[00:02:06] hm, ok [00:02:26] Thank you anyway. [00:02:31] : ) [01:46:26] New patchset: Jdlrobson; "Make special pages declare their modules" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56886 [01:46:26] New patchset: Jdlrobson; "RL: Add modules and module styles to page as modules found" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56887 [01:47:49] New patchset: Jdlrobson; "Make special pages declare their modules" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56886 [01:48:27] New patchset: Jdlrobson; "RL: Add modules and module styles to page as modules found" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56887 [01:51:42] New patchset: Jdlrobson; "RL: Add modules and module styles to page as modules found" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56887 [01:58:49] New patchset: Jdlrobson; "RL: Add modules and module styles to page as modules found" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56887 [02:05:20] New patchset: Jdlrobson; "Use OutputPage headElement function to generate head element" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56875 [12:57:37] [Commons-iOS] siebrand pushed 1 new commit to master: http://git.io/3XxLNA [12:57:37] Commons-iOS/master 9172ac6 Siebrand Mazeland: Localisation updates from http://translatewiki.net. [12:58:30] Change merged: JGonera; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56350 [12:59:16] [android-commons] siebrand pushed 1 new commit to master: http://git.io/0tnCZA [12:59:16] android-commons/master 7f34831 Siebrand Mazeland: Localisation updates from http://translatewiki.net. [13:01:48] [WikipediaMobile] siebrand pushed 1 new commit to master: http://git.io/IXn8Fw [13:01:48] WikipediaMobile/master 13b7a81 Siebrand Mazeland: Localisation updates from http://translatewiki.net. [13:14:19] hey MaxSem, you there? [13:14:26] of course [13:14:55] I just pulled from master and it doesn't work for me... "Go more ResourceLoader native for head scripts and styles" makes MobileFrontend a 0-byte blank page for me [13:15:16] update core [13:15:18] is it ok on your instance? [13:15:20] oh [13:15:45] and btw you really need to output PHp errors;) [13:15:53] it's odd that I didn't get any error, maybe something that makes PHP expldoe and is only in php.log [13:15:58] I do! [13:16:07] error_reporting( E_ALL | E_STRICT ); [13:16:07] ini_set( 'display_errors', 1 ); [13:16:15] let me double check [13:16:17] oh, right [13:16:24] I don't, not on this computer... [13:16:30] [android-commons] yuvipanda pushed 1 new commit to master: http://git.io/R0LHrg [13:16:30] android-commons/master b667bcb YuviPanda: Trim whitespace in one of the messages [13:16:44] no it exploded correctly for me before I updated [13:16:51] with errors galore [13:17:05] yeah, I just forgot to set error reporting here after installing everything [13:17:13] it would explode on my laptop ;) [13:17:16] that's why I prefer setting this stuff explicitly in LocalSettings [13:17:31] hm, not a bad idea, thanks [13:17:55] my setup is so complex that I carry over my working copy between my machines [13:18:16] hm [13:43:13] New review: JGonera; "(1 comment)" [mediawiki/extensions/MobileFrontend] (master) C: 1; - https://gerrit.wikimedia.org/r/56886 [13:55:47] New review: JGonera; "(3 comments)" [mediawiki/extensions/MobileFrontend] (master) C: -1; - https://gerrit.wikimedia.org/r/56887 [14:40:45] New patchset: MaxSem; "WIP: [Card 446] Don't vary page HTML on X-Device" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/55226 [15:01:02] New review: JGonera; "I don't have Protanopia or Deuteranopia but find it much less readable than previous colors. The con..." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56626 [15:17:32] jgonera: around? [15:17:36] jgonera: we are still generating the Month table [15:17:45] jgonera: and there have been no drastic changes to the SQL [15:17:51] will be in 15 minutes [15:18:07] then maybe there's some bug outside of sql [15:18:45] jgonera: yeah, because I see android being reported properly [15:19:08] jgonera: poke when around [16:19:17] YuviPanda, sorry, this turned out to be rather longer [16:19:37] so do you know what might be the reason for those two graphs being identical? [16:19:37] :) [16:19:48] jgonera: no, but i'm logging into stat1 now to check [16:19:53] do you have access to stat1? [16:19:54] ok [16:20:08] jgonera: were you seeing both the SQL things to be the same? [16:20:10] just try running a query for success and error for the same timeframe [16:20:15] yeah, doing it [16:20:17] maybe our logging is messed up [16:20:39] SQL is different, so I'm not sure how's this possible [16:20:50] hmm [16:20:55] you rang? [16:21:05] are you stalking 'logging'? [16:21:15] no, 'messed up' [16:21:28] hahaha [16:21:35] Change merged: JGonera; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56625 [16:22:15] (actually just happened to be glancing through channels) [16:22:20] :) [16:23:12] ori-l, I noticed that the graph for success and errors on our dashboard is identical for Web [16:23:21] which is very unlikely to be real data ;) [16:23:23] what's the dashboard? [16:23:32] er, where's [16:23:37] http://mobile-reportcard.wmflabs.org/ [16:23:42] oh, right [16:25:14] and where's the sql? [16:26:13] jgonera: hmm, 78 success since 29 mar and 17 failures [16:26:17] EL Seems okay [16:26:22] ori-l: gerrit.wikimedia.org/r/gitweb?p=analytics/limn-mobile-data.git [16:26:44] that's actually the best news I hoped for, that means that the handshake actually fixed something [16:26:51] ori-l: https://gerrit.wikimedia.org/r/gitweb?p=analytics/limn-mobile-data.git;a=tree;f=mobile;h=fab299e251b71b99f0438a62ad86d0673dd57a08;hb=HEAD more like [16:26:53] jgonera: :) [16:26:55] so we just need to find what's wrong with the dashboard [16:27:37] I should have access to stat1 today or tomorrow so if you have other things it can wait [16:28:30] jgonera: http://pastebin.com/PQi5nbj6 [16:28:45] New review: JGonera; "Nothing against this patch, but diff seems to be working really weird anyway. If at least some of th..." [mediawiki/extensions/MobileFrontend] (master) C: 1; - https://gerrit.wikimedia.org/r/56627 [16:28:48] so yeah, it is having an effect, I think :) [16:29:36] wow, that's even better, only 5 errors because people were logged out [16:29:49] thanks for running this query ;) [16:30:40] jgonera: :) it's only from 29th [16:30:43] want me to run it even backwards? [16:30:50] 'further' [16:30:51] not backwards [16:31:01] what's the left join sum(1) as web thing? [16:31:10] hm, I don't remember when we deployed the handshake [16:31:14] jgonera: ^ [16:31:31] I don't think I've written that, have I? [16:31:46] also, dat query... just needs a networking stack to become an operating system [16:31:52] jgonera: you did :P [16:31:54] hehe [16:32:03] let me check [16:32:06] well, run it from inside a terminal session in Emacs... [16:33:16] hm, I have [16:33:27] oh right [16:33:47] ori-l, so what's the question? [16:34:49] nothing, this query surpasseth my comprehension sadly [16:34:57] yeah, mine too, I