[00:00:01] at least in the response [00:00:09] the show/hide works on the mobile web yuvi [00:00:27] awjr, but not in the following requests [00:01:08] meh, time to sleep. I've got a mini-hackathon tomorrow with aharoni and vvv [00:01:10] which cookie mf_useformat=mobile [00:01:12] also the scrolling wheel isn't very prominent [00:01:22] I can barely see it, can you make it the same dark colour as the toolbar? [00:02:32] can we save images directly through the app instead of having to go to the mobile version? [00:02:36] mobile web* [00:03:02] Thehelpfulone: that's on the pipeline too - lightboxes for images [00:03:08] Thehelpfulone: scrolling wheel? [00:03:10] you mean the scrollbar? [00:03:32] I mean when somethings loading [00:03:45] the wheel that rotates [00:03:55] YuviPanda: i think he means the article loading animation [00:03:59] in the search box [00:04:00] ah [00:04:01] yeah that's it [00:04:03] right. [00:04:07] I didn't know the official name ;) [00:04:45] if I click on view saved pages [00:04:51] and then click on the x next to saved pages [00:05:00] the message is "index.html clear all saved pages?" [00:05:35] YuviPanda: i'm looking to head out in about an hour [00:05:41] tfinc: ok [00:05:54] Thehelpfulone: can you check the 'clear history' and tell me if you see the same? [00:06:05] YuviPanda: what's left before we can release to the market ? [00:06:13] that says "Confirm" as the title [00:06:21] and "Clear history?" as the message [00:06:38] Thehelpfulone: okay, looking into it [00:06:44] ty [00:06:45] tfinc: I see a RIL issue as blocking. [00:06:55] i'm working on it [00:07:03] Thehelpfulone: would you consider anything else blocking? [00:07:13] blocking == as a temp fix to the 1 star deluge [00:07:34] let me see hm [00:10:01] the problem with the settings is a blocker [00:10:01] Thehelpfulone: the index.html issue has been fixed [00:10:04] [WikipediaMobile] yuvipanda pushed 1 new commit to v3.1.2-iOS: http://git.io/wl0vVw [00:10:04] [WikipediaMobile/v3.1.2-iOS] Remove stray index.html title in 'clear saved pages' box - YuviPanda [00:10:07] how I can't get out of it [00:10:12] Thehelpfulone: i'm still unable to reproduce it. hitting back steems to close it for me :( [00:10:20] Thehelpfulone: does it happen everytime? [00:10:20] Project WikipediaMobile - Nightly builds build #327: SUCCESS in 8.8 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/327/ [00:10:21] yuvipanda: Remove stray index.html title in 'clear saved pages' box [00:10:27] yes [00:10:37] if I go to about us and then back to the settings then the back button works YuviPanda [00:10:44] but otherwise it doesn't if I only go to settings [00:11:06] Thehelpfulone: the back button to the left of the word 'settings' [00:11:11] yes [00:11:11] or the back button in the toolbar? [00:11:24] the back button to the left of the word settings [00:11:28] sorry, I'm talking about the iPhone now YuviPanda [00:11:39] theres no toolbar for settings in the iphone [00:11:50] MaxSem, preilly: i do not experience the same problem in chrome on windows vista [00:12:03] but it doesn't really work for me on the ipad either it seems [00:14:16] Thehelpfulone: works for the other overlays? [00:14:19] saved pages, history? [00:14:39] it seems so [00:15:31] also on the nearby YuviPanda, can you add one of those dots like you get on google maps? [00:15:39] and maybe zoom it out a little bit so you see at least one entry [00:15:56] Thehelpfulone: ah, right. file a bug? Will add it in for 1.2? [00:16:09] ok [00:16:24] this is kind of a blocker, if you click on the nearby optino YuviPanda [00:16:30] and then you go to an article [00:16:34] you can't get back to the nearby option [00:16:40] by simply pressing back [00:17:10] Thehelpfulone: that's an issue with all overlays. I understand that it affects nearby the most. [00:17:24] file a bug, and I'll make sure this gets fixed for 1.2 [00:17:49] ok [00:18:18] so i'm currently looking into the RIL issue, and will check settings back after that [00:18:26] Thehelpfulone: i'm still unable to replicate it on my simulator :( [00:18:31] the 'back button' not working. [00:18:39] brion: is the 'back' next to 'settings' working on your iPad? [00:19:13] YuviPanda: yep [00:19:20] brion: 5.1? [00:19:26] 5.1 [00:19:44] YuviPanda: sorry, back now - what was your question before? [00:20:06] philinje: nah, nothing. Just wanted to know if you had the mobile URL handy, no big deal. [00:20:13] preilly have you been able to replicate MaxSem's problems? [00:20:24] awjr: nope [00:20:37] hmm [00:20:44] YuviPanda: URL for? [00:20:55] awjr: but, I'm on the http://preilly.m.mediawiki.org/ domain NOT localhost [00:21:03] awjr: but, I don't really think that — that matters [00:21:06] the mobile designs. Again, nevermind, I found it and gave it to Thehelpfulone. [00:21:12] YuviPanda: do you need me to test anything ? [00:21:50] preilly well assuming Max's issues are unique to him, the only outstanding issue i see around the cookie stuff is for sites that use a single domain for both desktop and mobile view - there are browser caching issues [00:21:53] tfinc: confirmation that the 'back' button next to 'settings' works would be good. [00:22:02] k [00:22:03] preilly and i'm not really sure how to cope with them, or if there's even a sensible way to do so [00:22:06] tfinc: + ofcourse, general 4.2 testing. I'm figuring out ways to fix the RIL bug. [00:22:37] back button works fine [00:23:47] New review: preilly; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/4761 [00:23:49] Change merged: preilly; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/4761 [00:24:44] preilly what do you think about making having a separate mobile domain a requirement for MobileFrontend until we can figure out unique mobile paths? [00:25:15] awjr: well, I don't think we need to do it [00:25:34] awjr: you could just bust the cache with a custom query string parameter [00:25:53] preilly but that would have to happen on like, every request [00:25:55] awjr: just some random number converted to a MD5 hash or she! [00:26:04] s/she!/sha1 [00:26:29] awjr: just on the redirect [00:27:11] preilly with sites that don't have a separate mobile domain, there's no real redirect. so the URL for the mobile version of an article will be the same as the URL for the desktop version [00:28:09] YuviPanda: i did manage to replicate 'settings back button doesn't work, and hitting it a few times leads to a crash' on ipod touch (5.1) [00:28:18] brion: ow? [00:28:19] how? [00:28:31] just trying it a few times o_O [00:28:57] but i've not got it to happen on the ipad at all [00:29:31] LIFI [00:29:59] preilly so the problem is no necessarily just on redirect. you can have mf_useformat=mobile and be clicking around like crazy and then come to a page that was locally cached with the desktop version [00:30:21] brion: trying to reproduce now, switching between that and RIL bug (orientation switch not happening) [00:34:13] awjr: well short of cache control or vary [00:34:54] brion: I see it. [00:35:51] preilly well Vary is already set to cookie [00:36:09] jumping into 'about' and then back sometimes fixes it, but not always (may just be reopening the dialog, i forget) [00:41:04] awjr: but, that's for intermediate caches [00:41:44] brion: looks like the event is being unbound. Looking into it [00:41:58] preilly: is there something like 'vary' for browser caches? [00:44:26] awjr: not really [00:44:28] preilly i thought cache-control and vary were used by browsers as well as proxy caches [00:45:39] awjr: not vary [00:46:55] [WikipediaMobile] yuvipanda pushed 1 new commit to v3.1.2-iOS: http://git.io/A8aY5g [00:46:55] [WikipediaMobile/v3.1.2-iOS] Fix for 'back' from settings not ocassionaly working - YuviPanda [00:46:55] brion: that commit should fix it, can you take a look? [00:47:11] Project WikipediaMobile - Nightly builds build #328: SUCCESS in 9.6 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/328/ [00:47:12] yuvipanda: Fix for 'back' from settings not ocassionaly working [00:47:25] pulling [00:48:55] YuviPanda: doesn't fix it for me on the ipod touch :( [00:49:47] brion: damn, intermittent problems suck. It was working and now it isn't [00:49:54] YuviPanda: 10min warning [00:49:55] j [00:50:00] need me to test anything ? [00:50:14] yeah about 1/2 the time on the ipod touch settings close button doesn't work. WEEEEEEEEIRD [00:53:36] tfinc: Nokia — http://i.imgur.com/k0viD.jpg [00:53:47] brion: interlanguage list, are you able to scroll? [00:53:51] brion: 'read' in? [00:54:55] sec… app crashed on me while wildly tapping at settings close :) [00:55:23] this is reasonably infuriatig [00:55:46] yes, scrolling ok in language list [00:56:01] ok... [00:58:53] YuviPanda: need me for anything? [00:59:01] tfinc: not right now, no. [00:59:12] cya tomorrow, hopefully i'll have these two issues fixed [00:59:21] k, if i leave now then i'll be back around 10PM [00:59:25] maybe earlier [00:59:32] call if you need anything [00:59:36] tfinc: ok [00:59:47] i also setup a daily standup till this is in the market [00:59:48] brion: looks like it stops working right after the page loads. [00:59:54] you will have to deal with a daily meeting till we get this resovled [00:59:55] which makes no sense to me. It works until a page loads [00:59:57] hows that for motivation ;) [01:00:18] honestly, the fact that I can't look at the iTunes store or twitter without hating myself is motivation enough. [01:00:29] that's the spirit [01:01:51] maybe something in doFocusHack... [01:02:40] brion: i got rid of .cleanButton from there, still issues [01:03:05] .cleanButton or .closeButton [01:03:43] .closeButton [01:03:45] sorry [01:03:47] not cleanButton [01:03:54] cleanButton is there [01:04:19] yeah i can confirm commenting out the .closeButton in doFocusHack doesn't help :( [01:05:05] brion: not calling doFocusHack also doesn't seem to help [01:05:52] confirmed [01:06:11] and it doesn't seem to happen with the other pages [01:09:12] specifically make it a selector and binding it also doesn't seem to help [01:09:13] wtf [01:12:54] wtf indeed sir [01:18:18] * YuviPanda continues feeling like a failed voodo practiioner rather than a programmer [01:20:02] i've tried hooking the event on the particular element, checking touchstart/touchend…. nothing stands out. this is just damn odd [01:21:12] yes, tried those as well. [01:21:30] and it only happens at all on one out of four devices i've got. so wtf [01:21:40] but on that one it's so common, about 50% of the time. wtfffffff [01:22:32] more wtf. i commented out the event that bound click to showContent. other back buttons still work [01:23:53] ah, that was just an unclean build [01:23:56] not *that* wtf [01:24:05] :) [01:25:45] brion: binding to "button" also does not show up. [01:26:20] could there be some zindex thing or focus thing stealing events? [01:27:02] but why on settings and nowhere else… [01:27:17] * YuviPanda looks in settings [01:27:26] brion: the only difference is that we don't call doScrollHack for settings [01:27:51] * YuviPanda meant to say he was looking in CSS [01:36:29] brion: poke. [01:36:40] [WikipediaMobile] yuvipanda pushed 1 new commit to v3.1.2-iOS: http://git.io/uVchEA [01:36:40] [WikipediaMobile/v3.1.2-iOS] Adding Scrollhack seems to fix the back button in settings - YuviPanda [01:36:46] brion: can you try out this commit? It seems to fix the issue for me, but I can never be sure. Also, I've no idea why the fuck it works, if it works at all [01:36:56] Project WikipediaMobile - Nightly builds build #329: SUCCESS in 8.7 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/329/ [01:36:56] yuvipanda: Adding Scrollhack seems to fix the back button in settings [01:40:03] * YuviPanda pokes brion some more.  [01:57:14] sorry was afk [01:59:14] son of a bitch, that seems to fix it :D [01:59:17] lemme double-check on 4.3 [01:59:52] yeah, 4.3 is where iScroll does stuff [02:00:52] seems all as expected [02:01:00] \o/ you got it YuviPanda :D [02:01:13] 'keep shooting random people until you hit hitler' seems to be the way to debug [02:01:36] rofl [02:13:00] [WikipediaMobile] yuvipanda pushed 2 new commits to v3.1.2-iOS: http://git.io/VicJ5A [02:13:00] [WikipediaMobile/v3.1.2-iOS] Added warning about code in comments - YuviPanda [02:13:00] [WikipediaMobile/v3.1.2-iOS] Fixed bash misunderstanding - YuviPanda [02:13:09] right. now RIL and some more [02:13:15] Project WikipediaMobile - Nightly builds build #330: SUCCESS in 8.9 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/330/ [02:13:16] * yuvipanda: Added warning about code in comments [02:13:16] * yuvipanda: Fixed bash misunderstanding [02:43:34] YuviPanda: you about? [02:43:38] devgeeks: heya [02:43:41] hey [02:43:46] Just looking at RIL [02:44:20] devgeeks: awesome! [02:44:31] You won't say that in a second [02:44:32] lol [02:44:57] I think it's in RIL's "full" built in interface [02:45:34] yeah, that's the conclusion I came to too. Seems to be a bug there? [02:45:42] yeah [02:45:53] raise it with them? Have more weight coming from the big W [02:47:16] devgeeks: i'm not sure how iOS handles rotations, but if I understand correctly, it is done mostly automatically? [02:47:22] unless you want to do specific changes? [02:48:16] is that right or am I completely off my bonkers? [02:48:17] yeah, kinda [02:48:31] but they are creating their UI programatically [02:48:40] with frame widths and heights [02:49:28] interesting [02:49:33] i found a section of code marked with [02:49:33] // Allow for rotation [02:50:18] yeah [02:50:24] why isn't that working [02:51:25] devgeeks: maybe we're not passing it on to it? [02:51:32] we = phonegap [02:51:35] yeah [02:51:38] though it should [02:51:47] shouldn't be UP to phonegap [02:53:14] yeah [02:53:21] devgeeks: not like phonegap's hosting this [02:54:55] got an idea [02:55:20] devgeeks: ok? [02:55:49] gah [02:55:53] didn't help [02:55:56] :( [02:56:28] GOT IT! [02:56:30] BAM [02:56:43] * devgeeks does a happy dance [02:56:51] - (BOOL)shouldAutorotateToInterfaceOrientation:(UIInterfaceOrientation)interfaceOrientation [02:57:04] it's missing from ReadItLaterFull [02:57:25] rather: ReadItLaterFullView [02:57:39] Dantman: aha [02:57:44] err [02:57:47] sorry Dantman [02:57:49] devgeeks: haha! [02:57:54] heh.. tab fail ? [02:57:57] yeah [02:57:59] heh [02:58:04] I had d, hit tab. [02:58:13] what client do you use? [02:58:16] devgeeks: want to commit and gimme a pull request? [02:58:22] Irssi does that to me sometimes [02:58:23] devgeeks: textual, will migrate to weechat soon [02:58:34] well, here's the thing [02:58:40] it's still not right [02:58:45] see [02:59:00] it SHOULD be return [super shouldAutorotateToInterfaceOrientation:interfaceOrientation]; [02:59:09] but that doesn't WORK [02:59:28] I guess because of its plugin status it's not getting the correct answer from super [02:59:40] if I hardcode return YES [02:59:43] it works [03:00:06] but that would mean it would support orientations NOT supported by the app [03:02:02] Just gimme a sec to figure that out... [03:02:03] devgeeks: hmmm [03:02:05] ok [03:02:09] then the OTHER issue is [03:02:11] ... [03:02:22] To make it work we are changing THEIR lib [03:02:25] heh [03:02:31] we'll submit changes back up stream? :P [03:03:09] I am not sure what the licensing is on their lib. It's not clear [03:03:51] MIT, I think? [03:03:53] * YuviPanda goes to look [03:04:09] Not sure it says [03:04:33] hmm, it doesn't. will poke back and clarify. [03:04:37] k [03:04:46] i'm sure they're not going to proprietary it. Even if it is GPL that's fine for us [03:05:06] I'll make the changes on the assumption we can get away with it ;) [03:05:30] yup! [03:14:34] Which DO you support? [03:14:36] All? [03:16:02] devgeeks: looks like it [03:16:21] Trying to avoid a W specific hack [03:16:23] :/ [03:16:34] devgeeks: actually not 'all' [03:16:41] we don't seem to support 'reverse portrait' [03:16:44] not upside down [03:16:45] yeah [03:20:31] stoopid shouldAutorotateToInterfaceOrientation is getting called like 5 times just to pop up the modal, and answering differently every time ;) [03:29:33] devgeeks: ah [03:38:04] devgeeks: any luck? [03:38:15] [WikipediaMobile] yuvipanda pushed 1 new commit to v3.1.2-iOS: http://git.io/FYX0Dg [03:38:15] [WikipediaMobile/v3.1.2-iOS] Open attribution links under maps in external browser - YuviPanda [03:38:30] Project WikipediaMobile - Nightly builds build #331: SUCCESS in 8.6 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/331/ [03:38:30] yuvipanda: Open attribution links under maps in external browser [03:39:20] YuviPanda: trying stuff [03:41:18] devgeeks: ok! [03:41:22] :) [03:41:36] and crashing Xcode. I rule ;) [03:41:49] :D [03:44:01] infinite loop., oopsie [03:44:10] :D [03:44:12] they're fun [03:46:17] I like infinite recursion though [03:46:44] 'yo dawg, I herd you like a function call, so I'll put a function call in your function call in your function call in your function call in your ...' [03:47:02] hahahahah [03:49:11] Just trying to figure out how to get to the MainViewController from the ReadItLaterFullView [03:49:22] It's not as simple as super [03:49:47] ah [03:54:43] Or at least super isn't giving me the right answer ;) [04:09:12] devgeeks: think you can do a 'working fix first then do it right' type fix? :) [04:09:37] Yes :) [04:09:49] I have a 2 second fix that will fix it just for you [04:09:57] want that first? ;) [04:10:06] indeed sir! [04:10:10] hehe [04:11:37] ack! [04:11:55] It pops up correctly, just won't let me input [04:11:58] ough [04:12:00] not quite there :( [04:12:37] devgeeks: you might also want to work off v3.1.2-iOS branch instead of master [04:12:42] has a *ton* of iOS fixes that master doesn't [04:12:50] OK [04:13:07] devgeeks: might also be because of the lack of an API key... [04:13:31] My test app has one [04:13:34] I am using that [04:13:42] it's just a basic app to test the plugin' [04:15:37] gah! it's all messed up... their stuff is REALLY rotation unfriendly [04:16:10] failed login still gives an unrotated view [04:16:20] :( [04:16:47] probably why the "proper" method isn't working [04:16:58] one of the views is still saying NO to rotation [04:16:59] hehe [04:17:23] ah [04:21:39] This is getting a bit over my head, I am afraid [04:21:47] You might need to contact RIL [04:22:41] devgeeks: :( [04:22:58] devgeeks: diff what you have and give me so I have a starting point to poke them with? [04:23:23] yeah [04:23:24] sure [04:23:33] we're hoping to make a release tomorrow (1 star reviews ugh) and this is the final blocker. I might have to take this up myself when i wakeup [04:25:46] dang [04:26:09] yeah it's really messed up [04:26:32] if you are in landscape and pop it up, fine enough, but rotating after it's up does nothing, etc [04:27:09] ugh, this sounds nasty [04:27:15] It is [04:27:17] :/ [04:27:53] A proper iOS champ would probably be able to fix it, but there is a reason I use PhoneGap ;) [04:28:02] heh [04:28:07] And it's all to do with the UI parts [04:28:08] haha [04:28:11] :D [04:28:23] even then, I guess your iOS skills trump mine anyday [04:28:27] heh [04:28:47] dunno.. I am getting better everyday, but still [04:33:32] devgeeks: thanks! [04:33:37] now I shall get some sleepp... [04:33:51] hehe [04:34:04] take care, sorry I didn't have a better answer [04:36:07] thanks for your help! [11:45:23] New patchset: Jdlrobson; "redesign sections" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/4774 [11:45:23] New patchset: Jdlrobson; "toggling fix ups" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/4775 [11:45:24] New patchset: Jdlrobson; "jslint and styling fix ups" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/4776 [11:45:25] New patchset: Jdlrobson; "slight tweaks to footer text" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/4777 [14:36:13] hmm, those plugins are doing terrible things they're not supposed to be doing [14:36:36] if the problem is that you want the ReadItLaterFull VC to be rotating and it's not, [14:37:04] that's easy to fix, just implement shouldAutorotateToInterfaceOrientation: [14:38:57] this should do it http://pastebin.com/zxKJNuYi [14:39:27] because there's none in there, and the class inherits directly from UIViewController, it's falling back on the standard UIViewController implementation, which is portrait only by default [14:40:08] stick that somewhere between @implementation and @end in ReadItLaterFull.m and everything should be peachy [14:40:30] hope that helps :) [14:41:56] let's highlight devgeeks so he sees when he's around [14:44:49] those plugins are kinda badly mangling MVC; Views are supposed to be dumb terminals, and ViewControllers are supposed to handle all the view-data coordination, and classes inheriting from UIViewController should not have names ending in --View [16:26:17] gugh, bad time to get sick :| [16:28:35] hey YuviPanda- http://pastebin.com/Pt1C96Y6 [16:29:47] Should solve one problem [16:30:08] Rolken: reading [16:31:45] Rolken: so that was what I first did [16:31:52] that has no effect at all [16:32:00] and it was then that i poked devgeeks [16:32:04] Oh =\ Ah well [16:32:07] who also tried