[00:00:23] New patchset: Jdlrobson; "Avoid api request to lookup watchlist contents state (bug 44878)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48580 [00:00:24] New patchset: Jdlrobson; "Prevent looking up whether an article can be watched on special page" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48581 [00:02:02] New patchset: JGonera; "Add progress bar to photo upload progress notification (#379)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48050 [00:02:18] [Commons-iOS] brion pushed 1 new commit to master: http://git.io/gtxs0g [00:02:18] Commons-iOS/master 54ed82b Brion Vibber: basic pull to refresh, drop refresh button [00:03:32] New patchset: JGonera; "Run jshint on the whole javascripts directory" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48582 [00:03:40] Change merged: Jdlrobson; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48050 [00:04:46] jdlrobson: https://meta.wikimedia.org/wiki/Schema:MobileBetaWatchlist [00:07:37] New patchset: MaxSem; "Bug 36894: Mobile device detection for third parties" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48583 [00:07:38] my magnum opus for today^^^:) [00:07:45] New patchset: awjrichards; "(mingle 330) Styles upload count on dashboard" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48584 [00:10:20] YuviPanda|away: where was the reference page for the event tracking stuff? [00:10:28] jdlrobson: pm [00:11:28] i see https://www.mediawiki.org/wiki/Commons_App/Analytics but it's…. not very detailed [00:12:33] it's got a reference to https://meta.wikimedia.org/wiki/Schema:MobileUploads …. which doesn't seem to cover everything mentioned there [00:18:34] New patchset: Jdlrobson; "update event logging for watchlist star to use new schema" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48586 [00:19:23] Change merged: Jdlrobson; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48582 [00:45:08] awjr i'm out of my mtg. [00:45:13] how can i help? [00:45:26] hey munaf, just wanted to dbl check about the design asset for https://mingle.corp.wikimedia.org/projects/mobile/cards/330 [00:45:30] lemme send you a screenshot of what i did [00:46:57] munaf: just sent the screenshot [00:47:00] cool [00:47:39] question awjr [00:47:46] yah [00:47:48] in implementation notes it says "bolded sans serif" [00:47:57] yeah i wondered about that [00:48:04] i think that may have been a typo [00:48:20] it seemed totaly incongruous with the image, so i wasn't sure [00:48:23] this looks fine to me, and i believe it is what vibha intended [00:48:28] i'm guessing that was just a typo [00:48:31] i just did the easier of the two :) [00:48:32] ok cool [00:48:39] however i may speak to her about changing the design [00:48:40] i mean, easy enough to change if we need to [00:49:13] yup, but it works for me as-is :-) [00:49:17] cool thanks! [00:53:39] New patchset: Jdlrobson; "Avoid api request to lookup watchlist contents state (bug 44878)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48580 [00:53:39] thanks for your blog post edits, awjr! [00:53:49] np Maryana, looks good :) [00:54:28] Change merged: JGonera; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48580 [00:55:29] New patchset: Jdlrobson; "Prevent looking up whether an article can be watched on special page" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48581 [00:55:58] Change merged: Jdlrobson; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48581 [01:42:47] ok awjr finally drafted that mail [01:42:55] sent [01:43:01] that was painful [02:23:32] Maryana: http://english.stackexchange.com/questions/67720/how-did-the-pronunciation-of-the-word-derby-evolve [02:25:52] Maryana: http://en.wikipedia.org/wiki/Ca%C3%B1o_Negro_Wildlife_Refuge [04:12:52] tfinc: I archived and have a wip on the merge [04:12:59] YuviPanda|away: thanks [04:13:20] tfinc: I'll respond on that email thread once done [04:13:23] k [07:32:30] * Debloper needs a list of keyboards which are (or needs to be) in "beta status" (at least a couple) - wonders if there's any list for it. [07:32:33] https://github.com/wikimedia/jquery.ime/issues/14 [11:04:43] j #mediawiki-i18n [13:04:32] [java-morelangs] yuvipanda pushed 2 new commits to master: http://git.io/eJ5nQA [13:04:32] java-morelangs/master a8ac866 YuviPanda: Generate string labels for scripts too [13:04:32] java-morelangs/master 1e7c1c3 YuviPanda: Updated to latest jQuery.IME [16:28:43] New review: MaxSem; "Patch Set 3: Code-Review-1" [mediawiki/extensions/MobileFrontend] (master) C: -1; - https://gerrit.wikimedia.org/r/48164 [16:42:04] New review: MaxSem; "Patch Set 1: Code-Review-1" [mediawiki/extensions/MobileFrontend] (master) C: -1; - https://gerrit.wikimedia.org/r/48598 [17:08:01] * awjr waves [17:09:05] whee [17:12:10] hello brion [17:12:15] heyy [17:12:26] brion: https://www.mediawiki.org/wiki/Apps/Commons#Reports [17:12:52] ooh [17:13:12] for some reason the css is failing on wiki for me :D [17:13:15] it feels very retro [17:13:43] i started on an EventLogging implementation for obj-c but got distracted redoing all my async stuff using deferreds [17:13:53] brion: me too :P [17:13:57] hehe [17:14:02] brion: ooo objc has deferreds? [17:14:12] easy enough to make them now that it has blocks [17:14:29] ah [17:14:30] right [17:14:31] i'm just hoping i'm not fucking up memory management ;) [17:14:33] * YuviPanda curses Java [17:14:42] you can have anonymous inner classes ;) [17:14:53] they suck [17:15:06] haha [17:15:09] verbose [17:15:16] aoh well [17:15:23] brion: but on the plus side, background tasks :P [17:15:27] \o/ [17:16:08] so far i'm just working on the main thread using the event loop, any threading is hidden in the background for me :) [17:16:18] eventually i might have to move some encoding/resizing stuff to a thread tho [17:16:29] oh. async network with callbacks? [17:16:36] yeah [17:16:37] I think I've a *lot* of threads running around [17:16:41] hehe [17:17:20] brion: I'm sure that they'll bite me sooon [17:17:34] brion: did you see that section on the page? [17:17:45] YuviPanda: so for the event logging are we using the same schema as the mobile web uploads? or are we going to have our own [17:17:55] brion: we're going to have our own I think [17:17:59] excellent [17:18:00] i've been thinking of the schema [17:18:07] and thinking of how separate / together they should be [17:19:38] we have a lot of distinct stuff, like the different entry workflows [17:19:46] probably makes sense to have our own [17:19:57] then it's up to the data guys if they want to put them together :) [17:20:18] brion: yeah, about that. I dunno how much 'data guys' time we can get :P [17:20:23] heh [17:20:59] brion: also me splitting up the reports by things we can pick up from categories [17:21:01] vs eventlogging [17:21:08] *nod* [17:21:28] most anything on successful uploads can be got from the actual pages yeah [17:21:35] it's mostly failures and behavior leading up to them that we want to do with event logging i guess [17:22:04] brion: yeah [17:23:00] brion: it'll also be interesting to see how to do 'joins' across data from two types [17:23:08] like, 'number of failed uploads per user per successful upload' [17:23:10] or soemthing like that [17:23:24] heh [17:25:20] awjr: what say you on these last commits [17:25:34] jcmish: one sec [17:25:38] k no prob [17:25:39] of Jon's? [17:29:59] notnarayan: hello? [17:30:07] brion: saw pau's response? [17:30:59] haven't read it yet [17:31:23] jcmish: sorry, there's an ops meltdown and there was some speculation it's mobile, but i think they're beyond that now [17:31:39] oh no! [17:31:43] was it? [17:32:26] bits cache sending 503s, i think current theory is something related to wikidata deploy yesterda [17:32:26] or today, ratehr [17:32:33] ah well we haven't deployed anything [17:33:47] fun morning already [17:34:01] YuviPanda: hey [17:34:09] although mobile requests to bits are still being