think [16:35:06] not that my comprehension is anywhere near advanced.. [16:36:34] sum(1) is the same as sum(*) [16:36:58] it's just quicker in some databases, I'm sure about postgres, not sure about mysql [16:37:40] hm, or is it [16:37:53] I'll check the SQL when I can access stat1 [16:38:08] and if SQL is OK, then we'll start investigating more [16:38:36] ori-l, where is the data from NavigationTiming stored? [16:38:44] is it also stat1? [16:39:02] jgonera: the data is all in s1-analytics-slave, we just access it from stat1, I think [16:39:08] technically nothing is stored on stat1. you just need a machine on the cluster to tunnel through [16:39:14] do you have access to fenari? [16:39:50] true, I don't have access to anything because of my laptop burning down and my SSH keys being invalidated [16:40:18] and Daniel needs to verify my identity through a hangout before adding my new key [16:40:20] oh, right. sorry to hear that :/ [16:41:01] anyway, when I have it, where can I find it? is it also EL-based? [16:41:21] yeah, it's in the table "NavigationTiming_5336845" [16:41:25] also in mongo [16:41:41] is mongo on the same host? [16:42:31] no, mongo is on vanadium [16:42:45] ok [16:43:02] jgonera: http://www.mediawiki.org/wiki/Extension:EventLogging/MongoDB [16:43:15] probably not, but do we have a dashboard for that or just raw DB data? ;) [16:45:31] just raw db data [16:45:35] ok [16:46:03] so I just need to get my new SSH key added asap ;) [16:46:04] +1 yes :) [16:55:13] awjr, can you make the training in calendar start a bit earlier because we will need to prepare commits? say, 15 minutes [16:55:19] ok, restarting to Windows so maybe hangout won't hang this machine... [16:55:23] sure MaxSem [16:55:43] thanks [16:59:52] having some technical problems - i should be in standup in a minute [17:25:08] New review: Jdlrobson; "Yeh I've noticed this as well - especially with urls. Not sure if there's a better way to use librar..." [mediawiki/extensions/MobileFrontend] (master); V: 2 C: 2; - https://gerrit.wikimedia.org/r/56627 [17:28:59] New patchset: Jdlrobson; "Change diff colors after design input" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56626 [17:28:59] New patchset: Jdlrobson; "Add new diff test case for simple insertion" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56627 [17:30:23] New review: Jdlrobson; "I've sent you the email. Vibha and Munaf tested these out so they should be the most optimum for eve..." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56626 [17:32:16] mmm, am I right that jslint's "Implied global 'isMobileDevice'" should be STFU'd with /*global isMobileDevice:true */ ? [17:38:54] MaxSem: https://gerrit.wikimedia.org/r/#/c/56875/2 < did you get a chance to test that out? [17:39:15] MaxSem: you can edit jshintrc but it's bad practice to introduce globals in js [17:39:20] what introduces isMobileDevice ? [17:39:31] the startup module [17:39:43] that's how it's made [17:40:36] New patchset: Jdlrobson; "WIP: [Card 446] Don't vary page HTML on X-Device" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/55226 [17:40:53] the core startup module? [17:41:55] yup [17:42:19] https://gerrit.wikimedia.org/r/#/c/55446/ [17:42:28] New review: Jdlrobson; "(3 comments)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56887 [17:42:52] impressed you are doing js ;-) [17:43:52] doesn't mean I ain't hate it:) [17:44:29] i need to familiarize myself with this code [17:44:37] but it seems you can solve the problem by wrapping in a closure [17:45:14] (function(){ alert('something'); /* code goes here */ })(); [17:50:41] ok MaxSem so I'm familiar with how the startup module works now [17:51:25] Some of the regexs scare me - '240x240' for instance [17:51:58] i'd rather we started with ones we know don't work and add new ones as problems get reported [17:52:00] they've successfully worked in production for ages [17:52:25] jdlrobson: poke brion to come on irc if you see him? [17:52:38] YuviPanda: he's with monte [17:52:44] ah ok [17:52:51] MaxSem: yeh but I'd worry we'd kill the desktop experience for some people in process [17:53:08] we won't, I have tests [17:53:25] how do you imagine a breakage? [17:53:51] if they have a suspicious user-agent, they get redirectd to the mobile site [17:54:18] can we not have a whitelist instead [17:54:32] these regexes are taken from Squid ACLs for a reason;) [17:54:47] whitelist of what? [17:54:56] do we have the original squid ACL or are we just speculating? [17:55:04] I do [17:55:20] I personally copied these regexes from it [17:55:34] i really wish those squid confs were in git [17:55:50] er puppetized i mean [17:55:56] New review: Nikerabbit; "As an outsider this commit message leaves me puzzled." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56459 [17:56:49] New review: MaxSem; "It is merged with the main iPhone device type;)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56459 [18:02:03] hey brion [18:02:04] got my email [18:02:05] ? [18:02:27] not yet been showing monte around [18:02:55] yeah i figured we're skipping the syncup [18:03:14] brion: okay :) [18:04:08] jdlrobson, re https://gerrit.wikimedia.org/r/#/c/56875/2 [18:04:37] MaxSem: yup? [18:04:41] so you know yourself that you can't vary on device [18:07:00] New review: Jdlrobson; "Well going forward I believe we should match the desktop id's and classes. If these are inconsistent..." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56324 [18:07:35] New patchset: Jdlrobson; "Rename #mw-mf-search to #searchInput" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56324 [18:07:39] New patchset: JGonera; "Add missing dependency" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56943 [18:09:12] New patchset: Jdlrobson; "Bring watchlist css/html markup closer to desktop" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56515 [18:09:51] Maryana, we have qa today? [18:10:02] New review: Jdlrobson; "(1 comment)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56515 [18:10:33] jgonera, i think so. jcmish, is the stuff that's ready for testing on staging yet? [18:10:57] or beta now? [18:11:18] Change merged: Jdlrobson; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56943 [18:12:43] jcmish, is the stuff that's in the ready for testing column on staging/beta? [18:12:50] Maryana: Off Topic, the story of jdlrobson's lost slippers is totally true. rmoen|afk or tfinc can verify :) [18:13:00] wow, I see a tfinc but I guess that's just his bouncer? [18:13:09] Maryana: looking now [18:13:18] jcmish: you pinged me y'day? [18:13:20] or was that friday? [18:13:34] YuviPanda: that was friday [18:13:40] and now I don't remember why :) [18:13:47] :D [18:13:47] so it probably wasn't important [18:14:01] hehe :D [18:14:21] jcmish: perhaps the android testflight link? [18:14:29] oh yeah! [18:14:34] that was it [18:14:37] then I emailed it :D [18:14:48] Did you already know about it? [18:16:07] Maryana: I'm not sure [18:16:13] need to ask jdlrobson and jgonera [18:16:26] should we bump back the qa/design mtg then? [18:16:35] #wikimedia [18:16:43] New review: Jdlrobson; "Note that mobile-legacy will become redundant when Max's work is finished on https://gerrit.wikimedi..." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56887 [18:16:44] * mark_ is embarassed [18:16:53] * mark_ typed in the wrong window [18:16:58] yeah being off Friday threw my day off [18:17:01] no worries, mark_ :) [18:17:03] let's do it early tomorrow [18:17:10] okey doke [18:17:16] what's your day look like tomorrow Maryana ? [18:17:23] thanks Maryana [18:17:51] jcmish, i have meetings from 9 to 10:30, so it would have to be 10:30 or later [18:18:42] MaxSem: another way at looking at the startup module - is that anyway we could vary it on X-Device and simply leave it blank if the device doesn't support jQuery ? [18:18:53] MaxSem, and more appropriate :) [18:19:03] 10:45 :) [18:19:09] Changing it now [18:19:16] jdlrobson, overriding the core sounds scary [18:19:19] anyway, do you think we should have a dedicated machine for zero portal? [18:19:26] no [18:19:39] take a look how www.*.org portals are made [18:19:54] MaxSem, could you give an example? [18:20:41] operations/mediawiki-config/docroot/www.wikipedia.org [18:20:58] and relevant stuff in operations/apache-config [18:21:12] k, thanks jcmish! [18:21:46] yurik, these portals are served by every apache [18:22:25] MaxSem, is that a part of puppet GIT repo? [18:22:40] yes [18:23:32] jcmish: no i didn't :) [18:23:42] jcmish: i'll look into it :) [18:23:43] thanks [18:23:48] YuviPanda: WOOHOO [18:23:54] I'm glad to show you something new :D [18:23:55] awjr, I'm not sure who should I ask to add this SSH key if Daniel is not around [18:24:04] I can help ya get it up and running too! [18:24:26] MaxSem, i must be blind, which file is that config in? [18:24:44] what's adam's irc name? [18:24:44] yurik: ^ ? [18:24:47] dr0ptp4kt, [18:24:56] thankx :) [18:25:02] dr0ptp4kt: ok good news is i can now replicate it locally [18:25:04] dr-something :) [18:25:06] yurik, apache? operations/apache-config [18:25:07] jgonera: good question, try RobH or Ryan_Lane [18:25:07] gonna try and work out what's going on [18:25:16] www.wikipedia.org.conf [18:25:50] jcmish: :) [18:25:55] jcmish: for android the current testing sortof works fine [18:26:01] jcmish: which is just me pushing to market every week :P [18:26:18] yeah but it'll be nice to not have to do anything else :D [18:26:27] take something off your plate [18:27:01] :D [18:27:03] yeah [18:30:39] THERE"S NOT ENOUGH COLOURS! [18:30:52] * jdlrobson debugging by body { color: X; } [18:31:05] yeah, my whole IRC screen is mostly beige and black, with a few yellow smilies [18:31:25] i say every person should get their own color [18:32:00] awjr: dumb question... is there a reason [18:32:16] when you go into mingle it starts at that home screen [18:32:21] with Analytics? [18:32:26] oh, i dont know what that's about [18:32:41] it's really annoying thoug [18:32:54] i can ask james alexander who i think has been maintaining the mingle setup [18:33:12] yurik, found it? [18:33:41] MaxSem, i found www.wikipedia.conf in operations/apache-config project [18:33:49] yup [18:33:50] (was looking through the pupet git repo) [18:33:56] awjr: hehe I hate to be bitchy I just thought about it again and thought I'd ask :D [18:33:57] that's why the confusion [18:34:14] we've got a lot of repos:) [18:34:32] jcmish: i poked around a bit and i dont see any particular settings to allow you to get around that [18:34:33] yeah, i thought all config was in one "pupet" [18:34:45] :-) awjr that's okay [18:34:48] jcmish: personall though i just bookmarked https://mingle.corp.wikimedia.org/projects/mobile [18:34:52] I can suck it up just thought I'd ask [18:34:57] k will do [18:35:41] running to grab a bite [18:35:41] back in a bit [18:36:00] @embed is broken in ResourceLoader :( [18:38:14] ^ dr0ptp4kt appears to be a core bug [18:57:24] New patchset: JGonera; "Refactor the way we run mobile JavaScript tests" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/50442 [18:58:02] jdlrobson, can you check if this occasionally segfaults phantomjs for you when doing make qunit? ^ [18:58:36] if not, I'll assume it might be my 1GB of RAM [19:00:09] sure 1s [19:00:52] MaxSem, i have been digging through the wwwi.wikipedia.conf - is that file in prod? [19:01:01] yes [19:01:07] i mean, the ^/$ is redirected to extract2.php, which i can't fine [19:01:09] find [19:02:02] MaxSem, line 15 [19:02:03] [22:20:42] operations/mediawiki-config/docroot/www.wikipedia.org [19:04:22] MaxSem, not sure what you mean - I'm looking at gerrit.wikimedia.org/r/operations/apache-config [19:04:37] file www.wikipedia.conf [19:04:50] I was referring to operations/mediawiki-config repo, sorry [19:06:50] cloning... someday i really hope we can get rid of all the complexity... (wishful thinking) [19:07:28] nah, it's all very clearly factored to different repos [19:07:46] you can't keeep puppet manifests and MW config in the same repo [19:08:03] first of all, because puppet is ops-only to merge [19:08:11] do we have a doc describing the overall structure of the repos and what gets connected to what and how? [19:08:35] not that I've heard of [19:08:40] heh :( [19:09:08] ok, i got that repo, and extract2.php is weird - it contains its own name [19:09:57] but is there something specific i should take a look at? [19:10:08] btw, if you want, we could do a short hangout [19:11:09] sure [19:11:10] one sec [19:13:17] yurik, skype? [19:13:43] MaxSem, had to grab a webcam, sure, that works [19:14:04] MaxSemenik [19:15:53] sorry, laptop is running slow, sec [19:16:02] i miss my desktop :) [19:16:39] oh, its installing webcam, that's why [19:16:53] they've not given you a казённый laptop? [19:18:04] i'm getting one soon [19:18:13] haven't heard "kazyonny" in ages :) [19:19:09] grrr, now skype is updating... next time -- google hangout. No more bloatware [19:21:24] MaxSem, added [19:32:04] New review: awjrichards; "(2 comments)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/55226 [19:42:44] awjr, MaxSem, Maryana do you need anything before I leave? [19:42:51] no [19:42:56] no, jgonera - have a good night [19:43:05] seeya, jgonera :) [19:43:42] I was unable to get anyone from ops to change my key, they had a meeting or possibly still have a meeting [19:43:53] I hope it'll be possible tomorrow [19:44:40] I'm not sure how the verification process looks like, couldn't just Maryana tell Daniel that me = me? [19:45:05] jgonera: they'll probably want to verify over google hangout or something [19:45:25] jgonera: who have you communicated with about it so far? [19:45:39] but how do they verify this? should I print they public key and show it next to me in a webcam? ;) [19:45:48] jgonera: yes [19:45:51] jgonera: or, smoething like that [19:46:04] mutante, jeremyb_ [19:46:15] i dunno how they'd want to do it exactly, jgonera, but that would be my best guess [19:46:21] I tried RyanLane and RobH but they're in a meeting [19:46:36] jgonera: ok i can try to follow up with folks later this afternon to get