[00:14:15] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/32546 [00:14:47] New patchset: awjrichards; "nest table rules 2" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/32547 [00:15:20] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/32547 [00:16:28] grr, trying to update iOS to cordova 2.2 but having troubles. something's not completing loading [00:26:32] MaxSem or awjr: All of a sudden on staging I'm getting this [00:26:52] Welcome to Ubuntu 12.04.1 LTS (GNU/Linux 3.2.0-32-virtual x86_64) System information disabled due to load higher than 1.0 13 packages can be updated. 0 updates are security updates. The last Puppet run was at Tue Nov 20 00:09:59 UTC 2012 (13 minutes ago). Get cloud support with Ubuntu Advantage Cloud Guest http://www.ubuntu.com/business/services/cloud *** /dev/vdb will be checked for errors at next reboot *** *** Sy [00:27:12] Is that a preilly thing? [00:27:21] or do we have access to reboot? [00:28:09] we have access to reboot, but is it necessary? [00:28:14] sudo reboot [00:28:20] but it won't help:0 [00:28:24] :( [00:29:00] bummer so do i just have to wait it out until the load normalizes? [00:29:30] are you unable to do stuff? [00:29:40] I wanted to do another update of code [00:29:44] and run through some more testing [00:29:48] but I can sit it out :) [00:30:05] wait im confused - is it not letting you update code? [00:30:44] the machine is pretty slow these days [00:31:17] MaxSem was right [00:31:18] it was just really slow [00:31:29] oh gotcha - were you waiting for a prompt or something? [00:31:29] the prompt just returned [00:31:32] yupper [00:31:35] I've stopped a background import, might be a bit faster [00:31:45] I started creating new swear words :D [00:31:53] lol [00:31:53] so the time wasn't wasted ;-) [00:32:01] welcome to labs! [00:32:05] hahahah [00:32:15] I know huh, sorry for the false alarm guys [00:32:31] brion [00:32:44] I need to get the boys dinner started [00:32:48] no prob [00:32:52] but I'll get the logs for you in a bit [00:33:02] ping me when you get em :) [00:33:05] you bet [00:33:15] and I can't get to googledocs [00:33:20] :( [00:33:23] with my wikimedia log in for some reason [00:33:33] can you not download the .apk from there? [00:33:42] no sir :( [00:33:49] i'll stick it on dropbox instead, that seems more reliable for sharing files [00:33:57] sweet [00:34:04] I was gonna say dropbox has worked with me and Yuvi [00:35:05] durrrr i set the settings wrong [00:35:08] Jcmish: try it again [00:35:13] k trying [00:35:44] i forgot to actually share the file ;) [00:35:58] i see an anonymous user :D [00:36:11] brion success! [00:36:12] haha [00:36:18] \o/ woohoo [00:36:25] I so know the feeling :D [00:36:32] I'll get it going right after dinner thanks [00:36:57] New review: awjrichards; "Have you run this approach by product yet?" [mediawiki/extensions/MobileFrontend] (master); V: 0 C: 0; - https://gerrit.wikimedia.org/r/32552 [00:37:53] New patchset: awjrichards; "nest specific table rules / throw away strange rules" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/32548 [00:38:07] New patchset: awjrichards; "nest tbody under table" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/32549 [00:38:11] New patchset: awjrichards; "table styling cleanup" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/32551 [00:40:18] New review: awjrichards; "have you run this by product yet?" [mediawiki/extensions/MobileFrontend] (master); V: 0 C: 0; - https://gerrit.wikimedia.org/r/32548 [00:43:24] * tfinc is starting to lose the battle against jet lag [00:43:43] * tfinc goes on a coffee run before meeting with google  [00:54:28] greetings notnarayan [01:05:15] ok im outta here, see you all tomorrow [01:10:17] tfinc: greetings and salutations ! [01:10:41] tfinc: hope you had a good tour of India [01:18:07] getting back to my morning practice session. [02:14:42] the wireframes for the Commons app has been updated https://www.mediawiki.org/wiki/File:Commons_app_wireframe.pdf [02:20:56] New review: Jdlrobson; "Why would I.. this commit is refactoring the css into less syntax and throwing away redundant ones.." [mediawiki/extensions/MobileFrontend] (master); V: 0 C: 0; - https://gerrit.wikimedia.org/r/32548 [02:22:45] New review: Jdlrobson; "Nope.. but in my opinion the current experience is completely broken (it's a bug) and we should expe..." [mediawiki/extensions/MobileFrontend] (master); V: 0 C: 0; - https://gerrit.wikimedia.org/r/32552 [17:14:47] no brion :( [17:16:40] New review: Jdlrobson; "This is pretty important and first got uploaded at the end of October. Please get this deployed it's..." [mediawiki/extensions/MobileFrontend] (master); V: 0 C: 0; - https://gerrit.wikimedia.org/r/30309 [17:17:59] yuvipanda: is alive :) [17:18:11] notnarayan: sortof [17:18:24] yuvipanda: which city ? [17:18:29] aurangabad [17:18:45] ah! did you see the caves yet? [17:19:03] notnarayan: we saw one [17:19:13] notnarayan: rmoen is with me here, along with jdlrobson [17:19:18] notnarayan: we're going to the other ones tomorrow [17:19:22] notnarayan: hi shankur :) [17:19:37] shankar.. my mistake [17:19:44] rmoen: :) [17:19:52] New review: Jdlrobson; "Why is this not merged yet?" [mediawiki/extensions/MobileFrontend] (master); V: 0 C: 0; - https://gerrit.wikimedia.org/r/30811 [17:19:54] Good to know this is