pointed at as causing the overload [17:34:13] argh [17:34:39] rather, they are not helping [17:35:02] mark made one of the varnish caches serve 503s to help bring the load down, so we may hear some complaining [17:35:36] and testwiki has been disabled in varnish config [17:36:47] awjr, any thoughts on https://docs.google.com/a/wikimedia.org/document/d/13WASDu7PJKVY_DDfz4z9gir8_nIAAInUi-az4qBm4NA/edit ? [17:37:18] ah yeah, MaxSem, probably - i will look at it later [17:41:25] ok, the apache load still hasn't sufficiently dropped, mark is pushing a change that will cause all requests that have a referrer containing 'modules=mobile.startup' to 503 [17:41:37] there will be mobile site breakage [17:41:51] (https://gerrit.wikimedia.org/r/#/c/48645/1/templates/varnish/bits.inc.vcl.erb) [17:43:30] awjr, are you just lurking in #-ops or you have another source of information? [17:43:38] MaxSem: security channel [17:44:33] o_0 [18:01:02] MaxSem: there? [18:01:13] yup [18:02:21] awjr: there too? [18:02:26] apparently a problem with mobile right now [18:02:43] jdlrobson: yo, yeah there's an outtage [18:02:46] jdlrobson, it's a known meltdown [18:03:00] the ops sacrificed us [18:03:13] preilly is asking about doing mobile deployment earlier [18:04:14] why? [18:04:39] jdlrobson: yeah we're chatting - i was reporting about the outtage in here a while ago but havent emailed yet - i'll send a note [18:05:21] i've joined a state of confusion so am trying to work that out myself [18:10:50] ok so preilly is going to unblock mobile requests in about 15 minutes [18:11:09] and see if the meltdown goes - they are redeploying wikibase first [18:11:13] ^ preilly MaxSem [18:12:00] We've unblocked one box [18:12:07] Doing the next one right now [18:13:55] are mobile load.php requests so frequent that they're the easiest target to disable? [18:14:14] that was the impression i got [18:14:41] YuviPanda: https://twitter.com/ademianczuk/status/301372527354195968 [18:15:15] I'm not going to click that since that'll clearly be about how the app that nobody works on sucks and the mobile web part that about 5 kickass people work on is superior [18:15:17] :P [18:15:18] jdlrobson: ^ [18:15:46] YuviPanda: what is up? [18:15:51] ok i understand YuviPanda [18:15:54] let me quote it for you "Sometimes a mobile site is just superior to a native app. I'm looking at you, #wikipedia." [18:16:29] jdlrobson: you *do* realize that once we kick your ass that I'm not going to let any of this pass? [18:16:38] and you'll just have to hide behind 'but we have more page views!' [18:16:39] ? [18:16:41] :P [18:16:56] YuviPanda: I like your attitude I just don't believe the content [18:17:15] * YuviPanda considers what level of trash talking is appropriate [18:17:41] considering that if I trash talk mobile web that's about... 6 people? And I do not know how they'll take it (hopefully in the proper way) [18:18:08] and that trash talking the app is only one guy and the meanest thing I'll do is steal half a glass of special lassi from you :P [18:26:32] jdlrobson: are you in the office? is tfinc around? [18:26:45] YuviPanda: not yet [18:26:55] oh [18:26:58] ok [18:27:07] mobile requests to bits have been unblocked - ops is keeping an eye on load for the moment [18:27:19] jdlrobson: did brion come in? [18:27:19] (sorry to bother) [18:27:22] YuviPanda: nope [18:27:26] ah ok [18:27:28] thanks jdlrobson [18:32:33] so awjr MaxSem apparently our startup script includes device which is causing the fragmentation (http://bits.wikimedia.org/en.wikipedia.org/load.php?debug=false&lang=en&modules=mobile.startup%2Csite%2Cproduction-jquery%2Cproduction-only%7Cjquery.hidpi%7Cmediawiki.hidpi%7Cmobile.device.webkit&only=styles&skin=mobile&version=1360687729&*) [18:32:55] yeah - does it need to? [18:33:11] so on staging i'm seeing http://staging.wmflabs.org/load.php?debug=false&lang=en&modules=mobile.styles%7Cmobile.device.webkit&only=styles&skin=mobile&version=1360636444&* and