[09:48:52] MaxSem: what exactly is 'override' in the action=mobileview params? [09:54:26] yuvipanda, ask preilly with his hacks [09:56:26] hmmm hacks [09:56:58] MaxSem: It gets you a...token? [09:57:03] I'm confused [09:57:14] ignore it [09:57:26] or, better, press him to remove it;) [09:58:06] I'm possibly very confused. [09:58:07] sigh [09:58:40] hmm, so this was the hack for retreiving a token [09:58:41] ok [09:59:15] * yuvipanda makes a note to ask preilly [10:29:50] hmm, 'invalidtitle' is actually 'page not found' [10:29:51] ok [10:39:44] hmm, the app is only around 3000 lines of code. [10:39:46] not bad :) [10:40:56] are you testing not on labs? [10:41:21] cuz I've swapped invalidtitle with missingtitle now [10:41:29] should make a bit more sense [10:46:48] MaxSem: I was looking at a checkout of MobileFrontend. Was a bit antiquated [12:09:26] MaxSem: http://mobile-geo.wmflabs.org/ is down? [12:17:27] MaxSem: definitely down [12:46:32] rebooted [14:34:01] MaxSem: thanks. Main page transforms work :) [14:35:01] yuvipanda, next time you can just reboot the instance yourself. labs is too unstable:( [14:37:09] hmm, I've access but haven't really used it at all [14:37:14] MaxSem: will do. Thanks! [15:39:06] hello jdlrobson [15:39:11] hey YuviPanda [15:39:17] you've given me quite the backlog ;-) [15:39:26] :D [15:39:34] jdlrobson: many of them are quite small [15:40:26] i hope so for your sake ;-) [15:40:30] i've got a new hat for you today btw [15:40:39] and internet connection should be __lots__ better [15:41:33] ah [15:41:38] jdlrobson: mine should be a lot worse [15:41:47] i'll very probably be black rectangle today [15:43:58] :( [15:44:24] soon you'll be //that guy// who looks like a black rectangle [15:44:46] some day, it'll be a standup of 5 black rectangles [15:54:59] jdlrobson: when do you think you can look at the pull reqs? [15:55:04] I'm hoping to push out a beta today... [15:55:08] mm [15:55:09] to the Google Play store [15:55:13] i was hoping tomorrow, but we'll see [15:55:16] :( [15:55:34] i'll try my best to get them done today [15:55:50] i need to review the current situation with mobile site as there is a deployment today and i want to fix a lot of the screw ups I made last week [15:55:57] (relating to older browsers) [15:56:38] jdlrobson: ok :) [16:07:23] k YuviPanda reviewed them all and made some comments will pull and test them later when hopefully they are taken care of :) [16:10:47] jdlrobson: replied to / fixed most of them :) [16:18:27] jdlrobson: hi [16:21:40] jdlrobson: hi [16:25:54] yuvipanda: sorry i dropped my laptop so i didn't catch your last message [16:29:56] jdlrobson: hey Jon [16:30:00] hey philinje [16:30:07] how wa your weekend? [16:30:11] was [16:32:34] very good [16:32:43] lots of nature - relaxing [16:32:57] my mind is clear and feeling very productive today :) [16:34:26] * yuvipanda has a cold and is reasonably drowsy [16:34:40] jdlrobson: just replied saying fixed/replied to most of them [16:34:58] thanks yuvipanda - will merge them before I get started on anything else [16:35:06] hopefully will be done in next 2 hours [16:35:34] woo [16:46:40] greetings all [16:46:55] yuvipanda: iPhone3G came in [16:48:29] wooo [16:48:36] now just need to beam it here [16:48:43] hang out is up and running https://plus.google.com/hangouts/_/28a2c7a297ce7c76ea75cdb254b8865099462031?authuser=1&hl=en# [16:49:21] i'm staying pretty good to my word of having this up and running 10min before the actual standup [16:49:39] heh [16:49:47] jdlrobson: nice to hear, glad you had a good one. Sounds like I should bug you in a couple hours [16:50:09] jdlrobson: i surfaced a nasty bug in opera mini that we should talk about [16:50:38] excellent! (not that there is a nasty bug but that you found something and we are going to talk about it) [16:50:42] currently debugging on a nokia n95 [16:50:44] lots of fun [16:50:46] (another nasty bug) [16:51:14] are there any other type when you're dealing with client side mobile browsers? [16:51:27] BlackRectangle: seriously [16:51:47] jdlrobson: this one is really bad though as it brakes search on opera mini [16:51:53] breaks* [16:52:04] hopefully its the same one I'm looking at now then as search is also broken on nokia n95 :) [16:52:23] all hail the black