[00:31:12] awjr: why are you setting "$forceHttps = true;" [00:31:40] preilly: ? [00:31:44] oh for Userlogin? [00:31:48] https://www.mediawiki.org/wiki/Special:Code/MediaWiki/113202 [00:32:13] awjr: yeah, for that [00:32:20] preilly: that's essentially what it did before i made any changes to that code [00:32:28] awjr: oh, okay [00:32:33] i think... [00:32:47] awjr: well, I guess it won't really be an issue [00:32:53] awjr: so, it's probably just fine [00:33:01] awjr: give me 2min [00:33:15] it'll just ensure that Special:Userlogin is visited over https [00:33:35] awjr: I was just thinking of the case when we won't have ssl [00:33:37] preilly: that might need some attention though for non-wmf stuff [00:33:39] yeah [00:33:56] awjr: but, I think it's fine for now [00:33:56] tfinc_: no rush [00:34:36] preilly: k, we can add it to the list of de-WMF-ifying things to do [00:35:10] awjr: sounds good [00:37:22] awjr: dialing you now [00:37:43] i'm not seeing you on Skype so i'm ringing you on gtalk [00:38:17] awjr: i can do regular phone too [00:38:29] but i figure you have a head set [00:38:40] and that gtalk/skype would be easier [00:38:47] tfinc_: i do but it rang my phone and then behaved strangely... [00:38:53] one sec [00:39:01] so which do you want gtalk/skype/phone ? [00:39:14] im loading up skype i jsut need to make sure i have my headset configured right [00:39:18] k [00:53:21] preilly: what is the best way to track WAP traffic? [00:53:59] CIA. [00:57:13] tfinc: are you able to talk? [00:59:42] philinje: i will be after i finish my 1:1 with arthur [01:05:09] philinje: could you move our check in on friday to 2-3 ? [01:18:48] tfinc_: yes [01:19:36] can't modify your calendar [01:19:48] are you going to be a while [02:05:26] preilly: i just saw the response from Digi on the thread. [02:06:15] what can we speed up in our timeline to get it to production ? [02:10:39] preilly: .. ? [02:13:26] tfinc_: well, I want to talk to Ryan and Asher about it tomorrow [02:14:02] tfinc_: I think we can get it out as part of the Thursday push, but didn't want to commit just yet [02:15:18] binasher: do you have any initial thoughts on it ? [02:17:05] since we already figured out the varnish header mangling / acls in labs, it should be a pretty simple matter to port over to the cluster [02:17:40] hows your time tomorrow to take a look at it ? [02:18:00] i should be able to fit it in [02:18:31] k [02:18:41] i'm not sure if it will be testable until the thurs code push though [02:19:15] tfinc_: yeah, we'll need to wait until Thursday to fully test, but it will be partial testable tomorrow more than likely [03:51:53] http://speirs.org/blog/2012/3/6/we-need-to-talk-about-android.html [04:00:51] Very good article, Reedy. Thanks! [15:55:22] jdlrobson, do you imagine how much fun will be to implement your proposal without jQuery? :P [15:55:34] :) [15:56:43] you already know i enjoy pain from my attitude to pull requests in WikipediaMobile ;-) [15:57:16] to be fair... it could be a jquery only thing to start off with [17:34:19] jdlrobson, do you have any requirements for the new section retrieval API? [17:34:38] it's not restricted by action=parse limits [17:35:08] well.. i guess i would need to do the following 1) for a given article name request a summary which gives me the number of sections, the names of those sections and how to retrieve those sections and the summary of the article [17:35:22] 2) be able to request a given section via information given in 1) [17:35:36] all implemented [17:35:50] i don't think i'd need anything else.. [17:36:20] 1) api.php?action=mobileview&page=Foo&prop=text|sections§ion=0 [17:36:52] 2) api.php?action=mobileview&page=Foo&prop=text§ion=1 [17:37:39] what information about sections do you need, so that I could make it a default? [17:40:10] I'm not sure to be honest [17:40:18] what information is available? [17:42:10] heh, I receive this information from parser and isn't even sure about the exact meaning of some items [17:43:26] jdlrobson, example section information: http://dpaste.com/713115/ [17:44:21] * jdlrobson looks [17:45:50] mm [17:46:06] that is, "line" is section heading [17:46:44] "anchor" is section's id [17:47:26] many others are used for building a TOC [17:47:38] "byteoffset" is useless [17:47:56] surely i'd just need text when requesting a section? [17:49:33] do you want to differ between levels? [17:50:38] well... in my head i'm in control of h levels as an h is a section [17:51:03] okay [17:51:05] say I render

Section 1

and bind it to an ajax call to the api requesting section 1 [17:51:29] I guess I'd get back the text of Section 1 with its sub sections [17:52:37] a bit like blah blah

Section 1a

Section 1b

[17:52:58] I can then render sections as h3 and request subsections via the id etc etc [17:52:59] MF currently doesn't seem to differ between section levels, so

containing a couple of

s will result in 3 collapsed sections, all visible [17:53:54] you can request multiple sections with action=mobileview, by the way - unlike action=parse [17:54:08] MaxSem: the only links LinkBegin seems to have any effect on are footer links (like Privacy Policay, About $site, etc) [17:54:39] * MaxSem dives into parser [17:55:55] and that was the last time we saw a sane MaxSem [17:56:12] heyy [17:56:24] have you ever seen a sane me? [17:56:33] * MaxSem strongly doubts this [17:57:12] I doubt diving into our parser code can cause anyone to become saner [17:57:13] finishing up a call .. might be about 5min late for the stand up. feel free to start with out me [17:57:37] MaxSem: same thing for LinkEnd :( [17:57:58] hmm, Parser doesn't seem to call Linker::link() [17:58:05] how nice of it [17:58:18] i'm afraid to look - what does it do? [17:59:34] calls these hooks! ;) [18:03:42] ok .. done with my call.. lets start at 10:05 [18:05:20] hmm, always knew that parser is insane, but never knew how deliciously outrageously deep this insanity is [18:15:12] brainfarts_count++ [18:17:09] so preilly - if i check in minified js/css/ with a .min.js extension - we will need to update the ApplicationTemplate.php [18:17:27] to use some kind of flag [18:18:30] The only css changes I was thinking of making were the ones in https://bugzilla.wikimedia.org/show_bug.cgi?id=19921 - but I think that can wait till next week [18:18:45] awjr, I was wrong - Linker::link() still gets called for all links, you probably hit parser cache [18:19:00] parser cache! [18:19:06] * awjr shakes fist [18:19:23] thanks MaxSem - i'll dig a bit more. [18:19:55] basically, you need to make null edits to ensure parsing [18:20:17] yuvipanda, about those requirements [18:20:21] MaxSem: is that the only way to clear the parser cache/ [18:20:41] MaxSem: sections? [18:21:25] awjr, DELETE FROM objectcache; or restart Apache or restart memcached depending on your caching method [18:21:34] yuvipanda, yes [18:21:43] thanks MaxSem [18:22:17] my plans were that you initially do api.php?action=mobileview&page=Foo&prop=text|sections§ion=0 to retrieve first section HTML and section information [18:22:44] then just retrieve sections of user's choice with api.php?action=mobileview&page=Foo&prop=text§ion=1 [18:23:34] yuvipanda, the questions are: do you need anything else? what information about sections do you want? [18:23:52] hmm [18:24:12] MaxSem: what if one section refers to another? [18:24:18] also, did you figure out what to do about refs? [18:25:00] "refers"? [18:25:11] ah, #links [18:25:30] yes [18:26:17] you can retrieve section anchors so that [[#links]] don't get broken [18:26:18] jdlrobson: that is fine. [18:26:26] refs are funnier [18:27:29] preilly: so, it turns out the problems I've been having with keeping useformat=mobile sticky via MW hooks is not actually a problem. i wasn't seeing behavior i expected because of the parser cache. considering that, do you think it's better to keep the functionality using hooks or use a cookie? a cookie would certainly be more straightforward, but i think cookies tend to raise people's hackles even if they are as innocuous as just keeping you o [18:27:35] MaxSem: how funny? :) [18:27:56] MaxSem, jdlrobson, yuvipanda, tfinc - thoughts? ^ [18:28:07] we could retrieve zeroth section and references section on first request, but that would result in increased traffic and willdisplay a shitload of needless information for large articles [18:28:24] awjr: what hook are you using? [18:28:34] id use hooks if possible [18:28:35] I support cookies [18:28:52] preilly: where is mobile home page for en.WT? [18:28:57] with them, we could theoretically discard special mobile domains completely [18:29:27] MaxSem: true, retreiving refs for every req seems to beat the point of it [18:29:32] but is there another solution? [18:29:39] Amgine: it's just the main page [18:29:58] yuvipanda, hmm - hook up to clicks on references? [18:30:09] No, it's [[Wiktionary:Main Page]], just found it preilly [18:30:19] MaxSem: hmm, so first click on a ref gets all refs? [18:30:34] (they're weird there) [18:30:37] yes [18:30:46] preilly: LinkBegin [18:31:31] MaxSem: not possible to retrieve indiviudal refs? :D [18:31:40] actually no, just getting all refs on first click seems good [18:31:44] plsu we can display them in a usable way [18:31:48] unlike what we have [18:31:55] preilly: so basically append useformat=mobile to all links generated by Linker except for ones that probably shouldn't (eg when aciton=edit or something) [18:32:21] awjr: and not external links right? [18:32:46] preilly: correct, but perhaps with the exception of interwiki links [18:32:53] hmm [18:33:35] yeah, I think using Linker::link() makes a bit more sense then a cookie but we need to make sure that we vary the cache [18:34:50] cache fragmentation isn't nice [18:35:33] MaxSem: well, I don't think we want to cache use format links and regular links in the same cache pool [18:36:20] preilly: don't we get that for free just by altering the URL to have useformat=mobile? [18:37:16] preilly, I objected against usage of different parser output for mobile and non-mobile site at all [18:37:53] awjr: yeah, that's probably true [18:38:07] MaxSem: okay [18:38:53] thus I'm for cookies [18:39:08] awjr: what is the performance impact of LinkBegin [18:39:32] preilly: good question [18:39:48] MaxSem: actually i think using cookies would make sane cache handling a little more challenging [18:40:10] awjr: it varies on cookies already [18:40:22] oh well nevermind then [18:40:41] preilly: all cookies? [18:40:54] awjr: can you use xhprof or xdebug to profile the addition of the hook [18:41:16] awjr: $wgOut->addVaryHeader( 'Cookie' ); [18:41:55] preilly: sure - i've not really used either before but i've poked xdebug once upon a time so i'll give that a go. first i have to rewrite everything i unwrote when i thought the hook wasn't going to work. i wish we used git already [18:42:17] also, should we code freeze at some point today? [18:42:58] that sounds good to me. what have you done in the past? [18:43:26] awjr: never needed to in the past [18:43:45] oh right [18:43:56] awjr: do you want to try to get the use format persistence in before we freeze? [18:44:04] i'd like to, now that i know what im doing [18:44:08] it shouldn't take me too long [18:44:11] awjr: okay, cool [18:44:23] although i could probably use some help with profiling once i have the code rewritten [18:44:30] so, why don't we plan on locking things down at 1 pm [18:44:35] sounds good. [18:44:55] yuvipanda, so I can just tell you which section contains references? [18:45:00] MaxSem, jdlrobson, yuvipanda: are you guys ok with a MobileFrontend code freeze at 1pm PST today? [18:45:01] jdlrobson: can you get the minified versions checked in and modify the application template? [18:45:14] * jdlrobson does timezone maths [18:45:22] MaxSem: which sections contain all the references? [18:45:29] awjr: i'm okay if jdlrobson is [18:45:31] 2 hrs