you [17:19:57] notnarayan: https://commons.wikimedia.org/wiki/Special:Contributions/Yuvipanda [17:20:43] the names' notnarayan :) http://shankarnarayan.in/notnarayan/ [17:20:59] notnarayan: you don't like the name shankar? [17:21:13] yuvipanda: i dint have a choice :P [17:21:17] heh [17:21:29] this one is uploaded by commons app :) [17:21:47] notnarayan: i didn't have a choice with my name either... damn parents [17:22:02] rmoen: ha ha ha. true! [17:22:34] notnarayan: https://commons.wikimedia.org/wiki/File:Panorama_of_Outskirts_of_Aurangabad_from_Aurangabad_Caves.jpeg panorama! [17:22:41] rmoen: letting a kid pick a name is a worthwhile experiment methinks. [17:22:54] hello Ford_Prefect [17:22:58] o/ yuvipanda [17:23:07] Ford_Prefect: AUTOJOIN THIS [17:23:29] Hmmm [17:23:31] yuvipanda thats one good picture. [17:23:36] notnarayan: some cultures actually do this me thinks [17:23:44] yuvipanda: :) [17:23:51] tfinc, hi tom ;) [17:23:54] Ford_Prefect: https://commons.wikimedia.org/wiki/File:Recording_of_rickety_sleeper_bus_from_Pune_to_Aurangabad.ogg [17:24:00] audio upload from a bus in the middle of nowhere [17:24:02] i've to hear it now [17:24:18] * rmoen takes the bandaid off pointer finger [17:24:37] * tfinc looks around [17:24:40] Ford_Prefect: okay, that's a very boring upload [17:24:41] tfinc: have you adjusted to sf time ? [17:24:42] rmoen: which culture is this, may there tribe increase. [17:24:45] rmoen: yuvipanda jdlrobson : i see your still alive [17:24:48] thats a good update [17:24:52] * awjr waves [17:24:57] Thankfully so [17:24:57] just about [17:24:59] tfinc: aren't you glad we're on IRC? [17:25:06] hello tfinc [17:25:08] jdlrobson, rmoen, yuvipanda where are you guys? [17:25:09] i sure am [17:25:15] how was the taj ? [17:25:19] yuvipanda: but it's from The App, so it's awesome :p [17:25:25] bit dull [17:25:26] Ford_Prefect: true dat! [17:25:30] caves were better [17:25:36] tfinc: not as big as we expected, really. [17:25:37] Not what we expected [17:25:48] thats why i told you guys to make it a day trip at most [17:25:49] people seemed surprise to see white people [17:25:56] there isn't that much more to do there [17:26:04] what is the next stop ? [17:26:16] Agra is next [17:26:16] is this ajanta or elora? [17:26:25] Aurangabad caves [17:26:42] I think we've a bus to Pune? [17:26:48] grr buses [17:26:53] * yuvipanda <3 buses [17:26:56] jdlrobson: juliusz is eager to see you get back [17:27:02] Ford_Prefect: thoughts on video? [17:27:06] awjr: aurangabad [17:27:17] tfinc: he wants to hit me about my code already? [17:27:22] are you not going to ajanta + elora? [17:27:24] he's only been here 1 day..! [17:27:30] Elora tomorrow notnarayan :) [17:27:34] yuvipanda: good question. I need to figure out some transcoding stuff with upstream, but in the mean time, 2 problems [17:27:41] 1. What formats are okay to upload [17:27:44] We are definitely doing elora tomorrow and maybe ajanta if we dont make the train [17:27:56] rmoen: ah, nice. [17:27:59] 2. How we want to get the user to set resolution, etc. [17:28:08] i'm sure he will in due time. he'll be setting up his envs today [17:28:15] i think he'll enjoy having a desk mate [17:28:21] as you guys are sharing a desk [17:28:28] Ford_Prefect: I think for now, (1) will be good? [17:28:49] glad to see hammock still is intact [17:28:49] Ford_Prefect: and just re-encode into the same bitrate and resolution? [17:28:59] we think a dog just died outside [17:29:04] rmoen: ajanta caves are massive, but so is elora. [17:29:12] yuvipanda: Could be nasty with things doing 1080p recording [17:29:14] jdlrobson: get my photo of it? [17:29:26] we ended up putting it right next to the collab space [17:29:27] yuvipanda: also, transcoding cost for high res is high [17:29:28] fits quite nicely [17:29:29] rmoen: ajanta caves still showcase some good paintings. [17:29:32] yep tfinc [17:29:38] Ford_Prefect: well, it's 'experimental' [17:30:05] Ford_Prefect: but we need to figure out (1) and i'm not sure what's supported now, with TMH on [17:30:12] documentation seems super scarce [17:30:13] hello tfinc, you back home? [17:30:21] notnarayan: i am [17:31:02] yuvipanda: well, the sensible options to serve are really h264+aac+mp4 and vp8+vorbis+webm [17:31:11] Ford_Prefect: hehe 'sensible' [17:31:22] tfinc: i was looking at this https://www.mediawiki.org/wiki/Mobile_design [17:31:30] notnarayan: That's great, part of me wants to do both but I think we have a two night rule to abide by. If the gods keep us here, we will be making a day of each [17:31:45] yuvipanda: :) for reasonable quality and browser support [17:32:20] tfinc: is this still the active version? [17:32:23] jdlrobson: pm [17:32:35] Ford_Prefect: https://en.wikipedia.org/wiki/User:RobLa-WMF/Video_testing [17:32:42] is the samples from the new infrastructure [17:32:54] notnarayan: i was never a big fan of that page as it doesn't give you an easy way to identify whats happening right now [17:33:03] it includes some new and some completed items [17:33:27] oh wait .. this is a different page [17:33:46] notnarayan: yeah, it looks up to date [17:33:59] tfinc: yes, i was looking for a page that shows 1) whats happening 2) whats next so i can do something [17:34:57] yuvipanda: hum, all ogv [17:34:57] tfinc: last night in goa, this happened -> http://instagram.com/p/SLpYG0hx-v/ [17:35:24] rmoen: i would have been happier not to see that [17:35:32] rmoen: ouch! [17:35:41] tfinc, sorry pretty brutal [17:35:43] warning next time [17:35:49] deal [17:35:57] rmoen is a walking disaster [17:36:05] tfinc: whats next? that is after commons [17:36:05] Literally [17:36:22] I cut my finger last night unlocking a bathroom door [17:36:42] apparently they don't sand down the edges of metal in this country [17:37:15] rmoen: we don't waste metal in here [17:37:26] Ford_Prefect: https://www.mediawiki.org/wiki/Extension:TimedMediaHandler is the extension page [17:37:35] Ford_Prefect: i'm not sure who's the best person to talk about that to, though. I'll poke around [17:37:44] Ford_Prefect: but yeah, i'm perfectly happy doing just audio [17:37:52] notnarayan: check http://www.mediawiki.org/wiki/Wikimedia_Engineering/2012-13_Goals#Mobile for the annual roadmap [17:37:57] not forgetting rmoen banging his head on the door [17:38:02] in the taj [17:38:10] "taj" [17:38:15] yuvipanda: it does list webm [17:38:46] Ford_Prefect: also, ffmpeg - no gstreamer :P [17:39:23] yuvipanda: for the server-side transcode? Pretty typical. :) [17:39:31] tfinc: https://commons.wikimedia.org/wiki/File:Panorama_of_Outskirts_of_Aurangabad_from_Aurangabad_Caves.jpeg panorama from here. (No blood) [17:41:00] tfinc: any idea if brion is coming in today? [17:41:27] yuvipanda: this beautiful yuvi. i didn't know your phone could go into the past [17:41:29] "6 January 1970 (according to EXIF data)" [17:41:34] just days after the epoch [17:41:41] heh [17:41:54] yuvipanda: he'll be here today just like yesterday [17:41:59] i expect him in .. in the next hour [17:42:06] cool [17:42:22] tfinc: yup, we can travel to the past now since we're no longer limited by PhoneGap [17:42:36] More importantly, you can travel to the future :p [17:43:30] Ford_Prefect: I think anyone who invents a time travelling machine immediately goes to the future, with no way of coming back. Explains a lot [17:44:00] Ford_Prefect: i'm on 3g and can't re-download the gst sdk. Anything else I could od? [17:44:01] *do? [17:44:06] i mailed him the xml file [17:45:27] yuvipanda: hum, I could probably do a build with a newer SDK [17:46:29] Ford_Prefect: sure [17:46:47] Ford_Prefect: also you need to fix your git config. Still newer comits coming with the collabra id :P [17:46:56] o+O [17:47:11] Not that it matters, but that shouldn't happen [17:48:04] yeah [17:48:06] see your last commit [17:48:09] also i merged it to master [17:48:12] on wikimedia/commons [17:48:49] yuvipanda: I'm going to be talking to howie about our commons app discussion [17:48:56] yuvipanda: any updates that you want me to pass on? [17:49:07] margle: [17:49:18] netAction: ? [17:49:19] ups sorry [17:49:24] np. [17:50:08] tfinc: you're going to talk now? [17:50:29] in about 30 min [17:50:32] tfinc: no, nothing after that. That diagram and our discussion covered it [17:50:36] k [17:50:54] tfinc: do put up an etherpad if you can, I'd like to help shape the discussion from the start [17:50:59] i wasn't awake enough to update the wiki page so i'm just going to walk him through our prelim discussion [17:51:20] i'm sure he's going to wonder why we made certain choices [17:51:21] yeah, cool! [17:51:28] but all of it is clearly experimental right now [17:51:34] and were just tossing out ideas [17:51:34] tfinc: I'll be up for a while more if you want details [17:51:45] k [17:52:33] jdlrobson: i'm eager to see if i can get the CR backlog down to zero before we break for the hoildays [17:52:47] that would make me happy [17:52:52] i don't want to spend the first week rebasing [17:52:52] * tfinc ponders opening up the rest of his moving boxes  [17:52:54] ;-) [17:52:57] Ford_Prefect: you know the kaltura guys? [17:53:01] tfinc: can we add maryana and howie to mobile-tech? we keep forgetting to cc them on important emails they should see on mobile-tech [17:53:03] * tfinc dreads opening put he rest of his moving boxes  [17:53:10] yuvipanda: nope [17:53:13] awjr: do it [17:53:17] what's juliuz's username ? [17:53:19] tfinc: how? [17:53:20] (on irc) [17:53:31] granted that does mean that we lose the inter team discussion space [17:53:33] are you guys ok with htat? [17:53:36] awjr: --^ [17:53:39] intra* [17:54:00] Ford_Prefect: hmm, they + goog did most of the work on the vid stuff [17:54:04] tfinc i'll send an email to mobile-tech asking that question :) [17:54:17] gawd i wish i had photo uploads from mobilefrontend - have had so many opportunities this trip! [17:54:23] awjr: your a group owner just like i am [17:54:47] tfinc i probably knew that once [17:54:53] jdlrobson: where are you hosting the photo uploads that you were showing me ? [17:55:10] i'm eager to show that to howie as well [17:55:13] jdlrobson: use the app :P [17:55:31] awjr: start the discussion and see where it goes. you already have all the power to change it [17:55:31] yuvipanda: jdlrobson yeah, use the app :) [17:55:37] tfinc yup [17:55:38] thanks [17:55:45] don't like apps! [17:56:53] awjr: what do you and brion have for juliusz to dive into next week ? [17:57:26] tfinc talk to Jcmish as she's been managing our current bug/qa sprint which will go through next week [17:57:37] Jcmish: --^ [17:58:20] inline styles tfinc ? :P [17:58:25] can I just walk through them with him? [17:59:29] yuvipanda: well, they do webm according to the TMH wiki, and that should be a good enough starting point [18:00:12] Ford_Prefect: iirc it's disabled right now [18:00:28] jdlrobson: you realize we're now scrubbing *some* inline styles in the app? :P [18:00:34] Ford_Prefect: also, how does that affect us? [18:01:15] Ford_Prefect: we're just doing uploads no [18:01:19] yuvipanda: I see that mp3 transcode on-server is deferred until the patents expire, which doesn't bode well for doing h.264/aac transcode [18:01:43] yuvipanda: true, we could just as well be doing theora + vorbis + ogg, easy enough to switch over [18:01:46] Ford_Prefect: yeah, is that like, 'forever'? [18:02:14] Don't know. For non-commercial use, h.264 is available till 2015 iirc [18:02:23] available royalty-free, that is [18:02:45] yuvipanda: MaxSem|away won't let me scrub inline styles :-( .. at least not in the php [18:02:48] Ford_Prefect: I think it's still too politicized. I think brion would have a better idea when he's back [18:03:03] jdlrobson: well, I just add an extra class to the ones I wanna scrub and !important the fuck outta them :P [18:03:33] okay, pseudo-scrubbing [18:03:43] $('[style]').removeAttr('style'); [18:03:54] yuvipanda: UX wise, doing unaccelerated video encode on mobile is suicide, imo [18:04:25] It's doable as a toy, and we should do some quick experiments to quantify how bad it is before dismissing it [18:04:47] Ford_Prefect: true, audio was much faster than we thought, no? [18:04:54] It was indeed [18:05:26] Ford_Prefect: or maybe it was just that we were using (literally) top of the line hardware [18:05:40] It was still faster than expected [18:05:48] Hang on, let me get you that debug apk first [18:06:04] And then I'll modify it to do a video transcode [18:07:50] Ford_Prefect: ok [18:16:00] tfinc? [18:18:53] jcmish1: yes? [18:18:59] howdy [18:19:24] so Juliansz… I'll just walk through what we have on the board and see if they are things he can tackle? [18:19:29] does that sound good? [18:20:30] jcmish1: you should walk through the backlog with brion and see whats appropriate for juliusz to handle [18:20:33] brion is on boarding him [18:20:42] sounds good [18:20:52] that way brion knows what he'll have to support [18:21:13] speaking of the brion [18:21:26] heyyyy [18:21:33] brion: are you not coming into the office today? [18:21:49] do we have a jgonera? [18:22:02] brion: he's sitting right across from me [18:22:05] figured i'd be in in a bit [18:22:08] excellent. get him on irc :) [18:23:35] Maryana: desk is mostly clean now [18:23:49] heh. i will wheel over [18:25:40] also heads up tfinc Maryana awjr http://www.mediawiki.org/w/index.php?title=Extension_talk:MobileFrontend&offset=20121117181255&lqt_mustshow=20527#Howto_switch_to_mobile_Version_20527 / https://bugzilla.wikimedia.org/show_bug.cgi?id=36894 < would love to get this prioritized and fixed [18:26:21] * jdlrobson waves at jgonera  [18:27:57] i can't figure out if sj is requesting a visible article URL or an &action=edit URL [18:28:12] jdlrobson: yeah that'll be nice to fix. it shouldn't be that hard…. [18:28:25] yuvipanda: new SDK changes one of the gst class names :/ [18:28:38] Caught me by surprise, but oh well [18:28:47] yuvipanda: http://arunraghavan.net/temp/wmf-commons.git-debug.apk [18:30:00] Ford_Prefect: heh [18:30:02] hey Brooke [18:30:03] yuvipanda: hey did you get the contact information of the guy who was working on the objc mwapi library? [18:30:03] err [18:30:04] sorry [18:30:06] brion: [18:30:27] brion: he forked it off of yours [18:30:37] brion: I also got contacted by two folks who want to do the same thing today [18:30:52] ah good should be able to find him them [18:30:54] heh spiffy [18:33:17] yuvipanda: that thing won't crash for you [18:34:55] Ford_Prefect: when doing what? [18:34:59] Ford_Prefect: wav? amr? [18:35:47] yuvipanda: if you recall, it crashed on startup [18:35:52] ooh, virtual box update to install [18:36:09] yuvipanda: while trying to initialised the Android MediaCodec wrapper plugin [18:37:26] yuvipanda: redownload and install the apk please (included the plugin in the build this time) [18:37:39] Ford_Prefect: slloooow connection [18:37:43] redoing it [18:37:49] Sorry, my bad for forgetting the first time [18:39:06] yuvipanda: going to crash now, though [18:39:18] Video encoding will need to wait [18:40:22] heh [18:40:24] good night [18:43:22] o/ [18:57:58] * ori-l 's here, but semi-afk for a few [19:22:08] meh [19:23:12] MaxSem: is anybody working on JS issues right now? [19:23:43] what JS issues? *looks in email* [19:24:38] MaxSem: I was trying to ask is there anybody available to look at a JS bug [19:25:21] I can try [19:25:54] MaxSem: The banner isn't showing on http://en.zero.wikipedia.org/wiki/Special:ZeroRatedMobileAccess?useformat=mobile [19:25:57] oh, but we have jgonera now!:) [19:26:50] MaxSem: you will need to do something like: RequestHeader set X-CARRIER "DTAC" in Firefox to see a banner [19:26:59] MaxSem: but it isn't showing on the landing page [19:28:59] preilly, what FF extension do I need? [19:29:28] Here is the text: ​…​Free preilly Access by DTAC Thailand [19:29:40] MaxSem: Modify Headers extension for FF [19:31:09] jcmish1, how are we looking WRT deployment? [19:31:55] I think we're looking good [19:32:15] cool [19:32:30] MaxSem: https://gerrit.wikimedia.org/r/#/c/32864/ [19:32:37] I just finished looking at 251 [19:32:54] what time we starting deployment today MaxSem? [19:33:04] 1pm PST [19:33:13] k cool thanks [19:33:23] ori-l, I'll deploy it with other changes then [19:33:36] I haven't forgotten about it:) [19:35:39] cool [19:36:59] preilly: what is the problem exactly? [19:37:36] ori-l: we don't see the carrier banner at http://en.zero.wikipedia.org/wiki/Special:ZeroRatedMobileAccess?useformat=mobile [19:37:57] ori-l: something to do with jdlrobson's mw-mf-banner refactor [19:38:26] ori-l: for example in the console if you do a $('.mw-mf-banner').show(); you see it [19:41:14] ori-l: also note that it doesn't render in the correct position [19:43:15] unfortunately, Zero has very low visibility for all of us [19:46:19] MaxSem: well that will need to be fixed or at least addressed moving forward [19:46:39] MaxSem: basically this issue occurred with Jon's mw-mf-banner refactor [19:46:42] preilly, can you repro it locally? [19:46:49] MaxSem: yes [19:46:59] bisect? [19:47:36] MaxSem: Divide into two parts?!? [19:47:45] git bisect:) [19:47:56] MaxSem: делить пополам? [19:48:08] на ноль [19:48:17] it has an inline style setting it to display:hidden, but the js seems to presume it is visible by default: [19:48:22] MaxSem: you want me to search? [19:48:30] if ( zeroRatedBannerVisibility === 'off' ) { banner.style.display = 'none'; } [19:48:48] and zero ha ha [19:52:18] WTF, FireBug doesn't step inside of $( '.mw-mf-banner' ).each( function() { for me [19:52:34] even though $( '.mw-mf-banner' ) is not empty [19:52:38] "you go ahead, i'll stay right here" [19:53:07] * MaxSem tries Dragonfly [19:53:18] ori-l: notice how it works on any other page [19:53:34] ori-l: e.g., http://en.zero.wikipedia.org/wiki/Windsor_Castle?zeropartner=1002&renderZeroRatedBanner=true [19:54:10] ori-l: this code is in gerrit.wikimedia.org:29418/mediawiki/extensions/ZeroRatedMobileAccess.git and gerrit.wikimedia.org:29418/mediawiki/extensions/MobileFrontend.git [19:54:59] haha, Dragonfly rules [19:55:02] preilly: the HTML output (pre-js) doesn't have display:none for the inline style in the windsor castle example [19:55:12] so it's probably in php somewhere, let me check out the repo [19:55:34] ori-l: includes/MobileFrontend.body.php:88: $zeroRatedBanner = $this->zeroRatedBanner ? str_replace( 'display:none;', '', $this->zeroRatedBanner ) : ''; [19:56:17] dismissNotification = banner.getElementsByTagName( 'button' )[ 0 ]; [19:56:29] and it's undefined [19:56:35] preilly: that's a terrrible, terrible line of code [19:56:59] awjr_lunch: i'm going to be 5-10 min later as i need to pickup lunhc [19:57:01] ori-l, welcome to the real world!:P [19:57:16] almost as bad as the stuff i write [19:58:20] tfinc no worries [19:58:25] tfinc just lmk when you're back [20:03:34] awjr: nom nom nom [20:04:01] preilly: Ori.livneh [20:04:35] ori-l: added [20:04:38] tfinc om nom! [20:04:43] preilly: danke [20:14:09] MaxSem: it's all special pages, btw [20:14:18] ? [20:16:42] that the banner is missing from [20:16:57] the bug that preilly is talking about [20:18:14] ori-l: is createDismissableBanner not being called on special pages? [20:20:00] i don't see any calls to createDismissableBanner [20:23:02] hmm [20:36:05] ori-l: are you still looking at the issue? [20:36:45] sorta. 208.80.153.250 (staging) isn't fully configured, so i still don't have an instance to monkey with that is running the mobile setup [20:43:00] is it holding up deployment? is anyone else looking at it? not sure what the current status is [20:44:11] ori-l: it is broken for all the zero users around the world [20:44:21] ori-l: and is holding up our launch in India [20:44:54] ori-l: I was hoping that MaxSem was looking at it too [20:45:17] what do i need to do to get staging to work? [20:45:24] assign it a hostname in /etc/hosts perhaps? [20:45:32] it's not holding anything because it's already broken in production [20:45:43] ori-l: staging.wmflabs.org [20:46:07] oh, d'oh. i forgot the hostname isn't displayed in the instance list [20:46:48] preilly, if you can repro it locally you can git bisect for it [20:47:07] I'm preparing for deployment [20:48:03] http://staging.wmflabs.org/wiki/Disruption?zeropartner=1002&renderZeroRatedBanner=true <- not showing the banner, so i presume there's some salient staging/prod configuration difference. [20:49:00] ori-l: try http://staging.wmflabs.org/wiki/Disruption?zeropartner=1002&renderZeroRatedBanner=true now [20:49:22] ori-l: I added: RequestHeader set X-CARRIER "wikimedia" [20:49:23] RequestHeader set X-DEVICE webkit [20:49:23] RequestHeader set X-SUBDOMAIN "ZERO" to .htaccess file [20:49:54] still nothing [20:50:02] ori-l: clear your cache and try again [20:50:06] ori-l: it works for me [20:50:20] ori-l: feel free to come look at my screen [20:50:41] nm, works [20:50:49] ori-l: okay, cool [20:51:39] jgonera: also install php5-mysql in case that's not automatic [20:57:47] awjr, brion, jcmish1, jgonera, Maryana, preilly, ori-l: we're deploying in 3 minutes, is everyone ready? [20:58:02] yep :) [20:58:04] \o/ [20:58:15] I suppose ;) [20:58:22] i'm ready [20:58:52] ori-l, do we need anything in addition to https://gerrit.wikimedia.org/r/#/c/32864/ ? [20:59:42] no, afaik. iirc, jon has a makefile that pulls the script from our repo. did you run