things taken care of. when will you be around tomorrow? [19:46:45] rob halsell? [19:46:55] Maryana: rob halsell = robh [19:47:02] he's right here… :) [19:47:07] New review: awjrichards; "(2 comments)" [mediawiki/extensions/MobileFrontend] (master) C: -1; - https://gerrit.wikimedia.org/r/56887 [19:47:41] jgonera, you want me to nudge him? [19:47:42] hm, I was not planning to be earlier than 10am which is 1am PDT [19:48:16] ok jgonera [19:48:34] want to stick around for a few minutes and see if any ops are back from lunch or whatever? apparently robh is there [19:49:56] if you can catch someone tomorrow, just please tell them that I'd like to do that anytime before lunch hours PDT [19:50:12] RobH is not responding on IRC and I got to go [19:50:15] jgonera, i have robh here [19:50:18] i wish gerrit had a nicer way of display dependencies without having to dig into specific changesets [19:50:20] one sec plz! [19:50:23] oh ok [19:50:24] jgonera: Try using EU staff? ;) [19:50:41] I don't know who's in EU [19:50:48] EU staff would be mark (bergsma), paravoid (faidon), apergos (ariel) [19:50:56] and I was pointed first to mutante, later RyanLane and RobH [19:50:59] can't think of who else off the top of my head [19:51:03] jgonera: heyas, [19:51:04] that's it for ops [19:51:09] RobH, hey [19:51:12] so im told by maryana you are you [19:51:20] hurray for meatspace! [19:51:21] which is good enough since she can tell if you arent you [19:51:24] I am, I swear ;) [19:51:27] sooo, lemme pull up your request [19:51:41] 4854 [19:51:58] RobH, it's RT-Ticket: wikimedia #4854 [19:52:04] thx [19:52:11] i was searchin [19:53:41] jgonera: Hrmm, im about to break the rule we just set. since I know what you look like, can we do a hangout real quick? [19:53:44] that way i dont break any rules [19:54:07] if your system can do that now [19:54:29] if it is him, how will you know he's not under duress? [19:54:41] * awjr opens mouth, inserts foot [19:55:00] you could have him just go to .nl or .gr and verify in person [19:55:08] RobH, I have to reboot to Windows, my web cam is really crappy [19:55:09] ooor [19:55:13] I can use my phone [19:55:18] i can call you yes [19:55:43] if i call you, it means i verify its you [19:56:02] wait, but you mean Gtalk call? [19:56:11] nah, meant i would dial your phone [19:56:26] well, my US number doesn't have roaming [19:56:35] and you can't verify me using my Polish number [19:56:41] because you don't know it, right? [19:57:21] Maryana, can you make a video call through gtalk> [19:57:22] ? [19:57:29] OHHHH [19:57:33] jgonera: You have office wiki right? [19:57:39] jorm: just made a good potion [19:57:39] I do [19:57:41] login to office wiki [19:57:44] on your user page [19:57:52] put the SSH fingerprint [19:58:03] which you can get by running ssh-keygen -lf [19:58:13] then i can compare to whats in RT to confirm its not email social engineering [19:58:18] sorry you are guiniapig for all this dude [19:58:29] hah [19:58:41] so that and tell me when done [19:58:51] and i will push your new key right away (as in minutes after you tell me) [19:58:54] cuz you waited long enough [19:59:12] https://office.wikimedia.org/wiki/User:JGonera_(WMF) [19:59:30] jgonera, you're a trooper. thanks [19:59:35] if it's deployed in 10 hours, that's good enough ;) [19:59:38] now you just need to submit a saliva sample [19:59:46] via carrier pigeon [19:59:58] when will the saliva scanner arrive? [19:59:59] make sure it's a white pigeon, though [20:00:04] gray pigeons are no good [20:00:12] highly subject to social engineering [20:00:29] Maryana: RFC 2549? [20:00:30] ;) [20:00:39] hrmmmmmm [20:00:49] i hate to say they dont match, but they dont, but maybe i didn something wrong [20:00:50] checkin [20:00:51] is that the RFC for pgieon-base packet transmission? [20:00:53] jeremyb_: tcp over carrier pigeon/ [20:01:04] RobH: he put 2 keys in the ticket... which did you hash? [20:01:05] RobH, I sent a second key today, in my last e-mail [20:01:16] does RT receive attachments in followup emails? [20:01:16] ahhh [20:01:20] i hashed the first [20:01:23] i see it though [20:01:27] my bad [20:01:43] jgonera: huzzah! [20:01:48] hurray! [20:01:50] \o/ [20:01:50] that works, i'll get it setup now [20:01:55] thank you [20:02:01] ok, signing off [20:02:05] see you tomorrow! [20:02:07] i need to add the office wiki as an ok way to do this [20:02:18] good night, jgonera, thanks for your patience [20:02:22] New patchset: Jdlrobson; "Use embed the correct way, fix Nokia N95s" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56956 [20:02:33] ^ brion [20:02:34] robh, don't forget the part about the pigeons [20:02:42] african or european? [20:02:44] \o/ [20:02:45] arg wait a sec [20:02:52] lol RobH [20:03:23] reminds me of last year's april fool's FA: http://en.wikipedia.org/wiki/Pigeon_photography [20:03:32] New patchset: Jdlrobson; "Use embed the correct way, fix Nokia N95s" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56956 [20:04:15] nazi spy pigeon thanks you for your help, robh: http://en.wikipedia.org/wiki/File:Bundesarchiv_Bild_183-R01996,_Brieftaube_mit_Fotokamera_cropped.jpg [20:04:57] Maryana: Get that added as a WikiLove default [20:05:12] haha [20:06:31] jdlrobson: i was looking at https://gerrit.wikimedia.org/r/#/c/56626/2/less/specials/mobilediff.less and got concerned i might actually have color blindness as i suddenly couldn't tell the difference between current diff colors and that changeset; then i realized those changes are only for alpha [20:06:42] hahaha [20:06:46] that's funny [20:06:52] sorry i should have put alpha in the commit heading [20:06:57] New patchset: awjrichards; "Change diff colors after design input [Alpha]" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56626 [20:07:04] Did you know section on wikipedia homepage is amusing me [20:07:07] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56626 [20:07:19] New patchset: awjrichards; "Add new diff test case for simple insertion" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56627 [20:07:29] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56627 [20:07:49] limechat doesnt render thumbnails from wikipedia projects ;_; [20:07:53] had to actually click [20:07:55] jdlrobson: https://gerrit.wikimedia.org/r/#/c/56956/ fixes the super-whack nokia issue? [20:08:14] yup awjr - i'm not sure how best you can test it yourself though :S [20:08:30] jdlrobson: get somene in the office to verify [20:08:38] i've shown brion :) [20:08:44] does that count? [20:09:07] sure - brion, https://gerrit.wikimedia.org/r/#/c/56956/2 fixes the preiouvly super mysterious nokia problem? [20:09:44] jdlrobson: so why was it working with ?debug=true? [20:10:45] awjr: it puts it on a newline [20:11:12] New review: awjrichards; "LGTM but can't test - if someone can verify this, please go ahead and merge." [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 1; - https://gerrit.wikimedia.org/r/56956 [20:11:32] oh funky [20:11:36] i think the embed processing doesn't actually happen on debug=true [20:11:51] So awjr MaxSem