time? [18:45:42] correct [18:45:44] I just need to get this minification done so 2hr should be fine [18:45:48] cool [18:46:05] UTC Wednesday, March 7, 2012 at 21:00:00 [18:46:05] preilly - do we want to minify into one css and one javascript file? as that complicates things a little bit [18:46:13] jdlrobson: nope [18:46:17] awjr, okay [18:46:41] my latest module can be excluded from this deployment [18:46:52] yuvipanda, yes [18:47:03] also what's version=01182012210728 about in ApplicationTemplate.php [18:47:06] does that serve any purpose? [18:47:16] jdlrobson: busting the varnish cache [18:47:30] MaxSem: hmm, that sounds okay. [18:47:38] jdlrobson: just replace it with a unique string [18:47:46] i see [18:47:50] jdlrobson: probably a timestamp would work [18:47:54] same with all of the javascript/css files? [18:48:29] jdlrobson: yeah, just use a unixtimestamp like 1331149679 [18:48:37] k [18:48:38] jdlrobson: for all the css and javascript files [18:52:40] preilly can you just check a file page for me on your local instance? [18:52:43] i can't view them in mine [18:53:37] jdlrobson: File:Quark structure proton.svg works for me [18:54:00] jdlrobson: do you have Zero Rated Mobile Access turned on? [18:54:07] ahh yes [18:59:01] preilly - do we want to minify all the browser specific css files too? [18:59:04] e.g. iphone.css [18:59:31] jdlrobson: well, what do you think? [19:00:09] well.. since the files are small and the savings will be tiny i don't see much point as it's going to clutter the repository [19:00:39] jdlrobson: okay, well that is fine by me…. seems logical not to clutter things unnecessarily [19:01:02] ideally we need to find a way where the minified stuff doesn't have to live in the repository [19:01:12] jdlrobson: duly noted [19:01:23] jdlrobson: and I agree, but for right now we don't have that yet [19:01:28] * jdlrobson nods [19:04:29] yuvipanda: could you send me the user agent for the iPhone app? and is it different from the old one? [19:04:51] philinje: the new iPhone app, that is currently in beta? [19:04:56] or the one already in the app store? [19:05:14] yes, the new one, and if you could confirm the old one, that would be great [19:06:11] philinje: the old iPhone app sent no useragent, IIRC [19:06:21] yuvipanda: that's not true [19:06:30] okay, I remember wrongly? [19:06:42] preilly: do you know what it was? [19:06:42] I remember it sends an extra header [19:07:46] btw preilly - while you were away a guy called aharoni was asking about compatibility between Extension:WebFonts and Extension:MobileFrontend - some languages don't show up at all in the mobile site [19:08:04] e.g. https://bugzilla.wikimedia.org/show_bug.cgi?id=33648 [19:08:11] jdlrobson: okay, thanks [19:08:28] i said i'd make it a priority as it seemed like kinda a big deal [19:08:51] yuvipanda: so what is the new user agent? [19:08:54] would be good to look into it within the next few weeks if possible [19:09:03] * jdlrobson is almost done with minifying [19:09:08] yuvipanda: the old app now just does webView.customHeaders = [NSDictionary dictionaryWithObjectsAndKeys:@"Wikipedia Mobile/2.2.1", @"Application_Version", nil]; [19:09:36] philinje: new one is basically just 'WikipediaMobile/ ' [19:09:59] philinje: so whatever is the useragent of the iOS device the user is browsing with will be prepended with WikipediaMobile/ and be sent as UA [19:10:07] preilly: isn't that just an extra header, not UA? [19:10:15] yuvipanda: yeah, it is now [19:10:54] okay, so if I'm reading it right, it is just an extra header, and not the User Agent [19:11:07] yuvipanda: yeah, it used to be a user agent as well [19:11:12] btw php phpunit.php --configuration extensions.suite.xml is failing for me with an error 'PHP Fatal error: Call to undefined method FauxRequest::unsetVal() in /Users/jonrobson/Sites/MobileFrontend/tests/MobileFrontendTest.php on line 212' [19:11:16] does anyone else get that? [19:11:17] ah [19:11:17] preilly: any idea if Erik Zachte accounted for the old iPhone app header somehow? [19:11:39] philinje: I've got no idea [19:12:46] jdlrobson: FauxRequest doesn't have a unsetVal() method [19:13:13] heh i added unsetVal() to WebRequest [19:13:18] you can probably just add it to FauxRequest too [19:13:34] awjr: don't they share a base class? [19:13:47] jdlrobson, we can use ResourceLoader for CSS [19:13:48] actually i think you're right preilly [19:14:08] * preilly don't sound so surprised [19:14:21] i know my change to add unsetVal() was ok'd but i dunno if it made it into production yet. [19:14:51] we can make sure it does though when we deploy tomorrow [19:14:53] awjr, you should've receive a followup email if it's synced [19:15:04] MaxSem: i know, but i lose track of those things [19:15:06] MaxSem: doesn't ResourceLoader have javascript dependencies that we need to avoid currently? [19:16:12] preilly, no you can just insert one link a-la load.php?modules=mobile.core|mobile.android&only=styles [19:16:48] my unsetVal() change has not yet been sync'd [19:17:21] awjr: okay, well we need to make sure that is sync'd — can you do it or, do you want someone else to do it? [19:17:44] yeah i can do it [19:18:15] awjr: sweet [19:24:16] I could hack up MF to load CSS via RL after the deployment, if nobody objects [19:24:17] argh parser cache! [19:24:46] so I'm confused... shall I just do minification of the javascript? [19:25:09] jdlrobson, no [19:25:18] RL will do everything for you [19:25:24] ok, i think because of the parser cache + how MobileFrontend constructs output, doing sticky useformat=mobile just will not work with LinkBegin hook [19:25:43] including gzipping for user agents that support it [19:25:53] MaxSem: we can't use RL right now [19:26:02] for styles, we can [19:26:38] example: https://en.wikipedia.org/w/load.php?modules=mediawiki.debug&only=styles [19:27:08] MaxSem: yeah, but not for javascript [19:27:20] jdlrobson: so, you just need to minify by hand for now [19:27:42] why