http://staging.wmflabs.org/load.php?debug=false&lang=en&modules=startup&only=scripts&raw=true&skin=mobile&version=1360636444&*&target=mobile [18:33:27] so we seem to have fixed this already somewhat - but we can probably include the styles one to separate them into 2 requests [18:33:59] support [18:34:25] want me to look into that now? [18:34:27] we can even squeeze it in into today's deployment [18:34:32] or do you want to take control MaxSem ? [18:34:58] ok, I'll look [18:35:32] thanks MaxSem [18:35:43] i'll take a look as well to see if i can find some pointers [18:36:37] ok $headLinks[] = $this->resourceLoaderLink( $moduleNames['top'], 'styles', $target='mobile' ); is what's rendering the device [18:36:49] i'd suggest adding a hook that adds $headLinks[] = $this->resourceLoaderLink( 'device name', 'styles', $target='mobile' ); [18:37:56] MaxSem: it seems device is defined twice? [18:38:04] in hooks and body [18:38:38] getEnabledModules should not run $headModuleNames[] = $device->moduleName(); [18:42:48] ugh [18:43:12] jdlrobson: does the device name get appended in the appropriate place elsewhere? [18:43:55] at the moment it's treated like a head module that is static [18:43:58] that appears to be the only place it's invoked [18:44:01] i can throw a quick fix if you want [18:44:07] it's pretty clear to me how to fix this [18:44:14] jdlrobson: let's try it [18:44:44] ok, I'm lost in this insanity [18:44:52] :p [18:45:01] a bit proud that I started it:P [18:45:07] lol [18:45:15] ok i'm almost done with a fix [18:46:39] jdlrobson: it looksl ike some of those device-specific css files are empty [18:46:44] can we just exclude those as well? [18:47:02] android, iphone2, webkit [18:47:14] awjr, this will not help to remove the modules themselves [18:47:39] is there a reason to have those device specific modules though if we're not sending anything for them? [18:48:03] mmm push notifications from gerrit are broken [18:48:04] :( [18:48:16] awjr: MaxSem https://gerrit.wikimedia.org/r/#/c/48656/ [18:49:44] jdlrobson: can we run that only if there is actually device-specific css to return? [18:50:06] awjr: are you referring to empty files? [18:50:11] jdlrobson: yes [18:50:39] awjr: potentially in future we could make device return false when no file exists $device->getModuleName() [18:50:54] e.g. if ( $device->getModuleName() ) { //add module }} [18:51:44] yeah that's what i was thinking jdlrobson - that should be quick to do. we could just remove the css_file_name key/val for the specific devices and update getModuleName() to return false if there's nothing set [18:52:05] that would further reduce fragmentation, particularly for what i assume are the most common devices [18:52:10] feel free to do that - i don't feel that comfortable doing that right now so close to deployment [18:52:21] awjr: it shouldn't be such an issue [18:52:33] MaxSem: im happy to do this real quick - what do you think? ^ [18:52:43] at the end of the day as long as they are separate you are just caching an empty response [18:53:07] sure - I'll merge Jon's rev so you can build on top of it [18:53:12] MaxSem: groovy [18:53:29] testing [18:53:34] jdlrobson: true, but there are already so many things that cause cache variance. the more we can do to reduce that, the less impact we'll have particularly on the backend apaches [18:56:00] especially since the most popular devices don't have their own tweaks.. [18:57:03] New review: MaxSem; "Patch Set 1: Verified+2 Code-Review+2" [mediawiki/extensions/MobileFrontend] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/48656 [18:57:06] Change merged: MaxSem; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48656 [18:57:11] here we go awjr [18:57:14] thanks MaxSem [18:58:53] jcmish, we'll need to deploy ^^^ today, it's a matter of performance [19:02:37] MaxSem: jdlrobson awjr I've been following along [19:02:53] you have my blessing [19:03:57] eek just noticed editing is