i'd say https://gerrit.wikimedia.org/r/#/c/56875/ is the patch to merge [20:11:55] New review: MaxSem; "(2 comments)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/55226 [20:11:59] gives us the biggest win [20:12:20] also awjr https://gerrit.wikimedia.org/r/#/c/56618/ should be an easy one [20:12:31] jdlrobson, not with that fixme [20:12:39] cool - im about to run to the eye doctor so i can take a look when i get back [20:12:50] MaxSem: which fixme? [20:12:52] New review: MaxSem; "(1 comment)" [mediawiki/extensions/MobileFrontend] (master) C: -1; - https://gerrit.wikimedia.org/r/56875 [20:13:00] ^^^ [20:13:15] dfoy, are we meeting today? [20:13:24] oh MaxSem .. so I figured that would get addressed with the startup module [20:13:32] New review: Brion VIBBER; "Looks good to me -- this produces the output I expect to see in the minified CSS and works correctly..." [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 1; - https://gerrit.wikimedia.org/r/56956 [20:13:33] yurik: just sent out the invite [20:13:34] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56618 [20:13:50] jdlrobson: confirmed that that fixes the styles on the n95? output looks correct to me [20:14:20] MaxSem: don't we currently vary with that inline script anyway ? so i'm hardly introducing anything new [20:14:27] at the moment i can't see the wood from the trees though as there is so much going on [20:14:42] jdlrobson, don't introduce somethingI'll have to fight with the moment it gets merged;) [20:14:44] dfoy, could we do it at 2? [20:14:51] i have a meeting at 3 [20:15:08] MaxSem: i'm not introducing something with this though - i'm just rewriting an existing problem that i don't have a solution for yet [20:15:11] yurik: ok, I'll revise [20:15:31] jdlrobson, MaxSem do we need to start thinking about varying on JS vs non-JS? [20:15:36] dfoy, or even earlier, up to you [20:15:52] because i've notiecd that problem in at least one other patchset i looked at today [20:15:52] awjr_optometrist: i think so yes - but not in this patch set :) [20:15:53] awjr_optometrist, NEVAR [20:16:07] we need device specific js [20:16:15] with a little setting, every phone is potentially a non-js [20:16:15] well ew shouldn't be introducing changes that are going to be immediately updated when Max's changes get merged [20:16:33] we should use the word enhanced mobile site vs basic mobile site [20:16:40] but… i really need to go - back in a while! [20:16:40] mebbe set it unconditionally [20:16:42] ? [20:16:52] MaxSem: set what unconditionally? [20:17:06] that little js snippet [20:17:25] MaxSem: well by doing that it will expect js to correct when the js has finished loading [20:17:51] I suppose we could add a snippet to the end.. [20:17:52] 1 [20:17:54] 1s [20:22:02] Maryana: any key points about Commons and mobile? I want to finish the rough draft of the Welcome to Commons brochure today or tomorrow (and then seek community feedback) [20:22:17] http://commons.wikimedia.org/wiki/Commons:Welcome_to_Commons_brochure [20:22:26] bah, haven't had a chance to get to it [20:22:33] I just have the mobile section left to draft. [20:22:39] will do today [20:22:48] thanks! [20:32:10] New patchset: Jdlrobson; "Use OutputPage headElement function to generate head element" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56875 [20:32:14] ^ MaxSem that should make you happy [20:33:21] New patchset: Jdlrobson; "fix jshint error" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56961 [20:33:32] Change merged: Jdlrobson; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56961 [20:36:22] New patchset: Jdlrobson; "Bug 46760: Beta Regression: Fix styling problem for search in beta" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/55776 [20:39:40] New patchset: Jdlrobson; "Bug 46760: Beta Regression: Fix styling problem for search in beta" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/55776 [20:39:59] New review: Jdlrobson; "This does fix the bug for me. I've not made any changes..." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/55776 [20:45:48] New review: Jdlrobson; "Still doesn't seem quite right when running headless" [mediawiki/extensions/MobileFrontend] (master) C: -1; - https://gerrit.wikimedia.org/r/50442 [20:49:46] New review: MaxSem; "(1 comment)" [mediawiki/extensions/MobileFrontend] (master) C: -1; - https://gerrit.wikimedia.org/r/56875 [21:00:51] dfoy, want to do it here or there? [21:01:05] dr0ptp4kt, ^^ [21:01:39] yurik: just looking for an available room - all are reserved but not all are actually in use [21:02:00] dfoy, heh, we could just sit on the nice couches [21:02:04] yurik: I'll meet you downstairs [21:03:32] valeriej: heya! [21:03:36] I just filed a couple more 'easy'ish bugs [21:04:22] YuviPanda: Awesome! I'm working on writing a talk proposal now, but I'll definitely take a look tomorrow! [21:04:46] valeriej: sweet. I've marked them with the 'easy' keyword, and since you're bugmeister in training, I guess you'll find them easily enough ;) [21:05:09] YuviPanda: Yep! :) [21:05:15] hah, sweet :) [21:09:28] New patchset: Jdlrobson; "Use OutputPage headElement function to generate head element" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56875 [21:09:41] New review: Jdlrobson; "(1 comment)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56875 [21:19:52] ragesoss: ! [21:19:59] ragesoss: fixed that bug :) [21:20:05] goddamn that took a bit. [21:20:13] [android-commons] yuvipanda pushed 1 new commit to master: http://git.io/9rufcQ [21:20:13] android-commons/master 5e45cb2 YuviPanda: Disable the multi upload photos view while adding categories... [21:22:10] ragesoss: updated bit.ly/android-commons [21:22:18] ragesoss: now needs only event-logigng, then can go out [21:23:30] New review: awjrichards; "(1 comment)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56703 [21:30:59] jdlrobson: im a little confused by https://gerrit.wikimedia.org/r/#/c/56875/4/includes/specials/SpecialMobileMenu.php [21:31:26] how so awjr [21:31:29] shouldn't we only show SpecialMobileMenu if we're in the mobile mode? why check if we're in mobile mode, and then later force mobile mode? [21:32:08] actually that should just be $context->getSkin()->addPageClass( 'navigation-enabled navigation-full-screen' ); [21:32:13] it works on desktop [21:32:17] that's the intention :) [21:32:24] it's all part of my turning mobile into a skin plan [21:32:29] let me quickly fix that up [21:32:42] * jdlrobson has too many evil plans [21:33:16] mwahahahahah [21:34:43] oh wait a minute… there is no addPageClass just a getPageClasses [21:34:54] so that is right [21:34:57] ^ awjr [21:35:18] * awjr scratches head [21:35:22] but you are right we should force mobile mode and not do check [21:35:38] but in future i don't want to force mobile mode [21:35:43] it should just f'in work :) [21:36:05] although saying for that forcing mobile mode doesn't seem to be working for me.. [21:36:26] if you force the mobile mode before you get the skin, it whould be sure to get the mobile skin, right? [21:36:28] s/whould/should [21:36:50] apparently not awjr [21:37:15] oh probably because the skin has already been initialized by the time you're forcing mobile mode [21:37:24] yeh [21:37:31] not sure why that code is there or who introduced it [21:37:54] apparently i thought that would work [21:38:01] (i seem to have introduced it :)) [21:38:27] what does setContentTransformation do?! [21:38:28] isnt there a way to add classes from OutputPage? [21:39:01] awjr: not yet [21:39:17] well there is getPageClasses [21:39:32] which is part of the skin [21:39:53] theoretically i could just rewrite that [21:39:54] oh yeah - it's hard to keep track of this stuff :p [21:39:58] and not shave addPageClass [21:40:07] yeh that's why i'm trying to divide and conquer slowly :) [21:40:39] only skin can decide [21:41:02] might be able to add a hook that special pages on mobile can evoke? [21:41:13] are you happy letting this slide for time being [21:41:19] i expect lots more changes to the skins [21:41:25] but it's slowly starting to explode my mind [21:41:31] i have about 10 open patch sets doing improvements [21:41:42] i mean i'm not introducing anything new [21:41:55] just deprecating the horrible setForceLeftMenu ;-) [21:42:34] so when i go to Special:MobileMenu from desktop view, i dont really see anything, except for a blank article page that says 'Site navigation' [21:42:51] correct … at the moment [21:42:52] b [21:43:14] sadly the left menu on desktop is hardcoded in properties [21:43:26] right [21:43:34] everything in mobile should work in desktop though [21:43:43] the fallbacks in mobile we will need in desktop if our skin is to work [21:43:50] and in the words of obama YES WE CAN [21:43:53] * jdlrobson goes for his massage [21:44:31] ok, i dont disagree in principle, but i dont see the point in making things more complicated to support things that don't currently work, particularly when we haven't really discussed how these things *should* work with the rest of the team or with product [21:47:48] [Commons-iOS] brion pushed 2 new commits to master: http://git.io/V--xaw [21:47:48] Commons-iOS/master de47601 Brion Vibber: update readme [21:47:48] Commons-iOS/master 3387910 Brion Vibber: Add new localization files to project [21:54:08] awjr, have you seen my replies on https://gerrit.wikimedia.org/r/#/c/55226/ ? [21:54:22] not yet MaxSem, im digging into 56875 [21:57:16] New review: awjrichards; "(3 comments)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56875 [22:06:17] Yuvi: I can test in a few hours. [22:06:57] ragesoss: sweet [22:07:05] ragesoss: i'll do a final update and head to sleep [22:07:10] (adding EventLogging Now) [22:07:19] ragesoss: test in your 2.3 too :) once all done I'll just push it out tomorrow [22:08:55] awjr: could we use the hook OutputPageBodyAttributes to add the class from the special pages? [22:09:21] awjr: pm [22:09:21] maybe jdlrobson [22:09:21] * awjr looks [22:09:42] [Commons-iOS] brion pushed 1 new commit to master: http://git.io/CY1ksQ [22:09:42] Commons-iOS/master c43519e Brion Vibber: Fix upload button behavior after upload... [22:10:39] looks like it, jdlrobson [22:10:53] would i register a global hook and check the title? [22:11:03] brion: hey! how's the create-account stuff? [22:11:12] still waiting [22:11:22] you would register the hook in includes/MobileFrontend.hooks.php [22:11:33] jdlrobson: ^ [22:11:47] awjr: then check title from $out [22:11:49] i'm gonna have a go now [22:12:02] cool jdlrobson, i think that would be a better approach [22:12:41] brion: can't you threaten someone with a vial of smallpox and get it merged? :) [22:12:49] MaxSem: saw your comments, makes sense. im gonna try to do some testing on that patchset [22:18:26] awjr: remind me what's the safest way to check the name of a special page? [22:18:36] e.g. if title of page == 'MobileMenu' [22:19:10] $title->isSpecial( 'MobileMenu' ) [22:20:30] sweet [22:20:37] ok i've actually worked out a way to do this without hooks i think [22:20:38] 1 [22:23:29] * jdlrobson wishes MaxSem would move MobileFrontendDeviceDetectModule into a separate file for readability reasons https://gerrit.wikimedia.org/r/#/c/55226/9/includes/MobileFrontend.body.php - awjr can you be the deciding vote on this? [22:23:51] meh [22:24:11] can we at least put all those tiny modules into one file? [22:24:11] In Soviet Russia seperate file move you! [22:24:43] MaxSem jdlrobson my preference is to have separate files for the modules, it's easier to find them later [22:24:59] Reedy: +1 [22:25:10] Reedy, you have a typo [22:28:08] [Commons-iOS] brion pushed 1 new commit to master: http://git.io/AoUUDg [22:28:08] Commons-iOS/master ea996d1 Brion Vibber: Bump version to 0.16 [22:28:22] awjr: found a nice way of doing the class thing using Max's setProperty [22:28:28] cool jdlrobson [22:28:37] not using the hook? [22:29:07] the hook could be nice because it will allow amendements by other invocations of the hook [22:30:42] awjr: negative. We can do it in the skin itself [22:30:48] fancy [22:30:55] i prefer this way as the special page itself gets to add the class [22:31:00] which is much better for readability [22:31:17] skin has a handy function addToBodyAttributes which expects to be overriden [22:33:05] [android-commons] yuvipanda pushed 1 new commit to master: http://git.io/wkY9-Q [22:33:05] android-commons/master 1d1e7f9 YuviPanda: Add Categorization EventLogging [22:33:07] New patchset: Jdlrobson; "Use OutputPage headElement function to generate head element" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56875 [22:33:40] ^ awjr [22:33:51] MaxSem: http://urbanfunscape.tumblr.com/post/46154456922/the-pop-up-city-is-always-interested-in-mobile#.UVoLOFtUOp1 you are exciting people [22:33:52] ragesoss: okay, updated to latest version, with EventLogging [22:33:56] ragesoss: try it out later and let me know [22:33:57] ragesoss: < [22:33:59] ragesoss: <3 [22:34:02] good night everyone [22:34:24] :0 [22:35:59] ooohhhhhhhh Maryana i've found the greatest article to add a photo to https://en.m.wikipedia.org/wiki/Mobile_phone_throwing [22:36:14] hahaha [22:36:22] CAN WE CAN WE [22:39:00] Men's world record - 102.68 meters [22:39:01] :| [22:39:02] brion, hmm just downloaded the new version of the iOS app - I can see my file names, but all the pictures are showing up as black [22:39:14] I reckon I can beat that record Reedy [22:39:16] ohh and Maryana a fight for you - http://dtsonev.tumblr.com/post/46585612105/wikipedia-now-lets-you-upload-photos-from-mobile#_=_ [22:40:04] hmm [22:40:18] Thehelpfulone: does it clear up if you wait a while or scroll around the list? [22:40:34] also double-check that your photos didn't get deleted ;) [22:40:40] nope, and clicking on individual images doesn't do anything either [22:40:45] oh good point [22:40:46] * Thehelpfulone checks [22:41:14] MaxSem im having trouble testing https://gerrit.wikimedia.org/r/#/c/55226/9 - i can't seem to get device-specific resources to load even though im spoofing my UA [22:41:22] nope, they're still there [22:41:23] oh MaxSem heads