can't we serve CSS via RL for now? [19:27:47] yuvipanda: what is the version of the new iPhone app? [19:27:52] philinje: 3.1 [19:28:00] MaxSem: we can do CSS that way [19:28:02] philinje: it is not out in the wild yet, [19:28:05] MaxSem: just not JS [19:28:09] preilly: im throwing in the towel on useformat=mobile stickiness for now. after we deploy tomorrow i'll implement it with a cookie. [19:28:20] awjr: okay, sounds good [19:28:30] so be it [19:28:30] stupid parser cache grumblegrumble [19:28:37] so to clarify I should minify js but NOT css [19:28:53] awjr, $wgParserCacheType = CACHE_NONE [19:29:13] MaxSem: we'd have to do that all the time, which is probably not a good idea. [19:29:29] I meant for debugging:) [19:29:32] lol [19:29:45] i think it's a lost cause and an endless timesink at this point [19:30:09] but you're right, that does make it work :p [19:34:14] jdlrobson: can you start testing on http://zero-testing.wikipedia.wmflabs.org/w/index.php/Quark [19:34:36] preilly i'll finish the minification first - still a bit confused [19:34:39] just the js? [19:34:49] yuvipanda: any idea what the version number will be, or should we decide that now? [19:34:49] jdlrobson: JUST THE JS [19:34:52] k :) [19:35:08] philinje: the iOS one is going to be 3.1 [19:35:16] i think we decided that when I was in SF? [19:35:19] 3.1 iOS [19:35:20] android 1.1 [19:35:25] and they'll keep in step like that [19:35:32] jdlrobson: did you change some font properties? [19:38:18] preilly is something not rendering properly? [19:38:31] jdlrobson: stuff looks a bit bigger [19:38:31] yuvipanda: sorry, i thought we said 3.0 [19:38:45] i made a slight change yes to make it use browser default font settings [19:38:52] jdlrobson: also, the x in the zero banner exit button is weird now [19:39:05] * jdlrobson looks [19:39:25] yuvipanda: sorry again, i guess it was 3.1 [19:39:28] :) [19:39:53] which brings up the question, are we calling this major upgrade a point release? [19:40:23] philinje: probably not a good idea [19:40:26] philinje: it is going from 2.2 to 3.1 [19:40:31] 2.x to 3.1 [19:40:39] yuvipanda: why the .1? [19:40:44] preilly: android is 1.1 [19:40:49] same codebase [19:41:00] yuvipanda: hmm [19:41:02] hmm [19:41:55] will some people wonder if they missed 3.0, or will they be relieved we rushed into 3.1? [19:42:30] (in the old days, the dot1 release was when it became stable) [19:43:09] philinje: http://en.wikipedia.org/wiki/History_of_Microsoft_Windows#Windows_3.1_and_NT_3.x [19:43:10] :) [19:43:44] first version of WinNT was 3.1 [19:44:10] preilly can you remind me how to turn on the banners again.. i can't seem to display the banner at all [19:44:56] hmm, but there was no prior version or user base of WinNT [19:45:17] jdlrobson: RequestHeader set X-CARRIER "Verizon Wireless" in .htaccess file? [19:45:43] thanks man [19:45:51] anyway, if we think this iPhone release will be stable, then there is little risk in 3.1 [19:45:59] so it's up to you! [19:46:00] yeh easy fix [19:46:05] ;) [19:47:01] philinje: :) yes to 3.1 [19:48:02] MaxSem: is this if ( preg_match( '/]?class="references"/', $chunk ) ) {…. really the best way to detect sections with references? [19:49:35] jdlrobson: so, do you have a quick fix for it? [19:49:37] since this happens at the point where HTML is already divided into sections, we don't have DOM to help us. well, unless we want to parse all sections once more just for this check [19:49:44] pushing...its pushed [19:49:54] : jdlrobson [19:50:30] jdlrobson: much better [19:51:19] MaxSem: Okay, makes sense too bad we didn't have a little better way... [19:59:46] k pushed minified code [20:00:22] -> https://www.mediawiki.org/wiki/Special:Code/MediaWiki/113283 [20:01:42] also zero-testing has the same problem that mobile-geo had which was the xml on api.php is not enabled [20:01:49] it always returns json breaking the search [20:02:17] jdlrobson: do you have OpenSearchXml enabled? [20:02:37] i don't think zero-testing does [20:02:52] that should do the trick [20:03:49] jdlrobson: do you have labs access yet? [20:04:01] yes just logging in now [20:05:30] what's this boxes name? [20:05:34] ssh zero-testing-enwp.pmtpa.wmflabs doesn't seem to be working [20:06:30] jdlrobson: mobile-feeds [20:06:37] jdlrobson: after you login to the bastion [20:06:45] jdlrobson: require_once( "$IP/extensions/OpenSearchXml/OpenSearchXml.php" ); should be working now [20:08:45] mm aparently its enabled [20:08:55] jdlrobson: I just enabled it [20:08:59] :) [20:09:14] im still getting errors though.. [20:09:21] jdlrobson: really? [20:09:25] yeh [20:09:31] gonna force my cache clear [20:09:32] jdlrobson: like what? [20:09:59] k working now [20:10:01] it was just cache [20:10:12] jdlrobson: okay, coolio [20:10:26] doesn't seem to be many articles to test on here? [20:10:38] jdlrobson: not really [20:10:53] jdlrobson: you can export some from wikipedia and import them [20:22:29] jdlrobson: I'm importing some more pages now [20:28:00] jdlrobson: I imported a bunch of bridge pages [20:42:04] [WikipediaMobile] yuvipanda pushed 3 new commits to master: http://git.io/IuuUDw [20:42:04] [WikipediaMobile/master] import the common stylesheet from the MobileFrontend project - Jon Robson [20:42:04] [WikipediaMobile/master] minor tweaks to fix stying issues with introduction of common.css - Jon Robson [20:42:04] [WikipediaMobile/master] make sure jsEnabled class is added to body on startup - Jon Robson [20:42:33] Project WikipediaMobile - Nightly builds build #217: SUCCESS in 16 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/217/ [20:42:33] * yuvipanda: import the common stylesheet from the MobileFrontend project [20:42:34] * yuvipanda: minor tweaks to fix stying issues with introduction of common.css [20:42:34] * yuvipanda: make sure jsEnabled class is added to body on startup [20:46:07] so preilly there seems to be lots of annoyances with inline styling on tables and images [20:46:30] if