47:13] asher wants the API code to undergo a security review by Tim before it runs on the cluster and we need to figure out how we're going to handle deploying code as right now all we have is trunk on the TS svn repo (this would require either forking to WMF git, creating a deployment branch in the repo, or work with the current maintainers to migrate to WMF git) [19:48:55] awjr, tfinc: we could have a call [19:49:16] that works for me - i'm free until ~2pm pacific [19:51:16] lunchies [19:51:19] bbiab [20:02:02] jdlrobson: https://gerrit.wikimedia.org/r/#/c/16541/ [20:07:50] tfinc: One bug to rule them all https://bugzilla.wikimedia.org/showdependencytree.cgi?id=38648&hide_resolved=0 [20:19:25] tfinc: have a minute for a call with max and i to disucss the api migration/ [20:21:33] awjr: nope, in meetings till 3 like i said earlier today [20:21:41] i'm sure preilly can fill in just fine for me [20:22:45] tfinc: I can [20:22:48] tfinc: ok - missed that [20:22:56] preilly, MaxSem you both good for a call now? [20:23:01] awjr: what needs to be discussed? [20:23:03] * MaxSem is [20:23:10] preilly http://etherpad.wikimedia.org/WLM-API-to-WMF [20:23:22] and the timeline around all of that [20:23:55] perhaps we can get multichill on the phone too? [20:24:08] * preilly I'll be back in a few minutes [20:24:12] k [20:40:53] awjr: what would you suggest [20:42:09] jdlrobson: is the string for the 'current' language, or were you just trying to come up with a new key? [20:42:18] the current language [20:42:23] it prints the language of the content [20:42:33] I was just trying to come up with a new kry [20:42:37] jdlrobson then i think it's fine, it should just be explained in the qqq string and provide more context than just 'secondary nav' [20:44:20] New patchset: Jdlrobson; "rename language id follow up to 3002efab (bug 38595)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/16541 [20:44:22] awjr: how about ^ [20:46:15] jdlrobson i think that's fine [20:46:32] cool [20:46:32] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/16541 [20:46:37] thanks for fixing that [20:47:14] jdlrobson are you done with changes for today's deployment? [20:47:19] yup awjr [20:47:24] jdlrobson ok cool [20:47:30] * jdlrobson excited about next beta [20:47:59] * jdlrobson less excited about having to eventually reimplement browser scrolling to get it to work on older phones [20:56:26] jdlrobson does this look right to you? http://www.mediawiki.org/wiki/Extension:MobileFrontend/Deployments/2012-06-24 [20:56:49] oh forgot your most recent change [20:57:27] yup [21:02:54] http://ghostinthestream.net/404 < add this to wikipedia awjr ? [21:02:55] :) [21:03:18] zomg my brain just melted [21:03:33] zomg indeed [21:03:49] yeah i'll see if i can get that in today's deployment [21:05:24] its brion! [21:05:33] hey tfinc [21:05:50] how was memphis etc [21:10:30] brion: nashville as great. memphis was ok. [21:10:30] i'll tell you more in person [21:10:31] * jdlrobson waves at brion [21:10:39] * brion waves jdlrobson  [21:10:55] spiff. i'll be in tomorrow for the workshop fun [21:10:55] * jdlrobson was sadly no longer looking and missed the wave ;-) [21:14:55] jdlrobson, how about http://cat.en.wikipedia.org.meowbify.com/wiki/Main_Page ? [21:14:55] MaxSem: that's step 4 on the Wikipedia mobile roadmap ;-) [21:15:27] wtf [21:15:36] cats awjr . cats. [21:15:48] im allergi [21:15:49] c [21:16:53] preilly you back yet? [21:17:01] some this this just doesn't seem right… http://cat.en.wikipedia.org.meowbify.com/wiki/Vietnam_War [21:17:38] lol [21:19:16] http://24.media.tumblr.com/tumblr_lj73q2iGYH1qawarao1_500.gif [21:19:43] that made me sneeze all over my monitor [21:28:05] ok this is getting crazy now - http://bacolicio.us/http://nyanit.com/cat.en.wikipedia.org.meowbify.com/wiki/Wikipedia [21:28:31] that's like, meta-meta [21:28:44] or meta-meta-meta [21:29:14] I wonder how much of this till the web