broken :( [19:03:58] Uncaught Error: Unknown dependency: mediawiki.action.edit [19:05:09] RL magic breaking edit page MaxSem :( [19:05:15] editing is beta, soo.... [19:05:30] yeh but it's still annoying especially when people are using beta to edit [19:05:45] cheaters! [19:06:10] not sure why it is even a dependency [19:06:18] grabbing trevor [19:09:04] ugh, we broke unit tests meanwhile - fixing [19:10:02] heads up: the problem returned [19:10:32] arggh [19:10:51] Wikibase is very semantic lol [19:11:59] MaxSem: Stop swearing [19:12:12] hahahha [19:12:46] ok MaxSem so we can fix edit page easily by adding a hook [19:19:00] arggggghh damn you core [19:19:41] notnarayan: you got dc [19:19:51] browser crashed [19:20:00] that happens :) [19:22:10] notnarayan: :) [19:23:45] so awjr MaxSem turns out there is some horrible code in EditPage.php which assumes a certain module is available [19:27:41] notnarayan: i'm going to head to sleep now [19:27:59] YuviPanda: ah good to hear that :) [19:28:35] YuviPanda: good night! [19:30:01] notnarayan: :) [19:37:26] MaxSem, jdlrobson: https://gerrit.wikimedia.org/r/#/c/48659/ [19:38:55] awjr, can we return false instead of empty string? [19:39:04] MaxSem: yes [19:39:09] any reason for the preference tho? [19:39:48] and do you want the device definition to be 'css_file_name' => false and/or the return for moduleName() to be false? [19:40:06] semantics: string is a module name, false is no module needed [19:40:36] yeah that's fine with me [19:40:59] New review: MaxSem; "Patch Set 1: Verified+2 Code-Review+2" [mediawiki/extensions/MobileFrontend] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/48659 [19:41:01] Change merged: MaxSem; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/48659 [19:41:11] scrw it, can be bikeshedded later [19:41:26] :p [19:41:44] awjr, fancy a push right now?:P [19:41:52] MaxSem: was just about to ask the same thing [19:42:00] lol [19:42:07] MaxSem: lemme take a quick look at the other patchset [19:42:13] doesn't look like it's going to get any worse... [19:42:40] MaxSem i think we can safely cherry-pick those two patchsets [19:43:06] go for it [19:43:25] MaxSem: my local copies of deployment branches are way out of date and it might take me a while to git pull - are yours fresh? [19:43:55] 1 week old [19:44:02] updating [19:44:09] looks like wmf8 and wmf9 are in prod [19:44:17] Maryana: editing is broken in beta [19:44:24] :( [19:44:27] awjr, can you cherrypick these 2 into MF production branch? [19:44:46] should i disable the edit button or should we apply a horrible hack? [19:45:04] MaxSem: yes [19:45:07] Maryana: presenting horrible hack to https://gerrit.wikimedia.org/r/#/c/48663/ - awaits MaxSem to spit in disgust [19:45:47] jdlrobson, it has a dependency [19:46:22] jdlrobson, can you simply stick it into JS? [19:46:52] yeaj im getting a merge conflict [19:46:54] MaxSem: ? [19:46:57] ok, my branches are ready - will update as soon as prod is ready [19:47:05] it's an inline script MaxSem so it needs to be in the output page there :( [19:47:16] preilly has just disabled mobile scripts again [19:47:32] hmm this is an ugly merge conflict [19:47:32] we've had a huge amount of mobile edits today - https://en.wikipedia.org/w/index.php?title=Special:RecentChanges&tagfilter=mobile+edit would be a shame to disable those :( [19:47:36] awjr, can we just push master? [19:47:47] MaxSem: eg deploy early? [19:47:52] yes [19:48:07] i am comfortable with that if folks are able to commit to sticking around to help [19:48:12] jcmish, jdlrobson, jgonera, Maryana ^ [19:48:32] i'm more than happy to start deploying early [19:48:35] awjr: yup I'm here [19:48:37] right now? sure. has everybody gotten food? [19:48:37] this would be to deploy *now* [19:48:48] i haven't but i can wait [19:49:04] ok let's try it [19:49:14] on it [19:49:15] jdlrobson: jgonera : i'm going on a food run soon. do you want something ? [19:49:16] I can probably wait like 