the same bit and found it to not work [16:33:51] YuviPanda: I left a couple of comments here a day or so ago, did you manage to get them? [16:34:34] Jarry1250: which ones? [16:35:04] we put out a new one after that [16:36:30] Rolken: any other thoughts [16:36:31] ? [16:38:13] hmm, I'll download phonegap and see if I can fix it [16:38:57] Rolken: the RIL code isn't really part of PhoneGap, it's an SDK they released [16:39:18] Yeah but it's hard to do much real debugging if I can't compile it [16:41:17] right [16:41:45] well, every version of PhoneGap I've tried to download has been converted to 'CDV'; is the one you're using online somewhere? [16:43:09] Rolken: they make it hard to find it. Give me a second [16:45:35] YuviPanda: The comments on 1.1.2 [16:45:59] Rolken: https://nodeload.github.com/phonegap/phonegap/zipball/1.4.1 [16:46:33] looking at it again, my second best guess is that you're nesting view controllers, and rotate signals don't correctly propagate between them by default, you have to do it manually [16:49:41] Rolken: did you manage to get it to build? [16:50:02] YuviPanda: Sorry, I mean 3.1.2. Okay, three issues I notice with 3.1.2 on iOS 4.3.5 (and probably others): the main page is still saved as ?useformat=mobile; the show/hide buttons on [[Blah]] (literally that article, maybe others) don't work; and Nearby->Change Language->Nearby will show articles in the old language, not the new one." [16:50:16] fussing with linker errors now that it doesn't like having both cordova and phonegap installed [16:50:42] now it builds [16:50:47] hey YuviPanda [16:51:12] tfinc: heya [16:51:26] dialing you over skype [16:54:14] hey tfinc YuviPanda! [16:54:47] YuviPanda: so how hard is the RIL bug ? [16:54:51] tfinc: I can hear you just fine [16:54:57] YuviPanda: you were coming in and out [16:55:05] it was getting frustrating [16:55:18] greetings jdlrobson ! [16:55:22] hmm, interent is fine. I wonder what went wrong. Sigh [16:55:52] tfinc: Rolken is helping out with it, and I'm lookin into it as well. [16:55:52] yeah there's some REALLY funky setup going on in openSetup:title: [16:56:01] Thehelpfulone: the bugs you reported have mostly been fixed :) [16:56:03] greetings Rolken [16:56:12] hi :p [16:56:34] Rolken: i'm looking for a way to poke the guy(s?) who wrote that. [16:57:02] this right here [16:57:03] YuviPanda: I saw the log, see I didn't make the bug up :P [16:57:07] self.baseViewController = [[UIViewController alloc] initWithNibName:nil bundle:nil]; [16:57:08] baseViewController.view.opaque = YES; [16:57:08] baseViewController.view.backgroundColor = [UIColor clearColor]; [16:57:09] [16:57:10] [[[UIApplication sharedApplication] keyWindow] addSubview:baseViewController.view]; [16:57:12] [[[UIApplication sharedApplication] keyWindow] makeKeyAndVisible]; [16:57:14] that… shouldn't really be happening [16:57:18] do you have a new IPA? [16:57:28] tfinc: can you make a new apk for Thehelpfulone? [16:57:31] sure [16:58:17] tfinc: 3 more changes since actual rc2, not sure if you can call this RC2 or RC3 [16:58:44] RC3 [16:59:06] right [16:59:22] all the weird garbage in openSetup:title: needs to be cut out, and presentModalViewController needs to be called from the actual view controller that's active normally [16:59:42] that plugin doesn't seem to know where to find that view controller, so it's just creating its own stack [17:00:26] YuviPanda: Thehelpfulone http://dumps.wikimedia.org/iOS/Wikipedia-3.1.2-RC3.ipa [17:01:36] ty [17:02:40] Jarry1250: right, we made a few fixes after the time you checked. That was RC1, and as you see, we're not at RC3 [17:03:06] this is the fundamental issue: http://stackoverflow.com/questions/5386056/control-which-uiviewcontroller-a-uiwindow-sends-shouldautorotatetointerfaceorien [17:03:24] YuviPanda: gdgd :) [17:03:26] i'm not sure if there's a hackish solution that doesn't require rearchitecting [17:05:30] yeah, it is relatively non standard (to say the least) [17:06:42] Rolken: i'm guessing most of it would have to be redone to make it autorotate [17:07:07] I mean, you could do something grotesque like have an app delegate property with secondary root view controllers, and have the other view controller forward any rotate messages to that property [17:07:27] it would make baby Jesus cry, but it might get the app out the door if it's the last blocking issue [17:08:02] Rolken: think you can make baby jesus cry and help us push this out? [17:08:15] lol, I'll give it a shot [17:08:34] * YuviPanda goes to poke more testers [17:11:10] Rolken: thank you! [17:11:45] New review: preilly; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/4774 [17:11:47] Change merged: preilly; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/4774 [17:12:54] New review: preilly; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/4775 [17:12:56] Change merged: preilly; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/4775 [17:13:48] jdlrobson: ping [17:13:58] hey awjr [17:14:04] how's it going? [17:14:18] jdlrobson: 122 })(); 123 }()); in https://gerrit.wikimedia.org/r/#patch,sidebyside,4776,1,javascripts/toggle.js [17:14:18] not too bad looks like the beta's finally coming together [17:14:35] \o/ [17:15:44] jdlrobson: shouldn't it be })(); [17:15:55] not according to douglas crockford [17:16:30] jdlrobson: when's good for you to chat about contact us stuff? [17:16:34] makes it more obvious that the function is immediately invoked [17:16:38] it's a stylistic thingy [17:16:40] both are fine [17:16:43] yeh awjr now is good [17:17:09] jdlrobson: can you give me like 15 minutes? [17:17:22] i need to pop out in 15.. but i should only be gone for about an hour [17:17:45] so 7.30 GMT+1 would also be good for me (11.30 your time?) [17:17:56] jdlrobson: well, I just thought that proper self invoking anonymous function expression required })(); [17:18:02] jdlrobson: ok no worries - yeah that's perfect [17:18:12] jdlrobson: running the new user test now [17:18:13] New review: preilly; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/4776 [17:18:15] Change merged: preilly; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/4776 [17:19:26] jdlrobson: let me know if anything can be done about spacing around the middle text in the footer, but please no changes to the latest prototype for about an hour [17:23:47] no worries i've got to go out for an hour now anyway [17:23:53] i changed the middle text no? [17:24:01] New review: preilly; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/4777 [17:24:03] Change merged: preilly; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/4777 [17:24:13] i removed the top margin anyhow like you asked [17:28:12] * MaxSem is back from a trollathon with aharoni and vvv [17:29:38] I gave Amir an introduction in MF, as a result he posted several bugs and started figuring out how to add WebFonts support to it [17:30:17] among things we discussed was special-casing of main pages [17:31:23] we came to conclusion that we should output pages not containing magic IDs unchanged instead of empty [17:31:26] thoughts? [17:31:55] ? [17:32:01] unchanged == ? [17:32:05] alright [17:32:08] MaxSem: hmm [17:32:11] that ended up not having to be that hackish [17:32:31] MaxSem: even though they may look completely crazy? [17:32:31] Rolken: wooo! [17:32:57] mind, this ONLY works properly because you're only using one ViewController normally [17:33:06] unchanged == processed like any other pages [17:33:22] so you can just take the ViewController already in the appDelegate and present from there instead of inventing a new one [17:33:59] i'll fork from git properly and put in a pull request [17:34:09] Rolken: awesome [17:34:26] you should branch from v3.1.2-iOS rather than master. has a lot more iOS fixes [17:35:06] sounds good [17:36:37] preilly, better than nothing [17:36:55] jdlrobson: r u there? [17:37:24] user test is done but strangely, the image was too wide, even though it was fine for me after the fix [17:38:35] MaxSem: well, technically it's not nothing [17:38:57] MaxSem: right now we give them a message that tells them how to add the sections to the mobile view [17:40:16] preilly, why we give it with JS, by the way? [17:43:08] MaxSem: because that is how jdlrobson decided to do it [17:43:27] MaxSem: so, no real technical reason [17:43:40] jdlrobson - ??? [17:44:50] YuviPanda, that exposed a few more problems with the way that ViewController's handling its resizing, so I'm taking a look at that heh [17:45:10] yaks need shaving. Sigh [17:47:37] just usual autoresize strutwork, shouldn't be a big deal. just coming up because Interface Builder usually handles this stuff fairly automagically but this particular VC is entirely baked in code [17:51:00] :) [17:51:08] tfinc: did you get axel's message and my responses? [17:51:19] YuviPanda: to what? [17:51:42] to mobile-feedback-l [17:54:06] YuviPanda: subject: "Search bar in iOS app" ? [17:54:33] if so i see his mail and no response [17:54:34] [Mobile Feedback] 3.1.2-RC2 on ipod 2nd gen iOS 4.2.1 [17:54:36] tfinc: ^ [17:55:11] yup, got that [17:55:31] hows #2 looking ? [17:56:18] tfinc: yeah, been looking at it. Will put a hack in place. [17:56:38] YuviPanda: axel continues to be one of our best testers [17:56:43] chrismcmahon_brb: --^ [17:58:02] http://en.wikipedia.org/wiki/User:AxelBoldt :) [18:00:08] tfinc: one of the staffers responded, reported everything faster and smoother. [18:00:21] have poked back asking for iOS / device details [18:00:23] which iOS version did they test on? [18:01:47] haven't heard back yet [18:01:59] fun post about versioning http://phonegap.com/2012/04/12/rolling-releases-how-apache-cordova-becomes-phonegap-and-why/ [18:02:14] from brian [18:02:22] brion: you around ? [18:02:29] tfinc: yo [18:03:36] brion: http://www.eventbrite.com/event/3322761469 [18:03:39] interested ? [18:03:44] i signed up for two tickets [18:03:58] not sure if i can make it but i wanted to snag them for you [18:04:09] ooh that could be useful for us yeah [18:04:16] steve mailed me earlier this week so that we could get on the list [18:04:21] are you available to go? [18:04:47] yep,looks clear [18:04:51] woot [18:04:55] i'll cc you on the thread [18:06:07] \o/ [18:11:59] tfinc: Microsoft will be providing free licenses for Windows so feel free to bring your Mac. [18:12:05] yup [18:12:08] pretty nice [18:15:48] MaxSem preilly - special casing homepage - main reason did it in js was so it didn't come up in google results... although maybe that would be a good thing I don't know. [18:16:12] MaxSem: also, we'll need to figure out how to