explodes [21:30:11] brion: https://bugzilla.wikimedia.org/show_bug.cgi?id=38509 < how do I remove the button in the manifest? [21:30:41] jdlrobson: lemme check, i remember seeing it in the docs somewhere [21:30:50] can't seem to find anything :( [21:33:02] http://android-developers.blogspot.com/2012/01/say-goodbye-to-menu-button.html "Set targetSdkVersion to 14" sounds like what i remember [21:33:35] that should still leave things running on older versions, but will tell the system not to do back-compatibility stuff like the menu key [21:34:22] "If you set either minSdkVersion or targetSdkVersion to 11 or higher, the system will not add the legacy overflow button." [21:34:24] there we go [21:34:48] so [21:35:05] hmm [21:35:12] i think we want to keep minSdkVersion at 8 to run on gingerbread [21:35:25] targetSdkVersion is advisory for compat but will keep it running on older vers [21:35:41] mm doesn't seem to get rid of the button for me [21:35:46] try [21:36:11] ah i see there's a magic exception there [21:36:19] try [21:36:37] er [21:36:47] [21:36:49] ahh 14 does it [21:36:52] yay [21:36:54] awesome [21:37:31] i'll just check it on an old phone.. [21:41:33] jerith: ping [21:41:50] jerith: have you created a new vumi deb package [21:42:00] Heya. [21:42:36] calston (our devops guy) has, but we haven't packaged all the rest of the stuff we need. [21:42:59] https://launchpad.net/~praekeltfoundation/+archive/ppa [21:43:02] jerith: what is the timeline for that? [21:43:11] This week. [21:43:22] (That's the plan, at least.) [21:43:27] jerith: okay [21:44:04] jerith: because we will need this for our production hardware [21:44:18] The new vumi uses riak for some stuff, so we have to package the appropriate client libraries. [21:44:30] jerith: okay [21:44:54] (We don't actually need riak for what this app's doing, but the client libraries need to be there for various reasons.) [21:45:38] We also need to see about txredis -- that will probably need to be packaged. [21:46:10] Everything else should still be okay. [21:47:59] calston set up an awesome thingy where we can push successful CI builds to the PPA by clicking a button. [21:48:30] So I need to get the vumi-wikipedia project building like that as well. [21:48:54] The rest of the stuff shouldn't need to change much once we have the initial packages built. [21:52:41] jerith: okay cool [22:08:29] jdlrobson: updated code should be running on testwiki now [22:09:47] \o/ [22:10:14] jdlrobson anything in particular i should look for while testing? [22:10:31] Just check the non-beta hasn't been broken [22:10:42] that's the main risk [22:10:48] ok [22:11:23] looks good to me... [22:11:49] jdlrobson looks like there's now a way to rebiuld l10n cache on testwiki - im running it now [22:11:52] awjr: btw is it possible to fix the wikipedia logo in the bottom left [22:11:58] I have no idea why that got turned off [22:12:11] jdlrobson: wfm [22:12:22] i think, at least - what am i supposed to be seeing? [22:12:28] On test wiki it's fine but on en.wikipedia.org if you look in the bottom left there is Wikipedia ™ text [22:12:36] rather than the wikipedia logo [22:12:57] the config must have changed on live [22:13:03] i know it's been bothering heatherw a lot :) [22:13:06] hmm [22:13:08] let me see [22:13:12] since it's the logo that is ™ not the name [22:13:27] ;) [22:14:26] fuck… camera seems broken on android in phone gap 1.9.0 and 2.0.0 [22:14:36] yay backporting patches [22:16:08] jdlrobson: is it supposed to be http://bits.wikimedia.org/static-1.20wmf7/extensions/MobileFrontend/stylesheets/images/logo-copyright-en.png? [22:16:38] yep [22:16:53] ok yeah, config is just incorrect [22:19:41] or… is that code breaking on our end. wtf? [22:20:05] ? [22:20:51] brion map stuff is a bit borked as well - trying on an incredible and it doesn't work at all with released leaflet. latest master copy works much better but there are some