up.. i'm planning to experiment with turning off special casing the homepage on alpha within the next 2 weeks [22:41:26] should I delete the app and then install it again? [22:41:47] Thehelpfulone: try that… but there *may* be a bug where when you do that it shows none of your photos initially [22:42:00] jdlrobson, have you seen the main page on wikivoyage? [22:42:02] if you hit that, go into settings and toggle debug mode on and off to work around it [22:42:15] and it will be even worse [22:42:15] MaxSem: have you been able to get the correct headers to be returned for device-specific styles with that patchset? [22:42:21] dr0ptp4kt: i'm struggling to remember your handle / gerrit email :) [22:42:32] MaxSem: I'm ready for it [22:42:34] :P [22:42:39] confirming that none of the photos initially bug [22:42:40] RIP [22:42:47] jdlrobson: dr0ptp4kt [22:43:16] yep i know now but only because you were in the room :) [22:43:32] dr0ptp4kt: i added you to the gerrit patch for the zero bug [22:43:39] MaxSem: remember the only way to drive change is to break things XD [22:44:04] brion, okay so now I've got debug mode off - and one of my images on commons was deleted, but it still shows up in my upload list as a black image - is that intentional? [22:44:38] Thehelpfulone: that's a known issue at present. we intend to either remove them from the list or show an icon that indicates it was deleted [22:44:38] also brion did you see that thing about your meta adminship/cratship inactivity removal - do you still need it? [22:44:55] Thehelpfulone: yeah that's cool, i have all the permissions i need behind the scenes if i need to fix anything :D [22:45:24] thanks, I'll remove them - and yeah if we're thinking power users most of their images shouldn't be deleted so they might like to know if an image of theirs is deleted [22:45:32] and a newbie would definitely want to know if an image was deleted [22:46:07] yeah, a clearer indication might not make the first store release but it's a necessary fix at some point soon [22:46:17] New review: Jdlrobson; "(1 comment)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56887 [22:46:22] ^ awjr see that comment [22:46:41] ok jdlrobson, im in the middle of testing another patchset, i'll look soon tho [22:46:45] sure thing [22:46:47] head one [22:46:48] s [22:46:56] *head element one is the crucial one from my perspective :) [22:50:35] jdlrobson: awjr MaxSem how goes to code reviews? [22:50:45] we still looking good for 5PST? [22:50:47] it's goin, jcmish [22:50:50] hehe [22:51:15] i *think* so - is there anything unreviewed that *must* go out tomorrow still? [22:51:16] New patchset: Jdlrobson; "Rename #mw-mf-search to #searchInput" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56324 [22:51:16] New patchset: Jdlrobson; "Bring watch star css/html markup closer to desktop" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56515 [22:52:49] awjr: I'm looking through it and it seems like we're in good shape [22:52:55] but I thought I better doublecheck [22:57:54] awjr: I really think https://gerrit.wikimedia.org/r/#/c/56875/ should go out tomorrow as i'd hate to bundle all our RL changes out in one go [22:59:19] +1 [22:59:42] makes sense [23:03:12] k jdlrobson let me know if it doesn't make it in by 5 [23:03:23] roger [23:04:10] this is one of the rare occasions where i wouldn't mind a botched deployment tomorrow if it comes to that as it will highlight issues that need fixing! [23:06:09] HAHHAHAHAH [23:06:20] bite your tongue! [23:06:57] dammit, after jdlrobson's RL rewrite I don't see mobile.device.detect in page html even though that brnch gets executed [23:07:18] MaxSem: which RL rewrite? :D [23:07:37] the first one [23:07:41] $out->addModuleStyles( $device->moduleName() ); ? [23:08:21] MaxSem: you talking about https://gerrit.wikimedia.org/r/#/c/55226/ ? [23:08:33] $out->addModuleStyles( 'mobile.device.detect' ); [23:08:55] MaxSem: you'll need to add targets [23:09:09] class MobileFrontendDeviceDetectModule extends ResourceLoaderFileModule { $targets = 'mobile' [23:09:25] protected $targets = array( 'mobile' ); [23:09:51] actually it inherits [23:09:51] MaxSem: that might explain why i cant get this to work.. [23:09:57] * jdlrobson still having problems navigating the code [23:11:07] $wgMFMobileResourceBoilerplate sets targets to mobile [23:11:07] frak [23:11:13] fixed [23:11:19] \o/ [23:12:28] New patchset: MaxSem; "WIP: [Card 446] Don't vary page HTML on X-Device" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/55226 [23:13:25] MaxSem that looks a lot better [23:13:57] im getting proper variance now on the style :) [23:15:26] New review: awjrichards; "(1 comment)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/55226 [23:16:43] so jdlrobson does RL now load the robots policy automagically? [23:17:21] awjr: yes my comment explained it [23:17:28] no? [23:17:30] oh, didnt see your comment [23:18:20] arrgg it didn't publis [23:18:24] * jdlrobson shakes fit at gerrit [23:18:26] New review: Jdlrobson; "(1 comment)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56875 [23:18:26] *fist [23:18:30] ^ [23:18:45] https://gerrit.wikimedia.org/r/#/c/56875/4/includes/skins/SkinMobile.php [23:18:47] specifically awjr [23:18:53] thnx jdlrobson [23:20:22] hey jdlrobson, how come [23:20:30] Maryana: ? [23:20:36] so if we get these RL changes out tomorrow, i think we're gonna need a mobile varnish cache flush, right MaxSem? [23:20:40] i would suspect so awjr [23:20:43] big html changes [23:20:55] ah, sorry, hit return too soon :) how come there's no metadata on some mobile web uploads? [23:20:57] http://commons.wikimedia.org/wiki/File:A_double_rainbow_in_San_Francisco_2013-03-30_21-33.jpg [23:21:31] Maryana: *shrug* could be a browser or phone camera thing [23:21:40] are these HTML changes breaking? [23:21:47] * MaxSem is not so sure [23:22:07] * awjr shrugs [23:22:14] * Maryana opens bug [23:22:53] MaxSem: i think we should figure it out for sure, we'll need to coordinate with ops if we need a flush [23:29:35] New patchset: Jdlrobson; "Bug 45622 - Allow invoking of SkinAfterBottomScripts" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/57009 [23:33:23] jdlrobson, MaxSem: https://gist.github.com/awjrichards/16ce38383ec8ee2994ac [23:33:40] that's a diff of HTML output between current master and https://gerrit.wikimedia.org/r/#/c/56875/5 [23:33:45] ergg why is this bug https://bugzilla.wikimedia.org/show_bug.cgi?id=43628 back under MFE again [23:34:15] awjr: ? o_O [23:34:27] awjr: https://gist.github.com/jdlrobson/5281620 [23:34:31] that should be the diff.. [23:34:41] jdlrobson: that is not html output [23:34:48] oohhh [23:35:02] jdlrobson: im checking to see whether or not we'll need a mobile cache flush with that changeset [23:35:15] awjr: is this on beta labs [23:35:19] jdlrobson: no locally [23:35:21] note that the setup will be different there [23:35:31] other extensions might be adding tags/head items [23:35:53] in a way that would be different between what's currently in master and your changeset? [23:36:12] yes [23:36:28] if that's the case then i suspect there would be further differences in production [23:37:11] but jdlrobson given the diff i sent, is there anything concerning? more to the point, if old HTML was served, would it cause problems? [23:37:31] so if i'm reading the diff correctly just 4 new lines have been removed? [23:37:54] sorry having issue reading this diff [23:38:04] yeah it's wonky - but there are other differences [23:38:06] k got a better rendering [23:38:55] ohh actually yes [23:39:01] we definitely need a flush [23:39:08] document.documentElement.className.replace( 'page-loading', '' ); [23:39:14] otherwise the site will appear disabled to EVERYONE [23:39:14] MaxSem: ^ [23:39:19] @_@ [23:39:27] (well everyone from cache) [23:39:35] Maryana: hey, I hope https://bugzilla.wikimedia.org/show_bug.cgi?id=46765 is not a dupe [ missing string in "uploaded via Mobile Web ()." ] [23:39:42] if that line doesn't run all chaos will break loose [23:39:46] just wanted to give you a heads-up, though it's not urgent [23:40:29] thanks sumanah - i can't remember if we filed a bug for it or not :) it's not urgent but definitely annoying :) [23:40:44] out of curiosity, what's supposed to be in there? [23:40:51] ok, if we're going to need a cache flush to get this out, we may need to wait to deploy this [23:41:07] MaxSem: still there? [23:41:12] yup [23:41:30] sumanah - yuvi made the template specifically for apps, which come with a version number. that's what that param was for [23:41:37] MaxSem: so it sounds like we'll definitely need a cache flush if we push that change set out [23:41:37] oh, thanks [23:41:42] i just commented :) [23:41:43] then we just sorta ganked it for mobile web, but mobile web doesn't have versions :) [23:41:53] MaxSem: what do you think is the best way to proceed? [23:42:14] jdlrobson: is there a chance we can rework this so we wouldn't need a cache flush? [23:42:19] awjr: no [23:42:24] 56875? [23:42:31] why are we so scared to flush [23:42:31] yes MaxSem [23:42:31] (well, technically there are versions, https://commons.wikimedia.org/wiki/Special:Version says MobileFrontend (Version 0.7.0)) [23:42:37] jdlrobson: quite a number of reasons, actually [23:42:49] jdlrobson, we need to forget about flushes [23:42:50] & MW 1.21wmf12 (9eafdeb) [23:42:52] we're making big changes to the html we are going to have to flush [23:42:54] i suppose that's true, although version info is less salient for us than for the apps team [23:42:59] it's an emergency measure [23:43:08] Maryana: got it [23:43:10] MaxSem: sure on the long term [23:43:23] I'll just add that to the bug Maryana just for reference [23:43:24] if we're aiming at flish, we're screwing up [23:43:29] but i don't think we can avoid it in meantime [23:43:37] k, thanks, sumanah [23:43:59] anyway, we will not deploy 56875 tomorrow so no hurry [23:44:02] MaxSem: well if not varying HTML by X-Device is the goal.. [23:44:12] * jdlrobson facepalm [23:44:38] 1) flushing is more dangerous now that it was say 6 months ago and to do safely requires different procedures (doing one box a time, in a very controlled manner) 2) there are currently issues with some services (namely job queue) on the cluster, which ops and platform are trying to sort out, but i presume make deployments this week have extra performance considerations 3) if we *really* get in a situation where we need a cache flush for pushing [23:44:56] awjr: well at some point we are going to need to flush [23:45:03] if not this week next week [23:45:12] jdlrobson: or find a way to do this without requiring a flush [23:45:49] if that's really not an option, then we need to start planning with ops [23:45:51] awjr: why can't we stagger flush [23:45:52] w [23:46:23] we can, but i can almost guarantee they wont want to do it this week because of other things currently going on on the cluster [23:46:52] after further thinking, we set page-loading with inline JS, then the rest of JS doesn't get loaded because it's blacklisted - what's next? [23:46:56] and like i said, we'll need to coordinate with ops well in advance - they expressed concerns about even doing staggered flushes [23:47:10] MaxSem: we also remove page-loading with inline js [23:47:11] https://gerrit.wikimedia.org/r/#/c/56875/5/includes/skins/SkinMobile.php,unified [23:47:29] ugh [23:47:37] well this is one way we can avoid this on the short term [23:47:49] undo https://gerrit.wikimedia.org/r/#/c/56875/5/javascripts/common/mf-application.js [23:48:01] that way it will get unset at some point [23:48:29] if I put that back can we merge https://gerrit.wikimedia.org/r/#/c/56875/5 in ? [23:48:42] New review: awjrichards; "This looks good and seems to work - however this alters output HTML in a significant enough way that..." [mediawiki/extensions/MobileFrontend] (master); V: 2 C: -2; - https://gerrit.wikimedia.org/r/56875 [23:48:46] nooooooo [23:48:46] i *really* *really* would like to see us merge this otherwise tomorrow's deployment seems pointless [23:49:09] i don't think pushing the other changes we have out is pointless... [23:49:21] well what are we pushing out.. anything of interest? [23:49:25] and clearly this needs a little more thought and potentially coordination [23:49:29] considering deployment wastes a whole day from my perspective [23:51:23] New patchset: Jdlrobson; "Use OutputPage headElement function to generate head element" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56875 [23:51:53] there are quite a few patchsets otherwise to go out, plus tomorrow we're doing deployment training: https://gist.github.com/awjrichards/ec9be8d7f8b595e26232 [23:52:07] so… again im not sure how deploying without that patchset is 'pointless' [23:52:12] New review: Jdlrobson; "The above patchset will prevent the need for a varnish change. Keeping the line in mf-application.js..." [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/56875 [23:53:45] as the matter of fact, my revision can also be deployed cause the new functionality is protected by a config toggle [23:54:01] jdlrobson: can you point me to the specific difference for that patchset? [23:54:06] I forgot about first time user and upload going to stable - so i guess those are useful [23:54:21] https://gerrit.wikimedia.org/r/#/c/56875/6/javascripts/common/mf-application.js [23:54:27] compare with https://gerrit.wikimedia.org/r/#/c/56875/5/javascripts/common/mf-application.js [23:55:08] MaxSem: awjr do we have a flush on standby? [23:55:20] or will we play it by ear tomorrow? [23:55:38] we will avoid it [23:55:45] I truly believe however that we are going to need flushes of some form during this transitional phase and that they cannot be avoided [23:55:50] jcmish: we can't do a cache flush tomorrow [23:56:00] oh? [23:56:09] jdlrobson: if that's the case then we need to coordinate with ops in advance [23:56:18] we will [23:56:42] we'll need mark or binasher at that deployment anyway [23:59:27] jdlrobson: im still seeing that JS output from SkinMobile.php in the latest patchset [23:59:36] what JS output awjr ? [23:59:54]