that? [21:00:48] it's in Git already, dunno how recent [21:01:02] same as latest, i just checked [21:01:14] so no, nothing else [21:01:20] cool [21:03:21] ready! [21:07:53] Jenkins problems, poked Hashar [21:12:18] MaxSem: what's the prob? [21:12:46] jenkins-bot failng builds [21:13:09] MaxSem due to failing tests or for some random reason? [21:13:32] oh i see the failed build [21:13:39] Krinkle unleashed jshint on MW [21:13:50] doh [21:13:53] yeah [21:13:55] ick [21:13:59] which is fine! [21:14:03] Hey all. [21:14:06] but old wmf branches dont' pass yet [21:14:10] fixed [21:14:13] except for the fact that you broke our build :p [21:14:25] preilly: I hear you sorted out the vumi redis sadness? [21:14:36] jerith: I think so [21:14:44] \o/ [21:15:18] thanks for fixing it Krinkle :) [21:15:52] awjr: when is our next iteration kick off ? [21:16:04] awjr / jcmish1 --^ [21:16:08] tfinc the kickoff meeting is dec 3 [21:16:14] k [21:17:09] oh do we have jgonera added to the invite yet? [21:21:00] okay, we're live on http://test.m.wikipedia.org/ please test [21:21:33] jcmish1: i don't think so; i'll see to it [21:22:05] i can do it :D [21:22:40] jcmish1: i think the calendar event creator has to do it [21:22:41] ori-l, TypeError: Cannot convert 'mw.loader' to object mw.loader.implement("ext.Experiments.eventlog",function(){( [21:22:50] ah true :) [21:23:08] MaxSem: looking [21:23:42] heh, that module uses mw.loader in a non-Rl context [21:25:09] it's supposed to be loaded in an RL context [21:25:47] no mobile RL so far:] [21:25:47] gahh my mouse is spazzing out [21:25:58] did jon's image upload and table changes get merged? [21:26:06] no [21:26:36] ah, k [21:26:40] all is well with the deployment? [21:26:41] MaxSem: it doesn't have any dependencies, so you can just take out the RL module stuff. want me to submit a patch? [21:26:49] my computer keeps going to sleep so i miss the backlog ;_; [21:27:12] ori-l, would be great! [21:27:18] MaxSem: i see what happened -- jon updated the url in the makefile to grab the file from bits [21:27:24] which adds the wrapping code [21:29:53] New patchset: Ori.livneh; "Use non-RL-wrapped eventlog.js" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/34424 [21:30:38] Change merged: MaxSem; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/34424 [21:31:03] MaxSem, jcmish1: is the list of changes for today's deployment up somewhere/ [21:31:24] i see the contact page change. not sure what else there is... [21:31:42] awjr I spaced while testing [21:31:45] and didn't even look :( [21:32:18] jgonera: https://mingle.corp.wikimedia.org/ [21:32:22] it looks like watchlist viewing is in the ready for signoff column in mingle, but i don't see it. do you have to access it by going to special:watchlist? [21:32:26] ^ brion? [21:32:34] ok, aside from the usual, what should we be looking at for testing jcmish1/MaxSem? [21:33:31] awjr digging getting info together now [21:34:47] MaxSem: works; events getting logged. [21:35:06] \o/ [21:35:08] yeah, was going to report that I've deployed it:) [21:35:31] 150: Watchlist: logged-in viewing [21:35:31] 212: New contact email [21:35:31] 251: Suppressing general/article contact [21:35:57] I've clicked on a few sections, are they logged? [21:37:39] awjr: can you help us get jgonera set up on mingle? we think we can add him to the mobile group but can't actually create his account ;) [21:37:47] just chatted w/brion about watchlist - it looks like it didn't get reviewed, so it's not getting deployed today [21:37:48] brion sure thing one sec [21:37:53] thx! [21:38:14] Sorry Maryana [21:38:38] no worries. just letting you know not to look for it :) [21:38:45] hehe [21:39:02] yeah I did the on my machine thing but yup on to regression tests [21:39:08] I'm working on non beta [21:40:16] brion: done and i added him to the mobile project, see pm for creds [21:41:13] he's in, thanks [21:41:17] w00t [21:45:28] how do we look? [21:45:45] ori-l, does logging work? [21:45:53] MaxSem: yes! :D [21:46:37] the way random works in current beta causes some big problems. im not sure when the changes ewre introdruced that are causing the issues but im prettys ure sometime after the last deployment [21:46:50] http://kubo.wmflabs.org/mobile.html [21:47:18] k just finished non beta testing [21:47:22] and ever thing looks good [21:47:28] switching to beta [21:47:42] awjr: what's it doing now? we had some issues with the last (or next-last?) deployment when jon introduced some new goodies, not sure if it's changed recently [21:47:56] for instance if you navigate to a random page frmo the main page as a logged in user, then click the start to add the page to your watchlist, the main page gets added to your watchlist rather than the page you're looking at. a similar thing is true for the prepopulation of the email body for contact (the referring page will be 'main page' rather than the current page you're looking at) [21:48:08] d'oh [21:48:09] (you have to stay on the page to see events as they come in) [21:48:18] yeah sounds related to the way it's dynamically loading things [21:48:22] must not be updating some params [21:48:23] aye [21:48:31] but i dont think i noticed that issue when those changes were first introduced [21:49:55] jcmish1: how can we better catch issues like that ^^ [21:50:09] yay logging! [21:50:42] although