get the special cased home page via the API [18:16:32] on leaving it the same - it's not such a crazy idea. When things break it encourages people to fix them. We have a mobile class now on the body tag which allows admins to add special rules to make things hide in mobile mode [18:16:33] philinje: http://www.mediawiki.org/wiki/Mobile_design still needs an update [18:17:10] jdlrobson, in such case the message should be added to JS only for main pages that lack content, not to every page on the site [18:17:34] but the only time that is ever empty is on the home page [18:17:39] when a page is empty it shows the create text [18:20:03] jdlrobson, I mean we definitely should not serve "empty-homepage":"This homepage needs to be configured. Read more here<\/a>" on every page view [18:20:20] yeh i see what you mean [18:20:34] although if you are talking about throwing out the special casing... that concern goes away no? [18:20:40] and if you don't want your main page to be indexed, there's [18:21:05] no.. i want the page to be indexed ... i just don't want the text 'This homepage needs to be configured' to be indexed [18:21:09] preilly: any blockers for todays deployments ? [18:21:55] tfinc: done [18:21:59] thanks [18:22:20] so, we have 2 feasible proposals: output what there is if no special casing discovered, and have a separate mobile main page [18:24:43] jdlrobson, mobile pages don't have extra links that help spiders, so empty main page == unindexed site anyway [18:27:59] jdlrobson: take a look at the user test email [18:28:45] which part of it philinje - the img issue? [18:29:52] philinje: ^ [18:30:41] do you know what phone he was testing on? that will help narrow down the problem [18:33:00] [WikipediaMobile] yuvipanda pushed 1 new commit to v3.1.2-iOS: http://git.io/6imu4A [18:33:00] [WikipediaMobile/v3.1.2-iOS] Make sure that 'Show more results' is visible - YuviPanda [18:33:26] Project WikipediaMobile - Nightly builds build #332: SUCCESS in 12 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/332/ [18:33:27] yuvipanda: Make sure that 'Show more results' is visible [18:38:12] Rolken: woo, that works [18:38:15] testing it around [18:40:52] Rolken: the diff looks a bit iffy, with lots of seemingly unrelated chunks committed... [18:41:07] Yeahhhh I noticed that [18:41:32] I moved the 'private' categories into the implementation file and that seems to have caused it all kinds of whitespace processing confusion [18:41:45] Rolken: git add --patch? [18:41:46] All the other edits are related [18:42:28] "No changes." [18:42:55] okay, let me clean up that diff [18:43:50] awjr, have you had a chance to look at my mobile feedback rewrite? [18:44:28] MaxSem not yet, i've been investigating browser weirdness with the new cookie handling stuff for out-of-the-box MF install [18:53:36] [18:53:50] preilly i could use a hand figuring out how to deal with browser caching weirdness with the new cookie stuff if you've got a few minutes [18:55:04] jdlrobson: I think that user had an older Android - the video might show the model [18:55:24] can you send me a link to the video? I don't know how to access the user testing site [18:55:35] jdlrobson: but the problem is more about sometimes when the page loads there are partial loads of the CSS, i think [18:55:41] i've got to pop out now i'm afraid but will look at it later [18:55:41] yes, will send the login [18:58:02] tfinc: Rolken helped us with a RIL fix! [18:58:12] laters.. [18:58:20] i'm cleaning up the diff so jdlrobson dosen't twitch when he sees is. will put up in a few minutes [18:58:52] :P [18:59:17] Rolken: it does have reasonably messed whitespace :) [18:59:41] Amgine: I have sent Ryan my details, but I don't yet know how to actually access the labs machines to try and set things up. [19:00:11] I'm sure Ryan_Lane will get back to you quickly. [19:01:54] about what? [19:02:14] details? where'd you send them? [19:02:15] oh [19:02:16] OSM [19:02:17] heh [19:02:37] email is like a blackhole for me. gimme a sec [19:03:26] :-) Is there a better way to have sent you the details? [19:04:51] there's a request page, where a number of people pull account requests from, but email is fine [19:04:57] you just need to ping me if you email me :) [19:05:59] Ah, sorry, wasn't aware of the request form [19:08:10] apmon: you have an email coming with your account details [19:08:29] great, thanks [19:08:41] lemme add you to bastion and maps projects [19:08:50] got it. [19:08:52] apmon: you should subscribe to labs-l, and join #wikimedia-labs channel [19:09:34] I have subscribed to the labs-l list, although I am currently receiving it in digest mode [19:10:02] * Ryan_Lane nods [19:10:08] can you join #wikimedia-labs please? [19:11:41] [19:11:50] Yes, can do that too :-) [19:13:04] I'll see how far I can get later on, and ask questions if you don't mind and I can't figure it out... [19:13:17] but I have to go now again. Thanks for setting up the account [19:16:51] [19:18:04] yw [19:18:58] YuviPanda: woot. thanks Rolken ! [19:19:17] yw :) [19:21:03] okay, I give up. i'm letting the commit go in [19:21:19] Thehelpfulone: the RIL bug has been fixed, thanks to Rolken :) [19:21:31] [WikipediaMobile] yuvipanda pushed 1 new commit to v3.1.2-iOS: http://git.io/-FILuw [19:21:31] [WikipediaMobile/v3.1.2-iOS] Merge branch 'v3.1.2-iOS' of https://github.com/Rolken/WikipediaMobile into v3.1.2-iOS - YuviPanda [19:21:44] Project WikipediaMobile - Nightly builds build #333: SUCCESS in 6.9 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/333/ [19:21:44] janardan.yri: ReadItLater issues fixed by changing modal presentation targets and eliminating previous ex nihilo method. Associated issues also fixed. Completing registration needs testing with a working RIL API key. [19:22:02] * YuviPanda goes to add Rolken's name to AUTHORS [19:22:02] great :D [19:22:23] * Thehelpfulone gives Rolken a cookie [19:22:33] :D [19:23:19] YuviPanda: to me that sounds like we have an RC4 which should be final [19:23:47] that plugin still really needs to be refactored properly though [19:24:20] Thehelpfulone: yeah. [19:24:24] tfinc: yeah [19:24:31] tfinc: i'm making version bumps and adding AUTHORS file [19:24:49] i'm guessing we can do one round of testing with Thehelpfulone (5.1) and you (4.x) and we should be good to go? [19:25:37] yeah, I can do it on both the iPad and iPhone [19:26:18] also note, as mentioned in the commit, it wouldn't let me hit up the RIL API, so you'll want to make sure it disappears properly on successful submit (if you haven't checked already) [19:28:03] [WikipediaMobile] yuvipanda pushed 2 new commits to v3.1.2-iOS: http://git.io/sWWzAg [19:28:03] [WikipediaMobile/v3.1.2-iOS] Added Rolken to AUTHORS - YuviPanda [19:28:03] [WikipediaMobile/v3.1.2-iOS] Updated version to RC4 - YuviPanda [19:28:10] Rolken: we'll be able to when tfinc makes a build with the API [19:28:13] tfinc: pushed [19:28:17] Project WikipediaMobile - Nightly builds build #334: SUCCESS in 6.9 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/334/ [19:28:17] * yuvipanda: Added Rolken to AUTHORS [19:28:18] * yuvipanda: Updated version to RC4 [19:32:16] pulling [19:34:37] YuviPanda: http://dumps.wikimedia.org/iOS/Wikipedia-3.1.2-RC4.ipa [19:34:44] Thehelpfulone: ^ [19:35:12] hmm, no kaldari anywhere [19:35:23] i'm poking axel [19:37:05] thanks [19:37:20] it's really quite a slow download tfinc though [19:37:24] its there a reason why? [19:37:35] Thehelpfulone: nope [19:37:46] if its slow then i'd expect wikipedia to be slow [19:37:46] takes a minute to download [19:39:17] awjr: related to bug 35842, I can confirm similar cookie issue being present in language selection as well. [19:39:41] srikanthlogic: you mean outside of MobileFrontend? [19:39:46] yes [19:40:06] srikanthlogic: do you know if there is a bug open for that particular issue? [19:40:13] translatewiki.net, try changing the language and browse through [19:40:35] i hate cookies. [19:40:45] awjr: I was just checking how cookie was handled in language selector code, and chrome behaves same way [19:41:04] LIFI [19:41:10] I get pages in language = xy even after I select language ab because of cookies [19:41:11] srikanthlogic: does it work ok in firefox? [19:41:22] * srikanthlogic checks in firefox [19:41:50] i have not experienced caching/cookie issues with my changes in MF in firefox nor IE 9 on windows vista [19:42:44] [19:43:52] awjr: strange, now works on both firefox and chrome [19:44:15] sigh. [19:44:17] Amgine, that's the way to go :) [19:44:32] * awjr smacks all the browsers [19:44:53] [19:44:55] awjr: may be you want to poke Nikerabbit or see language selector code, solution might exist! or you may find a bug there as well [19:45:54] srikanthlogic: i will. it sounds like from the behavior you were describing that there's a similar bug there [19:46:37] Thehelpfulone: were you able to test it? [19:46:48] * tfinc imagines IE not even noticing that it was smacked  [19:47:48] awjr: not its fault it is retarded. /me gives IE ice cream [19:47:49] awjr: chrome beats me, I get interface in 3 languages randomly, I think chrome has some extra caching or something [19:48:05] Keep in mind, IE is now the most standards compliant browser. [19:48:12] what, seriously? [19:48:15] srikanthlogic: i can confirm the same behavior on translatewiki.net in chrome [19:48:21] Seriously. [19:48:32] srikanthlogic: can you file a bug about that? [19:48:40] Amgine: that won't stop me from making fun of it [19:48:49] Amgine: we have years of that left before they regain credibility [19:48:54] New patchset: MaxSem; "Page title for Special:MobileOptions" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/4801 [19:48:55] awjr: yes, will file one on lang sel and link to the mf bug [19:48:56] yes, yes indeed. [19:49:04] srikanthlogic thanks :) [19:49:19] It also means they *could* have been standards compliant all along, and *chose* not to be. [19:49:35] there's still 7.1% of the world using ie 6 >_< [19:49:53] WE ARE THE 93%? [19:50:00] lol [19:50:07] awjr: and 16% of all wmf traffic i guess if am not wrong [19:50:09] yuvi is https://bugzilla.wikimedia.org/show_bug.cgi?id=35843 api-move stable to test? [19:50:11] YuviPanda: --^ [19:50:15] i'd love to see it [19:50:17] tfinc: yes, it is. [19:50:29] you'll find style issues, etc but the api loading code is there [19:50:31] try it out [19:50:33] on 4.2 [19:50:51] YuviPanda: just waiting for it to be recognised by itunes [19:51:11] tfinc: though, the 3.1.2 changes haven't been merged into api-move [19:51:17] tfinc: it won't really work on 4.x [19:51:25] tfinc: also, it's on my fork, not on official [19:51:55] yeah, i see it [19:55:40] tfinc: are you testing it? [19:55:57] yeah. just built it [19:55:59] tfinc: you need to do a 'git submodule update --init' before using it. [19:56:00] putting it on device [19:56:26] hmm [19:56:31] I can't copy from the app YuviPanda? [19:56:52] and also, the list of authors [19:56:59] jsoby has a full name? [19:57:16] Thehelpfulone: there's a bug for that, was scheduled for a soonish release [19:57:28] quite a big thing YuviPanda IMO [19:57:45] Thehelpfulone: we were picking up names given to git, removing obvious duplicates [19:57:53] YuviPanda: its too broken to test [19:57:54] on 4.2 [19:58:06] tfinc: told you so, it doesn't have changes from v3.1.2 [19:58:16] scrolling would be just as broken, etc. [19:58:23] i'm going to put this down for now [19:58:27] good potential [19:58:27] tfinc: ok [19:58:34] YuviPanda: so if I hold my ipad with the home button at the top [19:58:41] and go to read it later [19:58:50] it's 180 degrees in the wrong orientation [19:59:09] Thehelpfulone: I don't think even the app supports the 'upside down' orientation [19:59:11] heh [19:59:20] tfinc: you can still test it on a 5.x device, it had perf gains there too [19:59:21] yeah the app does YuviPanda ;) [19:59:21] the app should support upside-down on ipad [19:59:24] but not on iphone [19:59:27] ah [19:59:46] brion: I'm happy you found the settings bug yesterday [19:59:46] i'm going to wait till you stabilize on it [19:59:52] tfinc: ok [19:59:52] else YuviPanda was thinking I was a madman :P [19:59:59] Thehelpfulone: i'm glad yuvi stumbled on a fix finally :D [20:00:02] was driving us all mad [20:00:12] brion: any idea *why* that fix works at all? [20:00:16] i'm utterly confused. [20:00:19] noooooo idea [20:00:23] voodoo magic [20:00:43] you know that delay YuviPanda [20:00:47] * Thehelpfulone finds the bug number [20:00:56] Thehelpfulone: the button click delay? [20:01:00] yeah [20:01:12] what's causing it? it really makes the whole experience slower [20:02:29] and the ontact us page still doesn't work [20:02:40] from a fresh install YuviPanda [20:02:45] I go to it directly from the home page [20:02:50] without navigationg to any other pages [20:02:57] and it's just a white page, no text, no images nothing [20:03:12] Thehelpfulone: so, essentially browser webkit implementations have a 300ms delay between end of touch and triggering 'click' event [20:03:44] jdlrobson, so why we currently load CSS both from RL and static files? [20:04:31] tfinc: The contact page is going to production when? [20:04:32] awjr: ^ [20:04:36] damn. it crashed on my iphone [20:04:42] Thehelpfulone: ouw. [20:04:44] steps to reproduce? [20:04:45] on a call [20:04:48] * Thehelpfulone tries [20:04:49] YuviPanda: when we get it done, i guess :p [20:05:06] awjr: :) okay. [20:05:14] https://bugzilla.wikimedia.org/show_bug.cgi?id=35779 [20:05:23] I think this is really quite annoying [20:05:27] * Thehelpfulone tries to reproduce his crash [20:10:08] Thehelpfulone: any luck? [20:10:36] nope hmm [20:10:41] it seems to be intermittent crashes [20:10:49] would a crash log help? [20:11:08] Thehelpfulone: most of our issues are in js, so probably not :( [20:11:15] :( [20:11:17] Thehelpfulone: was it just 'use it for more than x minutes' and it crashes? [20:11:32] hmm not quite [20:11:41] I was playing around back and forth between the settings and about us [20:11:45] and it just crashed [20:12:16] then whilst I was trying to reproduce that, it crashed because it was loading russia and at the same time I pressed the home page and somethng else maybe too as it was taking forever to load nad it just crashed [20:12:22] too many things = crash? [20:12:58] possibly. we do cancel prior network requests, but maybe too many async starts and stops crash it [20:15:20] New review: preilly; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/4801 [20:15:22] Change merged: preilly; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/4801 [20:21:12] brion: why do we not support 'upside down' as an orientation anyway? [20:21:15] we *should* work... [20:21:37] for whatever reason, upside-down isn't conventionally supported on iphone/ipod touch *shrug* [20:21:39] we could though [20:27:00] well this is suspicious in the read it later plugin: http://pastebin.ca/2135492 [20:28:53] brion: am unable to see it. pastebin.ca doesn't like me [20:29:18] YuviPanda: http://pastebin.com/27ciCjuQ [20:30:55] yeah, that was just added by Rolken. [20:31:11] brion: i'm guessing it should instead just 'fall back' and support what the app itself supports, rather than hardcode it [20:31:20] well if i change it to just return YES, then everything works great on ipad [20:31:37] preilly: si [20:31:44] preilly: is there an rt ticket about the LIFI ? [20:32:42] oh, yeah, if you're using that on iPad it should be changed [20:33:13] YuviPanda: if we want we can just have that return YES and then also update our iphone deployment info to say it's ok to run upside-down [20:33:20] then everything'll work and we get to work upside-down :) [20:33:25] yeah, I think that's what we'll do [20:33:36] however, i'm wondering why we weren't showing upside down on iPhone before... [20:33:44] it's just the default not to [20:33:44] brion: making the changes [20:34:01] yeah, it's sort of conventional not to support upside-down on iPhone [20:34:12] but do so on iPad [20:34:18] yeah it's kinda maddening [20:34:19] yeah [20:34:20] which is sort of due to it being a phone [20:34:33] ie- if your phone rings, and you've got it sideways, you know you need to reorient [20:34:35] i understand, not sure where you'd want to use a phone vertically upside down [20:34:40] hmmmm, true. but we're not talking on wikipedia :) [20:34:45] if your phone rings and you're holding it upside down and don't realize, pull it up to your ear, confusion ensues [20:34:46] heh [20:34:59] hehe [20:35:18] i'll make it support whatever orientations are declared in the plist [20:35:26] or Rolken do you want to do that? :) [20:35:31] +1 [20:36:07] yeah, plus if we support upside-down then whenever y ou follow an external link you get an upside-down safari. pah [20:36:20] you can take a look if you want, I'm working on another app atm [20:36:26] Rolken: sure! [20:36:27] feel free to ask me if you hit a roadblock though [20:37:07] Rolken: sure [20:41:44] tfinc: not that I know of [20:42:01] preilly: i filed it ticket #3076 [20:42:33] tfinc: sweet [20:45:18] tfinc: question about why we're not using Testflight on mobile-l; i sent a quick reply but feel free to add more detail. [20:52:37] * preilly — * /branches/wmf/1.19wmf1/extensions/MobileFrontend/ (39 files in 8 dirs): 1.19wmf: MFG Change-Id: I70736224edd3dde12dd37c46028b1cb6059603d2 [20:52:51] awjr_lunch, MaxSem ^^ [20:53:10] eek [20:53:23] do we still deploy from SVN? [20:53:30] MaxSem: for 1.19 yes [20:53:34] tfinc: http://www.urbandictionary.com/define.php?term=liefi [20:53:57] brion: thanks [20:54:11] preilly: did you just add that? [20:54:23] tfinc: nope one of the Wikia guys added it [20:54:25] ohh . you did not [20:54:26] nice [20:55:23] MaxSem: do you see any issues? [20:55:43] on testwiki? [20:59:09] preilly, ?? [21:00:26] MaxSem: yes, on testwiki [21:01:43] awjr: can you take a look at testwiki [21:01:59] awjr: I updated to * /branches/wmf/1.19wmf1/extensions/MobileFrontend/ (39 files in 8 dirs): 1.19wmf: MFG Change-Id: I70736224edd3dde12dd37c46028b1cb6059603d2 [21:02:38] preilly sure one sec [21:03:43] preilly did you mean to totally synchronize from master? [21:03:55] brion: fixed [21:03:59] [WikipediaMobile] yuvipanda pushed 1 new commit to v3.1.2-iOS: http://git.io/kpAJRw [21:03:59] [WikipediaMobile/v3.1.2-iOS] Support all orientations for RIL on iPad - YuviPanda [21:04:01] though not in the best way possible. [21:04:02] awjr: yes [21:04:10] Thehelpfulone: do you see any other things that could be 'blockers'? [21:04:12] Project WikipediaMobile - Nightly builds build #335: SUCCESS in 6.9 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/335/ [21:04:13] yuvipanda: Support all orientations for RIL on iPad [21:04:25] preilly ok [21:04:35] YuviPanda: is that RIL bug fixed? [21:04:39] the one about the upside down orientation [21:04:44] preilly what should i be looking for? i'll double check my cookie changes although it will be semi-useless on testwiki [21:04:45] yeah, that was just fixed [21:04:56] okay, what about the 300ms delay bug? [21:04:59] tfinc can make you a build once he's out of the call, but I just tested it on the simulator and it works [21:05:02] cookies look ok [21:05:03] YuviPanda: Incidentally, I definitely managed to break search on my Android version. Not sure quite how though. [21:05:20] awjr: well, I'm pushing this stuff live at 4 pm PDT [21:05:20] Thehelpfulone: that would affect scrolling and other things and I don't want to rush it. Will definitely be the first thing to be fixed for 1.2 [21:05:27] ko [21:05:29] ok* [21:05:31] awjr: if you and MaxSem don't see any issues [21:06:00] ah might have found a bug in the search bar [21:06:11] partly because Jarry said about his search, so I tried to break mine :P [21:06:45] Special:MobileOptions seem to work, too [21:06:51] YuviPanda: there's no way to clear out all of the search bar? [21:07:14] so if you type something really long, usually on iOS there's a little x you can press to clear the field [21:07:39] YuviPanda: also, try to search for something that is 3 lines in length for example [21:07:53] it comes out of the search bar, you can't really read the second line [21:08:00] Agreed. [21:08:03] the article loading wheel keeps goingeven though it can't find results [21:08:29] new feedback also WFM [21:08:40] preilly: we should also test MF against 1.20wmf1 [21:08:40] which i think is running on… test2? [21:08:50] preilly when were you planning on pushing these? i dont see MF deployment on the deployment calendar [21:08:50] and didn't realize we were going to be doing a MF deployment today [21:08:50] also no changelog on http://www.mediawiki.org/wiki/Extension:MobileFrontend/Deployments ? [21:09:03] bbiab [21:09:25] preilly it would have been good to know about this deployment before… right now [21:10:06] also YuviPanda [21:10:10] I got to settings [21:10:11] language [21:10:15] the one just about english [21:10:20] Emilian e rumagnol [21:10:23] and go the home page [21:10:32] there's just something about the CC license [21:10:35] licence* [21:10:42] no main page though [21:10:53] Thehelpfulone: that wikipedia haven't setup their main page. [21:11:02] ah