1hr, then I'll be starving ;) [19:49:17] maybe someone in the office can order sandwiches or whatever for folks? [19:49:31] i know that you guys are working and am hoping to pick something up [19:49:31] so before doing this 3 things - https://gerrit.wikimedia.org/r/#/c/48598/ didn't get merged Maryana meaning we won't get an accurate number for account creation data [19:49:50] horrible hack better than nothing :) [19:49:59] jdlrobson: we can push that later [19:50:03] edit page is broken [19:50:06] tfinc_, I was going to get something quick today, probably a sandwich from working girls + maybe a salad [19:50:08] :o [19:50:19] Maryana: are you ok with broken edit page going out now? [19:50:20] and we're not tracking isStable for watchlist usage https://gerrit.wikimedia.org/r/#/c/48586/ [19:50:20] boo, unfortunate, jdlrobson. but it's ok [19:50:24] we can push small bug fixes later [19:50:37] jgonera: jdlrobson: give me your order and i'll take care of it [19:50:38] ok let's do this then [19:50:38] yeah, i asked jdlrobson to just disable the edit button for now, awjr [19:50:45] cool [19:51:01] Maryana: i can't disable the edit button yet [19:51:08] that would have to be a small bug fix pushed later [19:51:46] brb food [19:53:33] ok we doing this awjr MaxSem ? [19:53:51] jdlrobson: yes, looks like MaxSem is merging to prod as we speak [19:55:55] MaxSem: im trying to confirm with matthias that it's ok for us to proceed (he has the current window) [19:56:33] awjr, it's an emergency [19:56:40] MaxSem: he's clear [19:56:49] pushing [19:57:15] going to update the code first, scap to get newest messages later [19:58:40] MaxSem: here if you need me [20:00:21] okaym we're live [20:01:04] jcmish, jdlrobson, jgonera, Maryana_food please test ^ [20:01:08] will do [20:01:14] on production or test? [20:01:15] on it [20:01:17] test.m? [20:01:21] assuming test? [20:01:35] production [20:01:40] oh [20:01:42] MaxSem: do we need a cache flush? [20:01:43] okay [20:01:51] seeing broken styles [20:02:01] awjr, lolwut? the cluster's already melting:P [20:02:06] :p [20:02:40] you may see broken/old messages [20:02:45] i'm not seeing new js either [20:02:49] although js is running now [20:02:59] MaxSem: touch/sync? [20:03:12] awjr, touched before syncing [20:03:15] k [20:03:17] it's funny seeing the star in the middle of the search box [20:03:22] @_@ [20:03:37] it may take time to settle [20:03:37] ha [20:03:47] yup i see that lovely star jdlrobson awjr [20:03:52] lol i like it [20:04:01] at least it works :p [20:04:26] looks like 503s are still coming from bits [20:05:06] logging in broken for me too [20:05:27] jdlrobson, the logging changes never went live [20:05:38] the config, I mean [20:05:51] sorry that should be login [20:05:58] it autocorrected me [20:05:58] i can't sign in [20:06:07] i can sign in jdlrobson [20:06:16] hmm i cant [20:06:18] we probably need that startup script enabled again [20:06:22] preilly is not here though [20:06:31] this is on prod right? [20:06:48] yeah i am still seeing 503s from varnish - were requests to bits from mobile disabled? [20:07:09] hmm well my log in is working :D [20:07:13] I'll keep moving along [20:07:20] jcmish: are you in stable or beta? [20:07:23] i'm testing stable only [20:07:25] awjr, see #-ops [20:07:33] log in works in beta [20:07:52] uh oh. but the settings page buttons aren't showing up [20:07:56] beta [20:07:56] so i can't get out of beta.. [20:08:32] Maryana: hahaha you're right [20:08:35] I can't get out either :D [20:08:51] ooh, they just came back for me. try now, jcmish [20:09:19] \o/ Maryana they're back [20:11:27] YuviPanda: https://www.mediawiki.org/wiki/File:Commons-Android-test-capture.webm \o/ [20:11:34] awjr, jdlrobson, I'm going to deploy https://gerrit.wikimedia.org/r/48669 asap [20:12:39] awjr, jdlrobson - objections? cause otherwise I'll merge in 60 seconds:) [20:12:51] looking MaxSem [20:12:54] temporary fix? [20:13:01] do it MaxSem [20:13:09]