they have a fixed width e.g. when viewed on a screen which has width 300px they jut out [20:47:29] i'm weary that people like the route guys rely on fixed widths though.. e.g. http://en.wikipedia.org/wiki/Wikipedia:Route_diagram_template so reluctant to mess with it in a rush [20:54:46] philinje: grab when you get back in [20:54:51] i want to talk about references [20:55:11] jdlrobson, you mean ? it's impossible to style images directly in wikitext [20:58:43] so take http://zero-testing.wikipedia.wmflabs.org/w/index.php/Bridge for example [20:59:05] the top image has width 337px [20:59:23] I can override this with a width: auto but in some cases this won't have the desired effect [21:02:21] tfinc_: would be nice to have an update on v [21:02:22] err [21:02:22] https://github.com/wikimedia/WikipediaMobile/pull/120 [21:02:40] yuvipanda: brion has the playbook and is working on it right now [21:02:47] cool! [21:02:55] awjr code freeze? [21:03:02] remember seeing a few tweets about it a while back, so keeping that pull req updated might be a good idea :) [21:04:42] yuvipanda: i responded to those [21:04:48] and then looked ay my calendar for this week [21:04:51] and gave up [21:04:55] :D [21:04:56] handed the playbook to brion :( [21:05:15] * yuvipanda notes to bring tfinc_ a 'I <3 Meetings!' T shirt the next time we see each other [21:05:25] * tfinc_ sneers at yuvipanda [21:05:25] jdlrobson & yuvipanda: https://www.mediawiki.org/wiki/Extension:MobileFrontend#action.3Dmobileview [21:05:51] cheers MaxSem [21:06:18] jdlrobson: i forgot to go eat lunch so im gonna go do that now. i should be back in 30 mins. if you need a little more time, just coordinate with preilly [21:07:00] MaxSem: cool, looking at it [21:08:55] MaxSem: so if I read it right, we'd be using action=parse for full html and this if we're doing per-section fetch [21:09:08] yes [21:09:47] action=parse also supports other stuff like treating page as main page [21:10:44] hmm [21:10:45] okay [21:18:24] jdlrobson: switching from wiki.css to common.css from trunk seems to cause problems for some infoboxes [21:18:37] example: check the one at Disco [21:22:21] we can push back the code freeze [21:23:04] ah, right. I can grab jon after the codefreeze. [21:23:08] * yuvipanda goes to write release notes [21:36:15] jdlrobson: the exit button on the banner is too low now [21:36:22] arghh [21:36:24] preilly: how's it going? code frozen or not quite? [21:36:45] awjr: not quite yet [21:37:22] how about if you remove height: 18px; from common.css [21:37:24] does that sort that out? [21:38:23] jdlrobson: it looks a bit better that way [21:38:42] jdlrobson: can you go ahead and make that change [21:38:55] k [21:40:21] pushed [21:41:43] jdlrobson: thanks! [21:41:47] jdlrobson: much better... [21:42:37] yuvipanda - there's likely to be a few problems with switching form wiki.css but hopefully the bugs we discover will help both projects! [21:43:07] jdlrobson: yes - so checked any of the infoboxes? [21:43:12] not yet [21:43:17] still looking through MobileFrontend [21:43:32] okay [21:43:38] i will do it straight after though [21:44:07] jdlrobson: They're issues with common.css, so I guess if you have them fixed in MobileFrontend, they should be fixed in the app as well [21:44:21] k ill take a look now then [21:44:24] ok [21:45:49] mm what's wrong with the disco page? Looks ok in MobileFrontend [21:47:12] hmm [21:47:17] let me get you a screenshot [21:48:18] jdlrobson: http://minus.com/m1BsKmPHM# [21:48:25] similar results in android as well [21:48:35] eek [21:49:01] * yuvipanda tries web [21:49:28] jdlrobson: same result in safari too [21:49:33] ah feck [21:50:13] * jdlrobson copying templates into a local copy frantically [21:51:00] jdlrobson: i just mailed out all the new mockups that i'm eager for us to work on [21:51:02] jdlrobson: i think you can export with templates [21:51:04] jdlrobson: take a look and then lets chat about them [21:51:08] some of them we'll attack in parts [21:51:28] for instance .. lets not do the previews within full screen search just yet [21:51:40] but grab me after you've had a chance to review them all [21:51:44] yuvipanda how do i export/import templates [21:52:05] jdlrobson: http://en.wikipedia.org/wiki/Special:Export [21:52:11] there's a 'Include templates' checkbox [21:52:12] ? [21:56:57] pchang: what happened to the 1.2 tracking bug ? [21:56:58] ergg my local version renders {{#if [21:57:17] Extension:ParserFunctions ? [21:57:49] sigh [21:57:50] jdlrobson: do you have require_once("extensions/ParserFunctions/ParserFunctions.php"); in LocalSettings? [21:58:40] still got some weirdness [22:00:59] hmm [22:01:43] pchang: do you have any links we could add for both product and ui on the roadmap http://www.mediawiki.org/wiki/Roadmap#Mobile ? [22:02:28] * preilly wonders why is tfinc and tfinc_ are here... [22:02:45] s/are/here/here [22:02:58] tfinc_: also something pointing to http://www.mediawiki.org/wiki/Glaucus would be nice. Possibly with a general timeline [22:03:15] hmm [22:03:19] possibly not there there, but still. [22:03:28] yuvipanda: Glaucus is still being worked on [22:03:32] not ready for a schedule yet [22:03:35] ok [22:03:48] do we have a schedule for when it'll be ready for a schedule? :) [22:03:50] where as all the mobile web bits can easily be worked on before the nav [22:04:12] yuvipanda: i'll let pchang field that one [22:04:26] but overall lindsey is attacking it this week [22:07:53] jdlrobson: still problems getting infoboxes to work? [22:08:05] nope already there [22:08:15] have 's all over the place despite the Cite extension being installed [22:08:19] nearly fixed it.. [22:08:45] ok [22:09:18] preilly: where do we have documentation on Zero ? [22:09:24] i want to link to something from the roadmap [22:09:27] yuvipanda: could you point me a list of pending work? I have become free now [22:09:45] heya shravan [22:09:50] tfinc_: yes, let me find something [22:09:58] yuvipanda: hey :) [22:10:40] shravan: let me go find that. Gimme a moment [22:10:46] yuvipanda: sure [22:10:48] shravan: also, have any nginx experience? [22:11:54] shravan: https://bugzilla.wikimedia.org/show_bug.cgi?id=33198 [22:11:55] ? [22:11:58] pchang: did you change your nick again [22:11:59] yuvipanda: no never worked on it, but i could try [22:12:25] shravan: basically, you could either pick that bug up I linked you to (CSS Animations yay!) [22:12:42] shravan: or help me setup a small proxy so we can host the app on the web and use it directly by going to a URL [22:12:56] yuvipanda: i will fix this CSS bug now :P [22:13:01] shravan: :) [22:13:10] yuvipanda: then move on to help you if that is cool [22:13:10] shravan: played with CSS animations before? [22:13:15] shravan: totally :) [22:13:23] yuvipanda: never but sounds interesting [22:13:31] shravan: it is, very much so :) [22:13:40] i've a spinner in one of my other projects I can point you to if you want [22:13:58] yuvipanda: the way I do spinners is, i just have multiple images which rollover using jquery [22:14:23] * shravan is pulling from the repo and building after a very long time [22:14:36] shravan: build instructions have slightly changed, check README [22:14:48] yuvipanda: ohh will do that [22:15:12] shravan: https://github.com/yuvipanda/POSM/blob/master/assets/www/style.css#L20 [22:15:16] check line 20-33 [22:15:28] you basically apply a rotate transform and set it to repeat [22:16:09] yuvipanda: ok but this will work if the browser supports CSS3 right ? [22:16:47] shravan: yes, and since on the mobile it is basically supported on all the browsers we support.. [22:17:11] shravan: http://caniuse.com/#search=animation [22:17:15] caniuse.com is wonderful :) [22:17:22] * aude waves [22:17:45] heya aude! [22:17:50] long time [22:18:15] jdlrobson: note: the table there seems to have a 'width: 22em' applied to it. If I clear that with web inspector, it all seems to work [22:19:12] yuvipanda: does this also belong to our team https://play.google.com/store/apps/details?id=uk.co.wikiapps.wikipro&feature=top-paid [22:19:25] shravan: nope, thats not us [22:19:33] its fixed yuvi [22:19:38] shravan: nope [22:19:47] try running make remotes and see if problem goes away [22:19:55] shravan: that used to be the top app in the market until we showed up [22:20:06] tfinc_: yeaa :P [22:20:29] 1.9 million users later..we come first for a search result :D [22:21:26] tfinc_: i hope it does not give people the wrong impression that we run that app dev [22:21:51] shravan: if anything it should up the quality in the market [22:22:00] there are A LOT of really BAD wikipedia apps out there [22:22:02] tfinc_ - will review your email as soon as MFE deployed :) [22:22:18] and it stems from lack of competition [22:22:35] jdlrobson: yup, fixed :) [22:22:36] i rather like having other apps rethinking about how people should navigate/interact with wikipeida [22:22:39] jdlrobson: k [22:22:50] jdlrobson: its for next week anyways [22:22:50] awesome yuvipanda - let me know if you see anything else as bad as that [22:22:56] jdlrobson: yes, am browsing around now [22:23:00] ahh ok tfinc_ that makes things easier [22:23:07] tfinc_: ohh cool :) [22:23:12] shravan: you can also help by beta testing :D [22:23:14] they didn't look too troublesome though when you previewed them to me [22:23:20] i fixed a lot the dead links and updated http://www.mediawiki.org/wiki/Roadmap#Mobile [22:23:23] looking forward to giving the page some shine [22:23:37] come monday phil, guillaume, robla, and i will look to merge our roadmaps [22:23:41] shravan poke at our app and keep hitting it until you make it break? [22:29:38] yuvipanda - on osx when you open sections on an article do you have issues scrolling to the bottom? [22:30:13] jdlrobson: iOS you mean? [22:30:22] yeh fs keep doing that [22:30:44] nope am able to [22:31:02] and on android too [22:31:10] do you have a specific page you want em to check? [22:32:11] i also have a flicker sometimes when i open sections [22:32:29] * shravan github is doing ssh key confirmation :S [22:32:43] might just be a glitchy simulator (using iphone Retina) [22:35:36] jdlrobson: no issues here. [22:35:40] just tried on retina [22:35:45] weird probably just me then [22:36:59] jdlrobson: when you're done testing, I've 3 pull reqs for you to merge before I can put out a beta :) [22:37:08] i saw one ;-) [22:38:39] jdlrobson: the internets must be slow there :) [22:38:40] tfinc_: how did the meeting with michal go? [22:38:51] aude: really well [22:38:58] we should catch up about sometime this week [22:39:04] yeah [22:39:05] i'm in the middle of something right now [22:39:09] ok [22:55:38] shravan: git remote [22:55:42] should show you a list of remotes [22:55:54] yuvipanda: yep [22:56:00] do you have the official one? [22:56:06] cannot remember how to merge [22:56:09] yuvipanda: yes [22:56:19] git pull [22:56:27] yuvipanda: it has not changed since I last worked on in feb right ? [22:56:35] the URL? [22:56:36] it hasn't [22:56:49] yuvipanda: after the pull step ? [22:57:05] shravan: assuming you're already in master, the pull will do both pull and merge [22:57:19] well, technically it does fetch and merge [22:57:23] but really [22:57:38] the next step if this finishes without conflicts is 'go ahead and write code' :) [22:58:40] yuvipanda: yes I recollect the steps now :) thanks [22:58:44] :) [23:07:02] * yuvipanda pokes jdlrobson [23:07:19] tfinc_ preilly do you know what price plan we are on on perfecto mobile? I want to reserve a handset [1] but I don't know how much that will cost [2] and whether I should care ([1] http://www.perfectomobile.com/portal/cms/tutorials/select_handset.html [2] http://www.perfectomobile.com/portal/cms/pricing/) [23:07:23] hey yuvi [23:07:26] you wanting me to pull stuff? [23:08:04] https://github.com/wikimedia/WikipediaMobile/pull/163 looks pretty straightforward.. mind if I automatic