ok [21:11:05] there are instructions on the wiki + admins were all poked, IIRC [21:11:10] the cookie stuff seems ok but we'll want to double check that pretty hard once this is live [21:11:27] brion: Don't know if anyone's mentioned this, but you don't have to create an adhoc distribution profile for Testflight, though you do need to load UDIDs. [21:11:28] what's the ordering of this language YuviPanda? [21:11:34] it seems like it's a bit random [21:12:02] Thehelpfulone: IIRC it is Unicode order. [21:12:16] can it be made alphabetical? [21:12:21] oh ok [21:12:27] >_< [21:12:32] I switched from english to isiZulu [21:12:37] and then on the home page [21:12:40] I cant scroll down [21:12:46] we need to get the test -> incubator thing fixed... [21:12:51] brion: I wasn't the one who set it up on a project I worked on, but it was certainly way simpler than standard ad-hoc (although testers have to register install the TestFlight app itself). [21:13:22] Thehelpfulone: hmm, looking into that [21:13:31] show/hide is also still broken on the contact us page [21:13:43] I'm looking at es wiki now [21:13:52] Thehelpfulone: iphone or iPad? [21:13:56] (can't scroll down on isiZulu) [21:14:04] iPhone now [21:14:23] if I say language: es [21:14:32] then I'd expect About Us, Font size etc [21:14:36] to all become spanish? [21:14:54] (btw, why is it still "About Us"?!) [21:14:57] or is that on the general iphone setting instead [21:15:31] "This application is copyright (C) Wikimedia Foundation and contribution developers" [21:15:44] needs a . at the end and can you make (C) the actual copyright? [21:16:22] Thehelpfulone: interface language vs content language. I admit it is confusing, but it is reasonably complicated and will be fixed with the upcoming UI i showed you [21:16:29] ok [21:16:52] Jarry1250: didn't change it this late into the release since I was worried about translations. Will be 'fixed' in the next one [21:16:56] and in this back button etc for settings YuviPanda, I think clicking on the whole settings work as well should take you back, or at least make the back button a bit bigger [21:16:59] awjr: it's been scheduled since last week [21:17:12] when's the next release scheduled? [21:17:14] Thehelpfulone: yup, part of the 'click responsiveness' work. All this goes together [21:17:51] * Thehelpfulone would much rather prefer fewer releases and more complete full features than this releases every couple of weeks [21:17:56] preilly oh really? this is the first i've heard of it and it's not on the deployment calendar. i heard tomasz mention that there was going to be a zero deployment today, but i hadn't head any more [21:17:59] better experience for the users I think [21:18:01] Thehelpfulone: it *was* scheduled at middle of next week. I think it'll be pushed back a bit. Will talk to tfinc and see about the schedule. We definitely want to keep it sub-monthly. [21:18:11] the beta looks sexy. [21:18:47] uh oh im suddenly seeing no content on test [21:18:50] Thehelpfulone: I understand, but we'll try to cut down on scope a lot and properly test. I do not like hating myself, so will make sure that another shitty release doesn't go out [21:19:07] Thehelpfulone: we're trying to hire a QA person to manage the QA process, so that should help as well [21:19:08] there appears to be a 500 error [21:19:14] ok :) [21:19:31] http://wikitech.wikimedia.org/view/Software_deployments#Week_of_April_9 [21:19:34] awjr: ^^ [21:20:03] okay so now there's a search bar bug [21:20:13] I'm in spanish mode in the language version of the app [21:20:18] my phone is running in english [21:20:25] but if I type in english "football" [21:20:33] and click on football, [21:20:33] preilly i'm not seeing anything about mobile frontend on there... [21:20:39] it takes me to futbol (in spanish) [21:21:02] actually that could be intended feature [21:21:02] Thehelpfulone: http://es.wikipedia.org/wiki/Football takes me to futbol as well. The spanish community has redirects [21:21:11] ah [21:21:16] awjr: PHP Parse error: syntax error, unexpected T_STRING, expecting ')' in /home/wikipedia/common/wmf-config/InitialiseSettings.php on line 2173 [21:21:39] preilly want me to fix? [21:22:58] awjr: why would I need you to fix it? [21:23:16] awjr: I was just pointing out the reason for the 500 status code [21:23:33] preilly: ok no worries, i just wanted to offer help if you had your hands full with something else [21:23:49] Thehelpfulone: I can make you a branch which supports a hacky version of 'fastclick' once tfinc is here. [21:23:52] awjr: I'm just wondering who broke it [21:23:59] ok [21:24:14] preilly: svn blame [21:24:45] awjr: it wasn't committed yet [21:24:52] awjr: it was a local modification [21:24:54] o [21:25:05] awjr: -rw-rw-r-- 1 reedy wikidev 379780 2012-04-12 21:24 InitialiseSettings.php [21:25:18] * awjr shakes his fist at Reedy [21:25:30] awjr: [21:25:30] + "বাটচ'ৰা" => 100, [21:25:30] + "বাটচ'ৰা_আলোচনা" => 101, [21:25:37] ^ do you want to try and deal with these? :p [21:25:42] hehehe no. [21:25:52] but those are very pretty white squares. [21:26:05] It seems on bugzilla it has no ' [21:26:11] but then, paste to shell, it's got ' [21:26:21] so 'foo'bar' doesn't pass so well ;) [21:26:28] i ? unicode [21:26:55] Let me sync and commit these lines and you can have the file [21:27:06] YuviPanda: is there a beta version of the new UI? [21:28:41] awjr: right, working copy is now clean [21:29:04] cool thanks Reedy [21:29:06] Thehelpfulone: not yet. It's still at the comps/wireframe stage [21:29:06] preilly ^ [21:29:44] ok [21:29:44] awjr: yeah, I don't need to modify it — but, that was the reason for the 500 status code [21:30:05] Thehelpfulone: feel free to get involved with it on the talk page though! [21:30:05] awjr: are you happy with the state of test wiki? [21:30:07] Jarry1250: ^^ [21:30:16] MaxSem: are you happy wit the state of test wiki? [21:30:25] state of test wiki preilly? [21:30:31] preilly: yeah i see no issues [21:30:40] preilly, yes [21:36:30] awjr, MaxSem: Okay, I'll be pushing these changes at 4pm [21:41:36] * YuviPanda goes to work on api-move for a while, waiting on tfinc to make a build. [21:41:52] Thehelpfulone: just to confirm, this is a 'good to go' 'fix 1 star reviews release', except for 'fastclick'? [21:41:55] atleast on your device/testing? [21:42:04] Jarry1250: ^ [21:42:38] hmm [21:42:42] * Thehelpfulone does one last check [21:42:44] Yuvi: on iOS you mean? [21:43:01] Well the only issues I found I already mentioned, so if you've fixed them then i'm a happy bunny :) [21:43:10] YuviPanda, is the contact us page fixed? [21:43:55] Thehelpfulone: it will be with the MobileFrontend push. I'll talk to tfinc when he's back and see if we want to do anything else about it... [21:44:02] hmm ok [21:44:09] also the search bar doesn't hide automatically [21:44:14] it can't be hid? [21:45:25] YuviPanda: same thin with the history like it is with the languages [21:45:54] no, it can't be hid. Again, part of the new UX you saw - there're discussions going on about that on mobile-l and you should weigh in there [21:45:59] Thehelpfulone: the 'white flashing'? [21:46:18] thing* [21:46:25] did I call it that? sorry [21:46:43] so basically if I'm scrolling up and down on thie history quickly [21:46:44] the* [21:46:56] the entries are lost, like they are for the languages [21:47:05] it seems like it's reloading them or something? [21:47:07] well, they come back when you stop scrolling [21:47:18] yes [21:47:25] but they shouldn't disappear at all surely? [21:48:03] so they seem to 'disappear' while scrolling. We're not sure why this is happening, there's a bug for it and this shall be fixed in the next release too [21:48:19] that may just be the webview's behavior [21:48:54] ah you fixed the map zoom out on something bit :) [21:49:06] but can you put a dot to show where I am YuviPanda? [21:49:29] Thehelpfulone: next release or one after. As I said, this one is to fix all the 'doesn't work' type 1 star reviews [21:49:44] okay, I'll just read some of those reviews [21:50:44] Thehelpfulone: filed https://bugzilla.wikimedia.org/show_bug.cgi?id=35930 [21:51:22] did you fix the "cant scroll" issue? [21:51:37] I think I experienced that when I was testing, although not sure how to reproduce [21:51:46] ah nvm, that was the main page that didn't exist yet [21:52:15] hmm [21:52:32] YuviPanda, go to "Saenuri Party" [21:52:39] on an iPhone simulator [21:52:44] then go to the languages [21:52:49] scroll down [21:52:53] Thehelpfulone: yes, the can't scroll was on 4.x devices that we fixed [21:53:01] and between Norsk (bokmal) and Pyccknn [21:53:05] there's white spaces [21:53:15] clicking on that white space brings back the languages [21:53:23] but why are they missing [21:53:54] Thehelpfulone: https://bugzilla.wikimedia.org/show_bug.cgi?id=35779 [21:53:59] we'll fix this in 1.2 [21:54:55] ok [21:55:26] damn it crashed again [21:55:36] no idea what I did [21:56:31] oh YuviPanda theres still that bug with RIL [21:56:43] I can only fit "Thehelpfulon" [21:56:47] nothing more [21:56:47] Thehelpfulone: the build has not been updated. We need tfinc to update it and he's in a call [21:56:53] ok [21:57:13] in landscape mode [21:57:15] it's [21:57:21] Username: Thehelpfulon [21:57:23] Thehelpfulone: hmm, that. I'll poke the RIL guys and see if it is a limitation on their username size count [21:57:29] Password: ... [21:57:35] sorry that should be the same spacing [21:57:40] but there's a massive unecessary space [21:57:50] I don't think it's a limitation on their username [21:57:59] I've got "Tho" to automatically become Thehelpfulone [21:58:03] and I typed that in [21:58:10] and it let me use "Thehelpfulone" [21:58:55] I don't quite get it - how do I view the save to read it later stuff? [21:59:35] Thehelpfulone: Read It Later is an app / service that you share to, just like instapaper [21:59:53] ah that's misleading, I thought it meant that it would save it on my phone [22:00:11] that needs a tweaking somewhere to make sure people can easily tell the difference between "save page" and "read it later" [22:01:49] hmm, 'Share to Read It Later' isn't really accurate. There was another thing that cropped up about copy writing in the list [22:01:58] Thehelpfulone: file a bug? We'll make it clearer in the next release? [22:03:12] MaxSem: still awake? [22:03:17] yup [22:03:47] * Thehelpfulone nods [22:03:54] so I think everything else is fine for the release YuviPanda [22:03:59] Thehelpfulone: awesome! [22:04:00] MaxSem: just went over your special page changes - i think they're totally fine and it shouldnt be a big deal for me to merge the contact-us-redesign changes with what you've done [22:04:15] my 'quick and dirty' fastclick experiment failed, so I'll punt it to next release [22:04:43] preilly, MaxSem: i don't remember - did we come to a decision about which SpecialPage approach to move forward with? (MVC-ish vs separate SpecialPages) [22:05:14] heh sure [22:05:55] awjr, I don't think that there was a "OK, let's do this", but my last argument that MW core doesn't use MVC [22:06:55] MaxSem, preilly ok, it would be good if we came to a decision before i start merging our changes and moving forward [22:07:00] personally, i'm on the fence [22:07:26] I like the MVC approach [22:07:55] i like the MVC style a lot better and it will allow us to keep mobile-related stuff well organized. but since the ultimate goal of this is to get things into core, it is probably a lot more pragmatic to use separate special pages [22:08:27] awjr: I don't really see a reason that core couldn't follow this approach as well for these sort of things [22:09:29] preilly agreed and i dont necessarily think it would be a problem if we did it MVC-style, but it might be more difficult to get other MW devs on baord [22:09:32] *board [22:09:39] lemme look at your changes again [22:10:43] that said, it might be nice to keep things more loosely coupled with SpecialPages [22:10:52] but then i think it's easier to have code drift and fragmentation [22:13:25] (MVC?