oddities between the new leaflet code and the marker cluster code :( [22:27:08] jdlrobson this is the correct footer logo, yes? http://en.m.wikipedia.org/wiki/Heart_Lake_(Soldier_Mountains) [22:27:22] yup [22:27:25] \o/ [22:27:28] ^ heatherw \o/ [22:27:36] ok i want to test just a bit more and then i'll push the changes out [22:27:43] actually … it seems a bit too big? [22:28:03] I might be wrong though [22:28:09] i'm pretty sleepy right now [22:28:19] w00t, camera isn't broken. it's a bug in our app, i'll track it down later [22:28:22] jdlrobson it looks ok to me [22:28:31] yeh sure it's fine [22:28:35] seems to break when uploading after selecting through the map view (or else a particular item i was testing0 [22:28:42] brion: who broke it this time? [22:28:50] not sure yet :D [22:29:00] can you give me a mtc and i'll try and reproduce on my phones [22:29:19] brion what does cordovalog say? [22:29:38] Error in success callback: Camera4 = TypeError: Cannot read property 'config' of undefined [22:29:42] cordova log says no [22:29:57] ooohh brion I think yuvipanda came across that one [22:30:01] i was seeing that issue when i was working on the description template [22:30:05] aho [22:30:07] it was rather odd but we couldn't reproduce [22:31:11] it was caused by any number of various js errors around the description template code - i thought i cleaned it all up. either i missed something, or someone else introduced a similar problem [22:31:22] mmm [22:31:55] jdlrobson i am satisfied that test is not broken - are you? [22:32:56] I am satisfied [22:32:58] jdlrobson: ok i can reproduce the error thus: [22:33:11] * select the nearby view [22:33:28] * drill down in SE corner of San Francisco until you reach 'South San Francisco Opera House' [22:33:37] * hit the upload button [22:33:42] * select file from gallery [22:34:30] not sure if it's the clustering or if it's specific to the item's data [22:38:51] jdlrobson wide tables are getting clipped on android 2.3.6 in stock browser. is that an old issue? [22:39:01] like, on the main page of testwiki, for instance [22:39:08] an old issue yes [22:39:18] a most frustrating issue that I want to sit down and solve [22:39:21] ok [22:39:24] first I want to solve dynamic sections though [22:39:37] taht's the only real issue im seeing right now [22:39:50] are you satisfied enough to push changes live? [22:41:15] jdlrobson ^ [22:42:01] yup [22:46:32] jdlrobson cool scap is running [22:47:38] sweet [23:26:34] awjr: any ideas what's down on https://bugzilla.wikimedia.org/show_bug.cgi?id=38653 [23:27:20] hmm im not sure jdlrobson, i dont know how any of that works offhand, i'd have to dig [23:27:25] jdlrobson in other news, scap just finished [23:27:32] your changes are now live [23:27:35] yep just seen [23:27:39] can i tweet about it? [23:28:16] jdlrobson let's let the dust settle for a bit in prod before you announce [23:28:27] k it's already to be sent [23:29:05] preilly, [23:29:13] or tfinc, around? [23:30:29] awjr: did you fix sticky cookies btw? [23:30:35] jdlrobson no [23:30:43] jdlrobson: care to peek at https://github.com/wikimedia/WLMMobile/pull/72 ? [23:30:45] it's pretty low on my priority list relative to the other stuff [23:30:55] james forrester just encountered it [23:30:58] he can't get on mobile [23:31:06] tell him to clear his cookies [23:31:09] and try again [23:31:21] if he still doesn't get to hte mobile site, then there's a problem with device detection [23:31:54] he doesn't want to clear his cookies [23:32:15] jdlrobson: he can delete just the one cookie [23:32:25] then tell him if he showers me with gifts and drink, the issue may find its way higher on the priority list. [23:37:57] preilly can we do that call now? [23:38:55] or tfinc ^? [23:39:00] i've got to take off in about 20 mins [23:39:13] awjr: can we do it tomorrow in the office? [23:39:49] ok let's shoot do it right after standup [23:40:09] awjr: okay