merge? [23:08:50] jdlrobson: go ahead [23:09:07] i'm guessing 2 of them are quite very simple [23:09:10] also care to explain the scrollHack and doFocusHack I keep seeing? [23:09:47] it just scrolls to the top of the page? [23:10:05] brion: ^ [23:10:46] jdlrobson: on some platforms we have to jump through some hoops to make scrollable sections work [23:11:05] iOS has a stub in there to initialize/reinitialize iscroll library if on iOS 4 [23:11:17] i can see the hoops :) [23:11:31] there should be a comment saying it gets overridden :D [23:11:56] i'm playing with the blackberry playbook stuff; it's a kinda weird and awkward sdk [23:12:09] [WikipediaMobile] jdlrobson pushed 2 new commits to master: http://git.io/bQg_qQ [23:12:09] [WikipediaMobile/master] Make iOS report as 3.1, not 1.1 - YuviPanda [23:12:09] [WikipediaMobile/master] Merge pull request #163 from yuvipanda/iOS-UA-fix - Jon Robson [23:12:18] doFocusHack is to show an 'active' class between touchStart and touchEnd [23:12:27] Project WikipediaMobile - Nightly builds build #218: SUCCESS in 10 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/218/ [23:12:28] yuvipanda: Make iOS report as 3.1, not 1.1 [23:12:58] it's possibly 'legacy' code, and something that either needs to be done in a cleaner way or gotten rid of [23:13:15] [WikipediaMobile] jdlrobson pushed 1 new commit to master: http://git.io/X_IxPA [23:13:15] [WikipediaMobile/master] Merge pull request #162 from yuvipanda/about-scroll-fix - Jon Robson [23:13:29] Project WikipediaMobile - Nightly builds build #219: SUCCESS in 8.1 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/219/ [23:13:30] yuvipanda: Fix scrolling issues in about page [23:14:35] yuvipanda: the application is crashing while building it [23:14:40] [WikipediaMobile] brion pushed 1 new commit to master: http://git.io/x_Jtzg [23:14:40] [WikipediaMobile/master] Merge remote-tracking branch 'gtanner/master' - Brion Vibber [23:14:55] Project WikipediaMobile - Nightly builds build #220: SUCCESS in 8 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/220/ [23:14:56] * gtanner: added in support for building to the PlayBook [23:14:56] * gtanner: removed permissions that didn't appear needed, updated viewport to fix scaling issue on PlayBook 2.0 and added build instructions for PlayBook [23:14:57] * gtanner: added ability to build for handsets, some scaling issue (tested on P9981) [23:14:57] * gtanner: added platform switch and phonegap.js file for blackberry handsets [23:15:04] yuvi - probably should remove https://github.com/wikimedia/WikipediaMobile/blob/master/assets/www/js/settings.js#L61 ? [23:15:16] hmm, I did? [23:15:31] jdlrobson: https://github.com/yuvipanda/WikipediaMobile/commit/448fafb2fc78c5d000275b8dfa4adaa640ca54d8 [23:15:39] sorry [23:15:41] my bad :) [23:16:27] brion: woot .. thanks for the merge! [23:16:45] it still needs some work but it's a clean merge (all new files) so easier to go ahead [23:16:56] menus are actually there, just missing the icons. ;) [23:17:47] though i'm not 100% sure _how_ the menus get created [23:17:51] brion_: any particular reason for re-introducing phonegap 1.3 js? [23:18:04] yuvipanda: just merging. should update it appropriately [23:18:09] ok [23:19:09] yuvi just trying out the settings fix [23:20:03] jdlrobson: use tab complete, so my client will notify me :) [23:20:15] plus it is then y, 2 keys instead of 4 :) [23:20:31] brion: send gtanner some feedback so that can clean bits up [23:20:35] yuvipanda: how are the betas going ? [23:20:46] tfinc_: last merge being done by jdlrobson [23:20:49] woot! [23:20:59] am writing the email here http://etherpad.wikimedia.org/android-1-1-iOS-3-1-third-beta-notes [23:21:30] tfinc_: i'd be able to sign the android build myself, but would need you to put it on dumps.* [23:21:35] [WikipediaMobile] jdlrobson pushed 3 new commits to master: http://git.io/aF1B8A [23:21:35] [WikipediaMobile/master] Tapping anywhere in settings dialog opense select - YuviPanda [23:21:35] [WikipediaMobile/master] Tap area for the selects in settings dialog much bigger - YuviPanda [23:21:35] [WikipediaMobile/master] Removed unnecessary code setting current language - YuviPanda [23:21:37] and you'll have to do the iOS build anyway [23:21:45] jdlrobson: thanks :D [23:21:49] and bam [23:21:50] Project WikipediaMobile - Nightly builds build #221: SUCCESS in 7.9 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/221/ [23:21:50] * jdlrobson: Tapping anywhere in settings dialog opense select [23:21:51] * jdlrobson: Tap area for the selects in settings dialog much bigger [23:21:51] * jdlrobson: Removed unnecessary code setting current language [23:21:52] yuvipanda: i'm sure one of our awesome ops folks can move it for you [23:22:03] i'll be busy 4-5 but can help after [23:22:12] even rebased for you this time ;-) [23:22:14] tfinc_: i'm blocking on you anyway. IPA [23:22:15] :) [23:22:18] or appe [23:22:20] apple [23:22:25] jdlrobson: yay you :) [23:22:30] yuvipanda: the application is not getting built. Here is the error log http://paste.debian.net/158928/ [23:22:48] tfinc_ takes a look at the change log to get a preview [23:22:54] tfinc_: am updating it now [23:22:56] jdlrobson: i [23:23:08] i'm going to commit the 'bumped version and CHANGELOG' commit directly to master. objections? [23:23:19] shravan: weird. uninstalled the app and try again? [23:23:26] preilly awjr what's going on with the code freeze? [23:23:36] no objections yuvipanda [23:23:39] :) [23:23:48] good question, i've been waiting to hear from you jdlrobson or preilly about it :p [23:24:00] * yuvipanda looks at brion_'s huge commit to make sure nothing other than blackberry code was touched [23:24:20] well I think I'm as good as I'm ever going to be today awjr - i'm sure there's some obscure device i've broken though :( [23:24:36] jdlrobson: this is why we deploy with a day to watch things [23:24:44] instead of deploying on fridays :D [23:25:52] tfinc_ im going to