screen that is yuvi on hangout [16:52:33] hey awjr [16:52:35] i should have better wifi today [16:52:36] \o/ [16:52:36] tfinc_: you're on mute [16:52:43] hi tfinc_+ [16:52:45] er [16:52:46] tfinc_ [16:53:01] tfinc_: ( 6.33%) unspecified Opera Mini - a substantial percentage of traffic [16:54:14] awjr: https://plus.google.com/hangouts/_/28a2c7a297ce7c76ea75cdb254b8865099462031?authuser=1&hl=en# [16:54:23] this saves even more bandwidth! [16:54:41] lolz [16:54:50] jdlrobson: how many hats do you have anyway? [16:55:12] currently 4, only 2 regulars though [16:55:17] ah [16:55:24] hi mobile team, can you give me an example user agent for the 1.0.3 version of the app? [16:55:29] the other 2 are being phased out [16:55:32] ;-) [16:55:55] jdlrobson: deprecated, you mean? :) [16:56:01] exactly [16:56:15] one seems to make me look a bit schoolboyish these days - think it might need starching [16:56:27] tfinc_: 36490 search is probably broken in more devices / older browsers [16:56:32] drdee_: WikipediaMobile/1.0.3 [16:56:38] srikanthlogic: looking into it now :) [16:56:45] managed to replicate on a nokia n95 [16:56:56] jdlrobson: great! [16:57:06] * BlackRectangle has zero hats. Should change before berlin [16:57:08] hoping to get it fixed for todays deployment [16:57:12] thanks BlackRectangle! [16:57:19] It might be summer there, but am pretty sure I'll be under multiple layers of clothing again [16:57:41] MaxSem: awjr : did you get the hangout link ? [16:57:42] BlackRectangle: does that apply to both the iOS and Android versions? [16:58:12] drdee_: yes, though iOS is 3.x+ [16:58:23] * BlackRectangle waves at MaxSem [16:58:29] just missing awjr now [16:58:41] where are you guys? [16:58:49] preilly wont be joing us today as hes not feeling well [16:58:56] linSmith: next to r34/r35 [16:58:59] BlackRectangle many thanks! [17:05:10] New review: awjrichards; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6753 [17:05:12] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6753 [17:05:16] BlackRectangle is a cheater! [17:06:09] philinje: https://bugzilla.wikimedia.org/show_bug.cgi?id=36017 [17:06:28] jdlrobson: is that tea? [17:06:34] MaxSem: how what why/ [17:06:34] ? [17:06:54] yup [17:06:56] you've seen us while we haven't seen you [17:06:59] very english [17:07:27] philinje: was this the page you were just talking about? http://www.mediawiki.org/wiki/MobileFrontend/Photo_upload i actually did look at this and went through the various links last week [17:07:31] MaxSem: Choice is between you seeing me or me hearing you everyone :P [17:07:50] justifications, justifications [17:08:01] MaxSem: I might be hopping to another city for a week or so in another week or so. Maybe I'll have working wifi there! [17:08:29] MaxSem: my current city has 2 hours of scheduled power cut a day. Outskirts have up to 9 hours of scheduled power cuts. Wifi might be a bit too much to ask for :P [17:09:31] :D [17:11:08] * tfinc boots up the new iPhone3G test phone [17:11:29] * tfinc is amused/annoyed by the "Do not update software - Phone will relock" [17:11:54] tfinc, philinje: when you'll be able to chat about PageImages? [17:12:32] tfinc: thoughts on how long it will take to get here? [17:13:09] rachel will coordiante that. lets see if it beats us meeting in berlin [17:13:10] ;) [17:13:28] tfinc: when do we want to setup intercontinental debugging? [17:13:39] * tfinc looks at calendar for today [17:13:52] 11-12 [17:14:10] jdlrobson: should we meet later today? [17:14:27] should be fine philinje - just depends how long this search problem takes to get fixed [17:14:44] ok let us know [17:15:04] MaxSem: can you give me a little background PageImages? [17:16:56] philinje, it does 3 things: associates images with pages to 1) make this information available via the API to match Geonames 2) replace lame thumbnalis of OpenSearchXml 3) search for pages with/without images [17:17:13] so 3) especially concerns you [17:18:30] shweet [17:18:33] * YuviPanda goes afk for a bit [17:18:37] tfinc: will poke you at 11AM PDT [17:20:24] YuviPanda: sure [17:20:36] MaxSem: did you wind up doing the deployment with Patrick last week? [17:21:11] awjr, no I fell aslep earlier [17:21:19] MaxSem: sounds awesome [17:22:24] YuviPanda: since this a new phone