i suspect logging may suffer the same problems in regards to what happens when you're viewing a 'random' page [21:50:57] awjr I need to add that to the regression tests [21:51:03] and that's a good candidate for automation [21:51:13] totally [21:51:25] I'm making a note and adding it [21:51:45] awjr: so the 'page' datapoint may be inaccurate is what you're saying [21:51:54] ori-l: correct [21:54:31] awjr: are you deeply familiar with how RandomPage works in general? not necessarily apropos of mobile [21:54:44] chrismcmahon: not deeply [21:54:48] random just pulls in the page and manipulates the dom to insert the content [21:54:52] it doesn't update mwMobileFrontendConfig.settings.title [21:55:12] yeah, i wish we had more data on the use of the random feature on mobile - i'm betting it's a power law curve, with a tiny minority using it extremely often and the vast majority completely unaware that this feature exists [21:55:17] chrismcmahon: but the way it's happening in MF beta is oh, as ori-l said ^ [21:55:47] ok we've got jgonera pretty much set up, he needs something in MobileFrontend to start hacking on :) any suggestions? [21:55:47] dynamically loading a random page into the dom instead of actually loading a fully rendered random article [21:55:47] ah, no I have a different weird issue than that I think. [21:57:50] chrismcmahon: what are you seeing? [21:58:03] New patchset: Ori.livneh; "(untested) Update title on random page nav" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/34428 [21:58:19] ^^ i think this does the trick, but can't test atm [21:58:47] awjr: I can't prove it exactly, but we have some automated browser tests that hit RandomPage, and for some reason IE10 lands on protected pages far, far more often than other browsers. [21:58:58] that is weird [21:59:10] :( [21:59:17] brion, jgonera how about https://bugzilla.wikimedia.org/show_bug.cgi?id=41514 ? [21:59:44] awjr: it is weird, and I'm not sure I have enough data to call it for sure, but it seems that way [22:00:09] chrismcmahon: im sure you can thank the MediaWiki gremlins :) [22:00:37] MaxSem, jgonera: that's probably interacting weird with some of jon's history stuff and the dynamic sections… should be fun to investigate at least :D [22:01:29] awjr, Maryana when I disable images they aren't disabled on the article [22:01:38] but the main page has them disabled [22:01:47] is it b/c the image in in a navbox? [22:01:50] this is in beta [22:01:52] jcmish1: in beta? also you may need to force a refresh/clear cache [22:01:52] those don't get disabled, iirc [22:02:03] *is in [22:02:23] ah tis possible [22:02:40] but I swear these were disabled before let bit a few other articles to test [22:03:20] ori-l that changes looks good (and i have confirmed that page name will currently show up wrong in logging) [22:03:30] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/34428 [22:03:33] awjr, is there an issue for that on Mingle or Bugzilla? [22:03:54] jgonera: for what specifically? [22:03:59] (I mean "dynamically loading a random page into the dom") [22:04:05] awjr: it _should_ fix the watchlist / random issue [22:04:06] jcmish1: seems to be working for me [22:04:21] ori-l it appears to locally for me [22:04:32] jgonera: no, that was something jon implemented on his own [22:04:32] Maryana yup you were correct [22:04:45] I picked a different article to use on my ipad [22:05:01] yeah, that's probably confusing users who try to switch off images [22:05:11] it confused a guy in a user test i ran [22:05:29] and it was in a navbox but my old tried and true article worked like a charm ;-) [22:05:52] updated once again [22:06:19] awjr, I'm not sure then if I understand the difference between "random page" and "fully rendered random article" :) Is there any place I can read more about that? [22:06:30] jgonera: one sec [22:10:09] Maryana, awjr: outside of the random issues not adding the right article to the watchlist the regression tests in beta look good [22:10:49] jgonera: in the non-beta version of MobileFrontend, if you select 'random' from the nav to load a random article, it will cause the whole page to reload with a randomly selected article. in the beta version, it functions a little differently. an ajax request gets made to find a random article title, and a dialog gets presented to the user basically saying "read this article or try again?" if you select 'yes' (read this article), then another aj [22:10:50] yeah, all looks fine to me. thanks for flagging the watchlist/random issue [22:11:13] this issue should already be fixed [22:11:32] ha! cool was that in your update above? [22:11:38] I'll retest if so MaxSem [22:14:37] MaxSem: im not seeing the updated code on fenari [22:15:56] awjrichards@fenari:/home/wikipedia/common/php-1.21wmf4/extensions/MobileFrontend$ git log | head -n 1 [22:15:56] commit ad60870bfc057333bb7001fa557b4edf787c7eb2 [22:16:26] i think the updated commit should be 4ee00c9e1ae822de09a2f2aec282d187008c08fa? [22:17:39] awjr, it is now [22:17:47] w00t [22:18:24] yah, the fix looks good on test [22:18:44] so, can I push? [22:18:45] MaxSem, awjr [22:18:48] looks good [22:19:10] awjr I think I just figured out the test case for the sticky cookie problem [22:19:31] that Jon reported… I need to try to repro a few times then I'll get it written up [22:22:02] ori-l thanks