reserve a nokia n95 and see what we get charged [23:25:53] any objections? [23:25:54] ok - preilly, is everything else in then? are we … frozen? [23:25:59] the website is useless for telling me this [23:26:06] jdlrobson: does it really not tell you any pricing ? [23:26:23] well there's this http://www.perfectomobile.com/portal/cms/pricing/ [23:26:28] i've been hoping to sneak this cookie handling stuff in but i still haven't gotten it right. [23:26:39] jdlrobson: just do it and we'll figure it out [23:26:41] but im not clear which plan we are on and what it costs to reserve one [23:26:50] rachel would know [23:28:11] "Reservation for this handset is not possible until 2012-03-08 23:27 GMT" [23:28:25] helpful. :) [23:32:28] [WikipediaMobile] yuvipanda pushed 1 new commit to master: http://git.io/G_SpgQ [23:32:28] [WikipediaMobile/master] Bumped version and updated CHANGELOG - YuviPanda [23:32:42] Project WikipediaMobile - Nightly builds build #222: SUCCESS in 8 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/222/ [23:32:42] yuvipanda: Bumped version and updated CHANGELOG [23:33:31] preilly are we frozen? [23:33:49] don't make me play madonna.. ;-) [23:37:38] another cultural reference that goes whoosh [23:37:43] * tfinc_ shudders [23:37:43] haha [23:38:21] * awjr secretly loves madonna [23:40:23] shravan: any lucky? [23:40:53] yuvipanda: ya it did get built [23:40:59] shravan: :) [23:41:05] yuvipanda: had to make some small changes to the manifest file [23:41:22] oh? [23:41:38] awjr: i should forward you my friends madonna cover band [23:41:44] ! [23:42:20] yuvipanda: i removed the android:installLocation="auto" and android:xlargeScreens="true" android:largeScreens="true" attributes. It would not compile with them. The run time error got solved magically I don't exactly know what had happened. [23:42:30] that's interesting [23:42:42] are you installing on a device or emu? [23:42:49] yuvipanda: on device [23:42:54] also can you put back the other two and just get rid of the first one? [23:43:04] shravan: do you have space in your internal memory? [23:43:05] :D [23:43:27] yuvipanda: does not work. Eclipse refuses to compile them :) [23:43:37] yuvipanda: yes lots of space it there arn 600 MB [23:43:47] hmmm [23:43:48] weird [23:44:51] tfinc_: changelog's in, you might want to look at it to see how i've separated iOS vs Android vs common [23:44:58] k [23:45:10] tfinc_: can you make the ipa? [23:45:17] sure. let me pull the changes [23:45:32] you'd need to run 'make remotes' before heading to xcode [23:45:46] arg .. ssh key audit on github [23:45:53] :D [23:46:06] tfinc_: just did that audit :) [23:46:23] * yuvipanda has one key, knows its fingerprint, and has a ~40char passphrase for the key [23:46:33] but i'm really glad that git hub messages about tit [23:47:01] the best way to make sure you never reveal the key to anyone is to make your key a phrase about some embarassing personal factoid that you'd never ever tell anyone [23:47:02] :) [23:47:16] s/key/passphrase/ [23:47:30] yuvipanda: did a make remote [23:47:34] remotes [23:47:42] it pulled the app the app js and common css [23:47:47] ok [23:47:56] now off to XCode and the usual :) [23:47:57] any udid's to add before i build ? [23:48:18] i'm assuming you took care of all the ones that emailed you / mobile-feedback-l [23:48:22] yup [23:48:50] one new one .. the helpfulone [23:48:57] doing a regen with it [23:49:00] Thehelpfulone: ^ [23:49:20] * yuvipanda makes android build [23:49:32] building [23:50:31] settings menu wasn't response at first [23:50:33] seems fine now [23:50:40] i won't have time to test more until after 5 [23:50:45] but let me build you the ipa [23:50:48] ok [23:51:00] yuvipanda: did you do a version bump ? [23:51:10] yup, you did [23:51:10] ugh, not in XCode [23:51:21] 3.1beta2 [23:51:21] i did in Android, and I did in iOS about screen [23:51:40] let me update the plist [23:51:42] ok [23:51:56] wait [23:51:57] no [23:51:59] no point [23:52:00] I really need to write that 'appversion' plugin. [23:52:06] i have it set to 3.1 [23:52:12] because it has to match what the app store says [23:52:15] so were fine [23:52:15] ahgh [23:52:16] okay [23:52:28] (hopefully) :) [23:52:33] ok .. let me try and validate the ipa [23:52:40] ok [23:53:53] No issues were found :) [23:54:02] :) [23:55:43] http://dumps.wikimedia.org/iOS/Wikipedia-3.1beta2.ipa (1112b34a4019758cd8d41eb3d58d122b) [23:55:54] test it well before sending it out to everyone [23:56:04] yes :) [23:56:04] i've only done some quick spot checks [23:56:17] but at least i'm not blocking you anymore [23:56:25] the testing list is also up to doate [23:56:28] date* [23:56:32] so feel free to nudge people [23:56:44] ok .. last meeting of the day time [23:57:05] tfinc_: my plan is to send out email to lists now, then send personal emails to everyone who signed up tomorrow. [23:57:08] okay! [23:57:13] k [23:57:24] but check the ipa with 1-2 people before even sending the mail [23:57:30] to make sure that it installs clean [23:58:55] yuvipanda: solved that css bug :) [23:59:04] yuvipanda: will commit it and send a pull request [23:59:05] tfinc_: yes [23:59:08] shravan: cool [23:59:11] Thehelpfulone: poke? [23:59:32] [WikipediaMobile] brion pushed 2 new commits to master: http://git.io/S01vIw [23:59:32] [WikipediaMobile/master] Update Playbook PhoneGap JS from 1.3 to 1.4.1 - Brion Vibber [23:59:32] [WikipediaMobile/master] Fix for menu toolbar in web: update to iOS platform current version - Brion Vibber [23:59:35] yuvipanda: i have made a spinner.png home that is okay. Should I remove the .gif from images/ folder or keep it that way [23:59:47] Project WikipediaMobile - Nightly builds build #223: SUCCESS in 8.4 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/223/ [23:59:47] * brion: Update Playbook PhoneGap JS from 1.3 to 1.4.1 [23:59:48] * brion: Fix for menu toolbar in web: update to iOS platform current version [23:59:59] yuvipanda: hi :)