i'll have to do a new build. what should i build form ? [17:22:25] from* [17:22:26] MaxSem: was the original motivation related to photo uploads, because it will help with that [17:22:39] philinje, tfinc, the question is about sufficiency of functionality [17:23:12] philinje, it was taken into account, yes [17:24:30] MaxSem: new project page in progress: https://www.mediawiki.org/wiki/MobileFrontend/Photo_upload [17:24:57] MaxSem: this might be useful: https://www.mediawiki.org/wiki/MobileFrontend/Photo_upload/Upload_Wizard_mobile_port [17:25:32] philinje, tfinc, I have the fundamental question of whether we will need finer-grained search capabilities. we discussed this stuff with you Phil, options being search for pages with less than some number of images or some other criteria [17:25:52] ah yes [17:26:13] to set a parameter for the number of images threshold [17:26:44] search by number of images is problematic because there is no set number of "enough" images per page [17:27:32] yes, that's where we left it last time, but perhaps just allow a parameter to be modified by experimentation [17:27:54] another thought: position of images in the page (or the lack thereof) [17:28:00] I tried experimenting with more complex scoring system, the outcome was some system that has the same problems as search by number of images, but with criterias less clear to mortal humans [17:28:57] we will have a somewhat different issue with Wiki Loves Monuments: [17:29:07] philinje, I actully use scoring for determining which image should be associated with a page, and scores take into account the position of image on page [17:29:16] i see [17:29:40] is the scoring system on w wiki page somewhere? [17:29:47] no [17:30:09] it's not something for end-users [17:30:16] could you create a page under Photo upload and put it there? [17:30:35] yes, some pages are not for end users but for developers [17:30:59] there's extension's page for such things (and I should really add more information to it besides the infobox) [17:31:18] ok, let me know where that is [17:31:34] awjr: how is our deployment looking for today? [17:31:45] tfinc: good question :) [17:31:49] awjr: let me know when the wiki page is up to date and i'll good through the changes [17:31:52] once i get through my mail backlog i'll let you know [17:31:58] awjr: k [17:32:05] tfinc: i have no idea how patrick was formatting the pages to look all pretty [17:32:09] MaxSem: the WLM issue is: number of photos uploaded per monument [17:32:12] do you? [17:32:28] tfinc, philinje, so the question is: do we need to search for something more than presence/absence of images on page? [17:32:53] philinje, this is tangential to this extension's purpose [17:33:09] not all photos uploaded are supposed to be on any given page [17:33:27] yes, though the WLM case may or may not be related to an article - the issue is monuments [17:34:59] it's like another type of article [17:35:10] so WLM will need a different image accounting system [17:35:40] my question is about the PageImages scope [17:36:22] MaxSem: they already have one - uploads are categorized by monument, then eventually the winners are chosen for the monument articles [17:36:58] MaxSem: ok, could you explain about the lame thumbnails? [17:37:30] tfinc: back [17:37:37] tfinc: gimme a second, will make you a branch. [17:37:45] tfinc: master's a bit behind, since there are 10 open pull reqs [17:37:52] MaxSem: happy to talk on Skype [17:38:12] philinje, sure [17:39:06] ok one sec [17:40:16] moving to a meeting room [17:40:24] YuviPanda: k [17:47:43] awjr: MaxSem : over the weekend we had some nasty twitter traffic around vandalism on the cinco de mayo article. it was quickly corrected on the desktop but it took forever for mobile to pick it up. what's the typical delay we should exect ? [17:48:36] YuviPanda: just fixing nasty search bugs then I will go pull crazy [17:48:58] jdlrobson: ok :) [17:49:09] * tfinc wonders what pull crazy jdlrobson looks like [17:49:31] tfinc: you'll basically see nothing in irc for a few mins ;-) [17:49:56] does pull crazy jdlrobson come with a hat? i'm imagingin an action figure here [17:49:59] so basically the search problem seems to effect browsers that don't support texContent [17:50:07] * textContent [17:50:55] btw tfinc I can't make standup friday due to my visa interview \o/ [17:51:23] bingo [17:51:28] fixed search on nokia n95 [17:51:30] tfinc: im not sure - looking at the response headers from a mobile request, the s-maxage for cache-control is 300s [17:51:34] jdlrobson:k, just mail out like normal [17:52:00] awjr: so in this case it was certainly hours. if you guys are sure that it should be quick then i'll have to experiment and see if i can reproduce it [17:52:04] this was certainly hours [17:52:10] tfinc: i am definitely NOT sure [17:52:32] tfinc: i'll have to ask asher/preilly [17:53:10] awjr: k, just find out the frequency. i'm still not sure about the extent of the problem [17:53:15] but it led to comments like this https://twitter.com/#!/gorelickingood/status/198967986353213440 [17:53:27] err [17:53:32] tfinc: pull from intercontinental-debugging-may-7-2012 on my repo? [17:53:41] this one shows it https://twitter.com/#!/meglanker/status/198968645819432960 [17:55:20] tfinc: i'll get the info from asher as soon as i see him online [17:55:25] awjr: thanks [17:55:28] np [17:56:38] tfinc is there a bug open about this? [17:56:52] awjr: i'll open one as soon as you tell me what the cahce settings should be [18:01:36] * YuviPanda pokes tfinc [18:05:27] tfinc: here's what we have so far for today's deployment: http://www.mediawiki.org/wiki/Extension:MobileFrontend/Deployments/2012-05-07 [18:05:31] * tfinc punches git for not pulling from remote branches [18:05:41] awjr: thanks [18:07:51] New patchset: Jdlrobson; "support browsers that do not support textContent (add get text function)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6803 [18:07:52] New patchset: Jdlrobson; "make sure nokia rule is important" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6804 [18:07:53] New patchset: Jdlrobson; "jslint codebase" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6805 [18:07:54] New patchset: Jdlrobson; "move updateQueryStringParameter into settings.js" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6806 [18:07:55] New patchset: Jdlrobson; "get rid of default list styling on search results" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6807 [18:08:11] philinje: should i try to fix the email address configurations for the feedback form today when i deploy, so that only the english addresses have the 'q' or 'c' suffix? [18:08:19] in git do i have to do anything special to update my remote branches? .. besides a pull [18:08:43] awjr: if ossible, that would be great [18:08:48] tfinc: pull should work, but I like doing fetch + checkout [18:08:56] doh labs is down again [18:09:01] git remote add yuvipanda [18:09:03] git fetch YuviPanda [18:09:17] YuviPanda: but for some reason git branch -a isn't showing your new branch [18:09:20] i see all of your old ones [18:09:36] i already have you as a remote [18:09:40] so tfinc git fetch https://gerrit.wikimedia.org/r/mediawiki/extensions/MobileFrontend refs/changes/07/6807/1 && git checkout FETCH_HEAD should hopefully fix opera mini search - it fixes s60 (cc @srikanthlogic ) - if labs comes back I'll stick it on there for testing purposes [18:09:46] * jdlrobson gets ready to go pull crazy [18:10:18] hmmm [18:10:31] tfinc: can't checkout yuvipanda/ [18:10:31] ? [18:11:25] nope .. even though my local checkout thinks its up to date [18:11:46] tfinc: did a git fetch? [18:12:08] ahh i see why now [18:12:13] intercontinental-debugging-may-7-2012 new (next fetch will store in remotes/yuvi) [18:12:25] that explains it [18:12:29] :) [18:12:42] hmm, I don't remember seeing that though [18:13:39] New review: awjrichards; "Aside from trailing whitespace, looks good to me." [mediawiki/extensions/MobileFrontend] (master); V: 1 C: -1; - https://gerrit.wikimedia.org/r/6803 [18:14:32] New review: awjrichards; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6804 [18:15:11] * YuviPanda pokes philinje about [18:15:11] ~f@wikipedia/Ironholds (freenode) [18:15:16] err [18:15:20] bad copy paste bad [18:15:36] * YuviPanda pokes pchang about https://bugzilla.wikimedia.org/show_bug.cgi?id=36017 [18:15:41] * YuviPanda pokes philinje ^ [18:15:41] New review: awjrichards; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6805 [18:16:54] awjr: have you ckedked in the