again for your fix :) want to join the mobile team? [22:22:27] jcmish1: that makes me simultaneously happy and sad [22:22:43] awjr hehe yes that's the way of it :D [22:22:44] awjr: np, i'm looking at the hash problem. [22:22:53] OuKB: i think so - jcmish1, Maryana - is it ok for max to push? [22:23:11] ok by me [22:23:55] * tfinc_ really likes having a whiteboard right next to him to sketch on  [22:24:17] yup me too [22:24:22] I think it looks good [22:25:24] OuKB: go for it [22:29:51] scapping... [22:37:50] brion: is https://gerrit.wikimedia.org/r/#/c/29621/4 the latest watchlist stuff awaiting review? [22:38:31] awjr: yep that's the last version since jon asked me to update it so it can be reviewed [22:38:38] brion cool thanks :) [22:40:48] jcmish1: when you have the chance can you please add the changelog to the deployment page for today's deployment? [22:41:02] yup will do [22:45:20] tfinc: ok I think i fixed the Facebook sharing on iOS 6 issue in cordova22-ios-update branch. [22:45:27] anybody want to review and merge? :) [22:45:52] https://github.com/wikimedia/WikipediaMobile/pull/314 [22:49:04] New patchset: preilly; "fix zero banner" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/34440 [22:49:26] Change merged: preilly; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/34440 [22:49:39] MaxSem: can you please push this change: https://gerrit.wikimedia.org/r/#/c/34440/ [22:51:01] awjr [22:51:07] hi [22:51:20] howdy I forgot how do I figure out which to and from changes i want? [22:51:33] for the changelog [22:51:35] ah [22:51:53] git log —no-merges .. [22:52:01] no i have that [22:52:04] oh [22:52:08] ? [22:52:17] I remember the command but what is the starting sha? [22:52:25] that is a good question [22:52:37] is it the last one time we did a release? since we skipped a few I'm a bit lost [22:52:38] i guess you can look at the merge history for the deployment branch [22:52:43] not necessarily [22:52:54] kayo [22:52:55] it's possible someone else deployed some MF changes since our last deployment [22:52:56] preilly, already scapping - will make a separate push once it's over [22:53:08] MaxSem: okay, thanks [22:53:26] MaxSem: It's very important that it goes out today if possible [22:53:27] jcmish1: you could do a query in gerrit like: owner:MaxSem project:mediawiki/core [22:53:44] ah okay cool [22:53:46] and see what he merged today for deployment [22:54:00] this looks like the first one he did today: https://gerrit.wikimedia.org/r/#/c/34410/ [22:54:21] if you look at the diff for that changeset, you can see the original sha1 from before his commit [22:54:25] -Subproject commit 263ca9bfc922621a835ee1611b6d479d146d6b26 [22:54:50] alrighty cool [22:54:53] there are other ways to do it but that's what i've done before [22:55:07] hehe I started just guessing and thought you know I better ask :D [22:55:26] :) [23:02:40] New review: awjrichards; "I haven't tested this yet; code generally looks OK (at a really quick readh-through) with some minor..." [mediawiki/extensions/MobileFrontend] (master); V: 0 C: 0; - https://gerrit.wikimedia.org/r/29621 [23:05:05] MaxSem: can you please let me know when you push that change [23:05:25] okay, we're getting events from prod [23:05:36] ori-l: sweet [23:05:51] sure [23:05:59] {"info": "17", "origin": "cp3020.esams.wikimedia.org", "event_id": "mobile", "seq_id": 458, "host": "en.m.wikipedia.org", "beta": true, "session": "1353452726505", "delta": 2111, "data": "TogglingReady", "page": "Soviet Armed Forces"} [23:06:03] max, is that you? :D [23:06:21] heh [23:06:27] no [23:06:39] ori-l: we are logging session identifiers? [23:06:48] tfinc are we meeting? [23:07:16] preilly: no [23:07:26] session is timestamp [23:07:52] ori-l, I've left a couple of entries [23:08:01] ori-l: okay cool [23:12:52] MaxSem, I checked https://bugzilla.wikimedia.org/show_bug.cgi?id=41514 and it seems to work on Chrome and Firefox. does it affect any particular platform? [23:14:17] jgonera, at a quick glance, it doesn't work in Opera. which mobile devices have you tried? [23:15:02] MaxSem, I didn't try any mobile devices, just the desktop, will do that [23:17:54] i can actually repro those bugs in Firefox [23:18:03] might be a beta-only thing? [23:18:39] yeah if i turn off beta mode, http://en.m.wikipedia.org/wiki/California#Population jumps to the section as expected [23:18:57] jgonera: go into the mobile site's settings menu and switch on 'beta', see if that helps [23:21:55] brion, MaxSem , yep, it's beta only, thanks [23:22:05] whee [23:23:20] off to starbucks, brb. espresso machine is on the fritz today :P [23:23:26] awjr: we lost you [23:23:35] gahh [23:38:39] back [23:46:31] how do I turn on the debug mode? I'd like the JavaScript not to be uglified/minified [23:48:23] jgonera: add ?debug=true on the URL, that… should usually work :D [23:48:52] brion, yeah, but can I tell mediawiki to just add it to all script tags by default? [23:49:00] agh that would be handy [23:49:04] there…. is a config switch lemme find it [23:49:39] jgonera: http://www.mediawiki.org/wiki/Manual:$wgResourceLoaderDebug [23:53:11] brion, that worked, thanks [23:53:21] yay [23:56:20] preilly, your fix is live on testwiki - do you want to test it or should I just push? [23:56:44] MaxSem: let me take a look [23:58:32] MaxSem: I think it is safe to push [23:58:39] ok