_ [22:13:27] ) [22:13:57] Jarry1250: http://en.wikipedia.org/wiki/Model%E2%80%93view%E2%80%93controller [22:14:36] awjr, once again: there's a huge amount of reinvention of the wheel in MobileFeedbackForm [22:14:58] MaxSem the plan was to convert it to use htmlform [22:15:04] just haven't gotten there yet [22:16:51] awjr: Ah, thanks. I'll get the hang of all these acronyms one day. [22:17:03] heh heh [22:17:06] :p [22:17:06] so what will remain of MVC once it's using HtmlForm and wht MVC can't use different special pages for diferent things, as it's customary with MW? [22:17:24] Jarry1250, are you coming to Berlin, btw? [22:18:33] MaxSem: It's a bad time for me, but I was think at a real push I could manage it if I got onto GSoC and could therefore get even more out of it. [22:18:44] But then, placces are filling up fast by all acounts :/ [22:20:17] So I'm torn. Also I haven't looked into the financial aspect of it, or indeed the travel time aspects. [22:20:27] preilly what advantages do you see with using something mvc-like over sticking with the traditional MW special pages? [22:20:59] Jarry1250, there are sponsorships available [22:21:48] MaxSem: Yus, but I don't know when the deadline is for them in relation to GSoC placement announcements. Presumably before? [22:22:16] interesting question [22:22:46] I'll ask Sumana [22:22:52] Thanks :) [22:24:03] awjr: just basically that it's a better pattern and it helps to isolate all mobile related activity to a succinct workflow [22:24:27] awjr: but, at this point I'm fine with whatever the two of you choose [22:24:28] ooh, pretty page pimping out firefox for android :) http://www.mozilla.org/en-US/firefox/toolkit/download-to-your-devices#sync [22:25:36] brion: nice [22:27:44] New patchset: MaxSem; "No more messages, decouple another not needed variable" [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/4817 [22:30:51] Jarry1250: poke [22:31:04] YuviPanda: Hey :) [22:31:22] Jarry1250: think you can give the new RC a go with your 4.x device? [22:32:15] YuviPanda: A quick test, yes. [22:32:28] Jarry1250: awesome [22:32:30] do we still need $wgMobileRedirectFormAction ? [22:32:38] back [22:34:01] tfinc: Thehelpfulone gave us a thumbs up, Jarry1250 is checkoing on 4.x [22:34:17] YuviPanda: Link? [22:34:50] Jarry1250: http://dumps.wikimedia.org/iOS/Wikipedia-3.1.2-RC4.ipa [22:35:03] so far so good on my 4.3.2 3gs [22:36:26] woot! [22:39:32] tfinc: i'm guessing we can submit to apple after brion and Jarry1250 give us a thumbs up? [22:39:41] tfinc: can you put it on your iOS device and test? (4.2) [22:39:44] last time, hopefully [22:39:46] +1 from me [22:39:56] tfinc: also, you need to make another build. A minor bug was fixed [22:40:51] YuviPanda: here is an idea. why don't i just give you the adhoc cert so that you can just sign the app yourself [22:42:06] tfinc: woah, you can do that? I thought it was locked to yours :| [22:42:19] were going to test it and find out [22:42:19] tfinc: go ahead, but this time you're the one who has to test anyway (4.2) [22:42:31] \o/ [22:43:13] YuviPanda: certainly seems to be working fine. One minor point: if you go onto nearby, open a popup, then change language and revisit nearby, the popup is still there, though the marker is gone. [22:43:20] *very minor :P [22:43:21] YuviPanda: sent [22:43:40] YuviPanda: what would you like me to test? [22:44:09] tfinc: in particular, test 1. settings page, check things can scroll [22:44:25] tfinc: orientations support, esp. with the 'Save to Read It later' window open [22:44:51] [[Blah]] still isn't working [22:44:56] Literally that page. [22:45:04] ok [22:45:04] The show/hide buttons on it. [22:45:59] hmm, *that* is weird [22:46:09] i can confirm on 4.3.2 :( [22:48:00] and on 5.1 [22:48:04] Also, I'm still getting incorrect markers if you go Nearby>Change language>Nearby. [22:48:04] might be something funky about the page, sigh [22:48:38] Jarry1250: you mean nearby -> back -> change language -> back -> nearby [22:48:39] ? [22:48:45] Yes. [22:49:04] Oh, wait, no. [22:49:22] have you guys been getting the weirdness where all of the scrolling happens super slowly? is that a known issue? [22:49:41] I think it's just loading Nearby in non-English still display English placemarkers (sometimes?) [22:49:45] * Jarry1250 tries to reproduce [22:50:43] No, more complex than that. [22:51:10] Rolken: on the release, or on the latest betas? [22:51:51] I got it both on the AppStore release and building the 3.1.2 branch myself [22:52:04] Rolken: what ios version & device? [22:52:21] Yuvi: Hmm, hard to tell, I think Nearby > Back > Change language > Back > Nearby problem *is* the common denominator but it's hard to tell. [22:52:27] iPhone 4, 5.1 [22:53:16] Rolken: would you describe the scrolling as "slow" like "jerky, stuttery", or more like "when flicking the page and letting it move, it doesn't go as fast as I expect"? [22:53:35] the latter [22:54:04] animation performance is ok, it just doesn't track with the finger properly [22:54:20] to clarify, it's with my finger on the page, not after I release [22:54:26] brion: Jarry1250 I just verified that the 'blah' page works on the api-move branch (not dependent on MobileFrontend, base for next release). Thoughts on filing a bug for this and not making it block this bugfix release? [22:54:41] hmm [22:54:46] it acts sort of like the rubberbanding that happens when you go off the top or bottom [22:54:50] well the overflow scrolling is known to be slightly funky: http://johanbrook.com/browsers/native-momentum-scrolling-ios-5/ etc [22:54:59] it ought to track w/ finger though [22:56:35] if I hit up Settings and promptly hit the back button away from Settings, the scrolling is fixed [22:56:35] feels ok to me on ios 5.1, iphone touch w/ retina display [22:56:42] curious [22:56:57] doing a search and selecting an article breaks it again [22:57:07] really weird [22:57:38] YuviPanda: Can you tell if it's just [[Blah]] or lots of other pages? [22:57:40] MaxSem: i think we should stick with regular MW special pages [22:57:46] i'm going to start merging our changes [22:57:55] Jarry1250: i've currently been unable to do it elsewhere [22:57:56] specifically I searched for 'Blah' (the article of the hour?), opened it, scrolling is broken; hit Settings gear, back to the article, scrolling is fixed; search again, pick the article again, scrolling is broken again [22:57:57] Rolken: oh ho, i got it to happen at least once [22:58:13] brion: uh. how? [22:58:52] tapped into search bar, typed 'golf', selected 'Golf' article. once loaded, it's doing the slow scrolling thing [22:59:12] on 5.x?! [22:59:17] yyyyep [23:01:01] brion: on 4.x? [23:01:14] only trying this on 5.1 so far [23:01:15] * preilly — pushing changes  [23:01:24] (p.s. apart from the Nearby-ChangeLanguage bug, there's also my favourite gripe from testing about the Main Page being saved/shared under titles other than "Main page"... otherwise I haven't noted any other issues that THO didn't) [23:01:41] restarted the app and it's not reproing. sigh. [23:01:41] Ooh, zoom on landscape is interesting. [23:01:44] preilly: woot. lets us know if you need us to test anything [23:02:12] Is the font size supposed to increase when you rotate? [23:04:01] Jarry1250: no, just that the bars are fixed height causes that to look so [23:04:45] Bars? [23:05:12] it is a little odd; the text gets out-of-scale bigger when you go to landscape, such that less text fits on a line in landscape than in portrait mode [23:05:22] whereas the images remain the same size [23:05:28] Rolken: scroll thing seems intermittent but is definitely there [23:05:40] Indeed. (Also, THO mentioned not being able to scroll the full length of the homepage, right?) [23:06:03] I think that was because it was a page that hadn't actually been created Jarry1250 [23:06:11] so that was the full length of the home page [23:06:41] Oh, well if I rotate into landscape, I can't scroll the full length on the en.wp homepage. I should mention that. [23:06:55] ...or indeed any page. [23:08:17] brion: Rolken i'm trying to repro the scroll issue, can't [23:08:20] hmm, let me do a clean [23:08:27] also, network provider change. brb [23:08:42] Jarry1250: i can reproduce that on ios 4.3.2. iscroll probably needs to be refreshed on orientation change [23:09:58] yeah, I see Jarry1250's issue on the simulator [23:10:04] brion: scrolling on 5.x still feels smooth to me... [23:10:06] ah [23:10:08] *gah [23:10:16] YuviPanda: it's smooth, but it doesn't track the finger [23:10:30] moving the finger across the whole screen scrolls it by half a screen [23:10:32] Right, before I have to go I might as well churn out one of my previous gripes about the fact that (X) says "Close" to me rather than "Clear history" (rubbish bin/trash can?) [23:11:24] Jarry1250: the discussion that happened when we added that was 'x is also clear', but I'll reopen that bug and poke people [23:12:12] Sure, X is also clear. But why be ambiguous? (I'm speaking from an English perspective with regard to icons - no idea about other cultures on that point) [23:12:32] here's a stub bug entry for the slow scrolling: https://bugzilla.wikimedia.org/show_bug.cgi?id=35932 [23:12:43] I don't think there's anything else on 4.3.5 that needs reporting. [23:12:53] MaxSem: Apr 12 23:11:51 10.0.11.4 apache2[3107]: PHP Fatal error: Call to undefined method MobileFormatter::useMessages() in /usr/local/apache/common-local/php-1.19/extensions/MobileFrontend/MobileFrontend.body.php on line 948 [23:13:41] preilly, botched up merge? [23:13:59] The basic reading function seems fine, history and saved pages seem to work fine. [23:15:24] https://bugzilla.wikimedia.org/show_bug.cgi?id=35933 <- and for the other scroll / orientation issue w/ 4.x [23:16:10] preilly, looks as if MobileFrontend.body.php didn't get updated to the tip of 1.19wmf1 [23:17:55] MaxSem: it's the same [23:18:03] MaxSem: I did a git pull and diff [23:18:03] Description for generic component reads "If your not sure where the bug should go" [23:18:29] *shudders* [23:18:31] preilly, http://svn.wikimedia.org/viewvc/mediawiki/branches/wmf/1.19wmf1/extensions/MobileFrontend/MobileFrontend.body.php?view=markup#l948 [23:18:47] it isn't eve in DOMParse [23:18:57] https://bugzilla.wikimedia.org/35934 < Minor zooming bug [23:19:00] s/eve/even/ [23:19:09] * YuviPanda looks at tfinc missing apostrophes.  [23:19:20] YuviPanda: huh? [23:19:21] and no useMessages() call for sure [23:19:23] MaxSem: it's the same as in GIT [23:19:35] [04:48:03] Description for generic component reads "If your not sure where the bug should go" [23:19:35] [04:48:33] *shudders* [23:19:37] tfinc: ^ [23:19:51] MaxSem: maybe that error is old [23:20:13] Jarry1250: then suggest what you would like that to read [23:20:18] Jarry1250: its used all the time [23:20:29] "If you're not sure where the bug should go"? [23:20:37] :P [23:20:44] I was just being nitpicky. [23:20:47] tfinc: you're missing an apostrophe. [23:21:29] brion: so apparently phonegap and orientationChanged don't go together well o_O [23:21:30] debugging [23:21:39] preilly, if it's just one time, could be a temp glitch due to one file being new at the time of request and another not yet [23:21:50] fixed [23:21:53] if it repeats, scap was borked on some server [23:22:31] Okay, that's a wrap for me. [23:22:39] I'll stop bothering you now :P [23:22:43] See y'all. [23:22:46] Jarry1250: thanks! [23:23:22] MaxSem: http://en.m.wikipedia.org/?lglg [23:25:10] awjr, fun: I click on mobile view on regular page, get to useformat=mobile, then immediately click on desktop view. it gets to useformat=mobile&mobileaction=toggle_view_desktop [23:25:47] [WikipediaMobile] yuvipanda pushed 1 new commit to v3.1.2-iOS: http://git.io/s22DoA [23:25:47] [WikipediaMobile/v3.1.2-iOS] Fix scrolling after orient change on iOS 4.x - YuviPanda [23:25:51] MaxSem that's not supposed to happen - you should not get useformat=mobile in the query string any longer [23:25:57] it is probably a cache artifact [23:26:00] Project WikipediaMobile - Nightly builds build #336: SUCCESS in 6.7 sec: https://integration.mediawiki.org/ci/job/WikipediaMobile%20-%20Nightly%20builds/336/ [23:26:00] yuvipanda: Fix scrolling after orient change on iOS 4.x [23:27:09] brion: fixed the orientation issue with not scrolling. [23:27:16] \o/ [23:27:17] still trying to reproduce scrolling issues on 5.x, can not :( [23:27:36] it's intermittent but reasonably often on my ipod touch [23:28:02] MaxSem i'll take a closer look once preilly is done with the deployment [23:28:36] awjr, you're right - it's caching [23:29:06] awjr: scap is still running [23:29:13] k [23:29:29] awjr: but, look at this: http://en.m.wikipedia.org/?cachebust=1 [23:29:34] awjr: no home page content [23:29:39] uhohs [23:29:49] what is that about? [23:30:21] "This homepage needs to be configured. Read more here" [23:30:24] hmmmm [23:30:32] yeah that ain't good. app's scraping empty page too [23:31:35] sommething wrong with testCanonicalRedirect()? [23:32:37] argh http://en.m.wikipedia.org/wiki/Main_Page?asdfasdfasdf [23:32:48] wtf [23:33:11] however, http://en.m.wikipedia.org/wiki/Poop?wregerg works [23:33:34] MaxSem: yeah, it's just main page [23:33:43] YuviPanda: confirmed scrolling after orientation fixed on 4.3 [23:33:44] brion: woo, ok. [23:33:56] brion: i'm still unable to replicate 5.1 scroll issue [23:34:12] preilly any reason to think it will just work once scap is done? [23:34:14] yeah let's worry about that one later. it's intermittent and can be worked around by scrolling more ;) [23:34:25] awjr: scap just finished [23:34:30] o [23:35:05] brion: file a bug? [23:35:11] brion: also, think this is 'good to go'? [23:35:13] tail -f /home/wikipedia/syslog/apache.log | grep php [23:35:14] YuviPanda: this one? https://bugzilla.wikimedia.org/show_bug.cgi?id=35932 [23:35:15] looks okay [23:35:43] brion: cool, thanks [23:35:54] YuviPanda: good to go on my end [23:35:59] critical stuff seems mostly fixed :D [23:36:47] brion: awesome! [23:36:57] brion: think you can do a bit of upgrade testing on iOS 4.x? [23:37:09] upgrade testing? [23:37:12] MaxSem: Any ideas? [23:37:36] YuviPanda: is it stable enough for me to test now ? [23:37:45] tfinc: api-move? [23:37:59] iOS [23:38:02] tfinc: no, i've been getting v3.1.2-iOS stable. [23:38:07] k [23:38:29] tfinc: v3.1.2-iOS has always been stable. api-move branch is the one that isn't stable on iOS 4.x, since the fixes haven't been applied on that branch [23:38:35] brion: yes, upgrading from prior versions of the app [23:38:40] preilly, can you repro it locally on a multi-domain configuration? [23:38:51] MaxSem: no [23:38:56] * MaxSem notes that he should use it too [23:39:12] not quite sure how to test that. delete, reinstall from app store, save some stuff, reinstall from xcode and make sure it's ok? [23:39:21] we should get a mobile version of testwiki [23:39:39] test.m.wikipedia.org [23:39:41] MaxSem: we've got an rt ticket open for it [23:39:49] i poked it earlier today [23:39:58] hopefully we'll get it sorted soon [23:40:08] MaxSem: http://test.wikipedia.org/wiki/Main_Page?useformat=mobile [23:40:10] brion: it is slightly weirder. switch to older branch -> archive -> organizer -> export -> drag ppa into iTunes -> sync [23:40:16] brion: do that again with newer branch [23:40:25] i'm doing that with 5.1 now... [23:40:38] preilly, so it seems to be reproduceable only on mobile site [23:42:31] MaxSem: yeah [23:43:05] YuviPanda: can't expand sections on the first article i loaded [23:43:16] MaxSem: could it be something around $formatter->setIsMainPage( self::$title->isMainPage() ); [23:43:17] tfinc: which article? [23:43:50] ok lemme see if this works [23:44:09] Red Army [23:44:21] it works fine after i navigate away and then back [23:45:38] same thing happens when i load IRC [23:45:50] but i'm seeing a much worse problem [23:45:51] tfinc: 4.x? [23:45:54] yes [23:45:56] preilly, true gets passed (as it should): we see a mobile main page. false gets passed: we see desktop main page. but not emptiness [23:46:12] YuviPanda: i'm only seeing the footer [23:46:14] nothing else [23:46:17] .. [23:46:33] thats what i get when i tap the W [23:46:39] tfinc: what language are you on? [23:46:50] English [23:47:02] YuviPanda: ok seems to work [23:47:17] tfinc: clean on xcode and build again? [23:48:03] brion: which bit? [23:48:21] tfinc: only seeing the footer for main page is because MobileFrontend is or very frecently was broken [23:48:38] YuviPanda: installing from master via itunes, then installing v3.1.2-iOS via itunes [23:48:41] because of todays push ? [23:48:58] tfinc: right. see http://en.m.wikipedia.org/ [23:49:10] yup, mf seems broken [23:49:14] tfinc: tes [23:49:14] preilly: --^ [23:49:17] tfinc: yes [23:49:19] k [23:49:24] i'll ignore that for now [23:49:28] other pages should work [23:49:33] yes [23:49:40] agh [23:49:42] argh [23:49:51] are there 1.20 wikis beside test and mw.org? [23:50:04] MaxSem: nope [23:50:12] dammit [23:51:00] if there was mobile domain for test, we could have just bisected it [23:51:05] MaxSem: why does this work http://test.wikipedia.org/wiki/Main_Page?useformat=mobile [23:51:37] YuviPanda: can't expand sections on [[Stoichiometry]] [23:52:31] YuviPanda: after tapping show 20 times .. they *finally* all opened [23:52:44] i've seen that a few times [23:52:54] YuviPanda: and then i was stuck between two sections .. couldn't scroll past them [23:52:57] tapping show doesn't seem to do anything, then suddenly *bam* you've got a billion sections open [23:53:00] MaxSem: why does http://test.wikipedia.org/w/index.php?title=Main_Page&stable=0&shownotice=1&useformat=mobile work [23:53:04] brion: exatly [23:53:07] exactly* [23:53:09] poop! [23:53:12] works on my device and on my 4.3 sim [23:53:22] its really bad on 4.2 [23:53:25] ... [23:53:25] it's been intermittent for me, not reliably repro [23:53:29] tfinc: does that ahppen on every page? [23:53:50] i wonder if it's 4.2 or if it's the 'phone 3g' model [23:53:54] hmm testwiki doesn't have any of the mp- ids but it has some mf- ides [23:53:56] *ids [23:54:11] it never ran 4.x well... [23:54:11] wikipedia mainpage only has mp- ids, no mf- ids [23:55:12] it happened on two pages that i tested [23:55:18] MaxSem, preilly: locally if i add mp-tfa and mp-itn divs on my main page, they are not showing up in the mobile view [23:55:18] now i'm just getting more empty pages [23:55:26] on non main page articles [23:55:32] :( [23:55:35] so this is issue is not just the main page [23:55:38] preilly: --^ [23:55:46] it's been intermittent on me [23:55:49] for the show/hide [23:55:58] for* [23:56:06] Thehelpfulone: right now? [23:56:16] it was earlier YuviPanda [23:56:19] i'm not sure if this is an MF issue as well or just an app issue. [23:56:20] but then it started working again [23:56:37] i just got a blank for [[Hopetoun Falls]] in the app, but it loads ok in browser [23:56:39] navigating away and then back to the article makes it work [23:56:59] awjr: $featuredArticle = $mainPage->getElementById( 'mp-tfa' ); [23:57:00] $newsItems = $mainPage->getElementById( 'mp-itn' ); [23:57:15] MaxSem, doing a var_dump( $newsItems ); on line 345 of MobileFormatter.php gives me null even though i have a
on my main page [23:57:47] awjr, on dev wiki? [23:57:53] MaxSem: locally [23:58:03] brb [23:58:54] brion: try the same article in the app again? [23:58:56] MaxSem actually doing a var_dump( $mainPage ) in the same spot is giving me an empty object [23:59:06] which seems… not right [23:59:21] and dumping $this->getDoc()? [23:59:34] YuviPanda: comes up blank multiple times in a row [23:59:37] MaxSem: same - empty obj [23:59:47] frakfrakfrak