email change already? [18:17:00] git is still not behaving [18:17:21] philinje: no, it's something i'll do during deployment. it's just a configuration change. [18:17:38] ohh .. i think i have it [18:17:54] Switched to a new branch 'intercontinental-debugging-may-7-2012' :D [18:18:13] git fetch yuvi FTW! [18:18:21] i see - Tomasz was wondering if we should bundle up changes for next monday, but if it's part of the deployment, then sounds fine [18:18:26] binasher: what is the TTL for mobile pages? looking at page response headers it seems like it ought to be 300s but it sounds like in reality it's more on the scale of hours [18:18:34] yes, for some reason 'fetch ' always is funny :) [18:18:45] awjr: that would leave one small text change for next Monday [18:19:01] yeah, i thought a 'git pull' would hit all origins but clearly it didn't [18:19:12] philinje: the email address generation logic happens at the configuration level - i can wait til next week to do it though [18:19:14] i had to explicitly say the remote [18:19:17] philinje: what's the one small text change? [18:19:24] ah, no - it doesn't [18:19:59] * tfinc goes to build a new ips [18:20:00] New review: awjrichards; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6806 [18:20:01] ipa* [18:21:00] New review: awjrichards; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6807 [18:21:50] New patchset: awjrichards; "support browsers that do not support textContent (add get text function)" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6803 [18:22:32] New review: awjrichards; "i went ahead and fixed the trailing ws as a commit amend." [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6803 [18:22:36] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6803 [18:23:01] YuviPanda: i have the app running now [18:23:07] tfinc: how's scrolling? [18:23:09] YuviPanda: where would you like me to focus testing [18:23:11] tfinc: plus what version of iOS is? [18:23:26] 4.2 [18:23:41] tfinc: so, two things: 1. scrolling on articles 2. scrolling after expansions / contractions of sections 3. performance during expansion / contraction of sections [18:23:49] YuviPanda: spin up an etherpad and i'll take notes there rather then cluttering up IRC [18:23:59] New review: Jdlrobson; "Thanks Arthur for the whitespace - I'm kicking myself for missing that!" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6803 [18:24:18] jdlrobson: no worries, i make WS mistakes all the time :p [18:24:27] tfinc: tfinc http://etherpad.wikimedia.org/intercontinental-debugging-may-07-2012 [18:24:34] jdlrobson: you need to rebase the outstanding commits before they can be merged [18:24:56] k [18:24:57] 1 s [18:26:14] tfinc: spinner not going away means something is very broken [18:26:18] tfinc: did you do submodule update? [18:26:32] nope. let me do that [18:27:00] * tfinc has no idea why he can't ever remember to do that [18:28:39] jdlrobson: how's your timing for a talk today? [18:28:41] YuviPanda: scrolling is super jittery [18:29:32] hey philinje - i probably need about an hour to merge yuvi's stuff so he can get a beta out [18:30:49] no problem [18:31:14] YuviPanda, do image pages look ok in the app? [18:31:56] MaxSem: I tested on labs, it seemed okay (I didn't error out) but need to test more. Fixing iOS issues now [18:31:59] MaxSem: why? [18:32:11] New patchset: Jdlrobson; "jslint codebase" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6805 [18:32:12] New patchset: Jdlrobson; "make sure nokia rule is important" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6804 [18:32:12] New patchset: Jdlrobson; "get rid of default list styling on search results" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6807 [18:32:13] New patchset: Jdlrobson; "move updateQueryStringParameter into settings.js" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6806 [18:32:16] YuviPanda, I was concerned about styling [18:32:23] awjr: did that do it? got myself confused [18:33:10] New review: awjrichards; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6804 [18:33:12] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6804 [18:33:40] New review: awjrichards; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6805 [18:33:42] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6805 [18:33:49] YuviPanda: my initial notes are in the etherpad [18:34:34] New review: awjrichards; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6806 [18:34:36] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6806 [18:34:49] New review: awjrichards; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/6807 [18:34:52] Change merged: awjrichards; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/6807 [18:34:53] jdlrobson: looks like it :) [18:34:55] tfinc: pull now and try? [18:34:58] sweet [18:35:22] jdlrobson: i've found it a lot easier to try and not make commits that are dependent on one another [18:35:25] * tfinc builds new ipa [18:35:35] sometimes it's unavoidable though [18:35:51] but if you havea bunch of dependent commits, sometimes it's easier to just squash them into one commit for review [18:36:07] i don't believe in squashing [18:36:10] I like small commits [18:36:32] you can still do small commits locally [18:36:55] but if I squash and do a git blame do i get the squashed commit on the smaller more specific commit? [18:37:03] * YuviPanda doesn't like squashing either [18:37:14] i don't mind rebasing so much - usually my pulls are small enough for it not to be an issue [18:37:16] YuviPanda: better [18:37:19] jdlrobson: no, you would get the squashed commit [18:37:26] jdlrobson right on [18:37:33] tfinc: how much better? [18:37:41] labs is still down? :( [18:37:59] YuviPanda: really good [18:38:32] tfinc: article expansion? [18:38:33] i need to load a long article and expand all its sections to really test this [18:38:37] err [18:38:40] err [18:38:43] s/article/section [18:38:44] section expansion, I mean [18:38:46] yeah [18:41:39] tfinc: pull again and try? [18:41:45] k [18:44:59] YuviPanda: that update made the app very unhappy [18:45:20] i *cant* scroll anymore [18:45:54] sigh [18:46:38] tfinc: pull now? [18:46:43] k [18:46:52] tfinc: scrolling should be okay and the 'can't scroll below a certain point sometimes' should be gone [18:50:06] YuviPanda: yes. both look good now [18:50:23] tfinc: test the 'can't scroll' a bit more? [18:50:26] tfinc: it's *very* intermittent [18:50:41] and *very* device specific [18:50:45] which sucks [18:51:07] sure [18:51:32] awjr: did you see philinje comments on http://www.mediawiki.org/w/index.php?title=Talk:Mobile_design/Contact&diff=0&oldid=519041 ? [18:52:02] jdlrobson: yeah, im going to fix the email address issue during deployment (it's a change i'll have to make in configuration) [18:52:07] sweet [18:53:04] philinje: is "Unless you agree to have your information made public, please do not include personal information in your comments."" final? [18:53:29] YuviPanda: i'm not seeing it happen again [18:53:49] tfinc: pull again? [18:54:03] k [18:54:04] awjr: cache ttls are set by mediawiki [18:54:14] hacks.css ;) [18:54:22] binasher: s-maxage, right? [18:54:25] yeaaah [18:54:32] yeah [18:54:46] YuviPanda: what can we do about the flickering that happens when you expand a section ? [18:54:51] tfinc: so this is all because of css hacks needed for inline styles [18:55:12] binasher according to tfinc pages are being cached for durations in the range of hours rather than minutes. [18:55:17] tfinc: good question. Let me see if I can repro that in the sim [18:55:27] binasher, even on enwiki which has a ttl of 300s [18:56:08] ? [18:56:31] Cache-control: s-maxage=2678400, must-revalidate, max-age=0 [18:56:35] that's the default from mediawiki [18:56:45] mobile is no different [18:57:17] YuviPanda: i'm going to grab lunch @ 12 ... should be back within 10-15 [18:57:30] binasher i am seeing Cache-Control:s-maxage=300, must-revalidate, max-age=0 on enwiki [18:58:04] you aren't going to see the apache headers from either varnish or squid [18:58:28] look at your test mediawiki instance [18:58:35] YuviPanda: what changed with this update? [18:58:59] tfinc: ok [18:59:04] tfinc: i got back hacks.css [18:59:10] tfinc: but killed a few styles [18:59:20] tfinc: we need to figure out *which* styles exactly are causing the problem and kill them [18:59:25] easier now since hacks.css is fast [18:59:26] err [18:59:27] smaller [19:00:09] k. as soon as the old iPhone3G finishes charging i'm going to hand the new one to our admin to send out to you [19:00:55] ok [19:00:57]