[00:40:34] Wiktionary, WIkinews and Wikisource now show the mobile view from mobile devices (except tablets) - mobile-feedback-l at wikimedia dot org [00:40:57] hmm [00:42:07] preilly: does twitter do anything sane for posting email address inside a tweet? [00:42:16] preilly: so that they don't look like another user [00:58:03] Reedy: just as we've been seeing http://techreport.com/articles.x/23149 [14:02:17] Hm. Where's that page on how to write a main page for Mobile Frontend? [14:14:00] preilly: Where should I send wiktionaries who want to fix their main page? [14:31:37] Also: how will this (global user js/css) affect MF? https://gerrit.wikimedia.org/r/#/c/7274/ [14:39:30] Amgine, http://meta.wikimedia.org/wiki/Mobile_Projects/Mobile_Gateway#Mobile_homepage [14:39:45] [14:40:44] Yes. That page is not at all intuitively titled. [14:43:02] Also, now that you gave me that one, [[Mobile Projects/Mobile Gateway/Mobile homepage formatting]] [16:41:05] jdlrobson: that hat that looked like a cactus - was it *meant* to look like one? [16:42:11] i'm wearing it again today [16:42:17] no it's just a green beanie hat :) [16:42:41] so it's my 20% day today but I want to sort out that ios 5.1 bug [16:42:42] jdlrobson: am I the only one who thinks it looks like a cactus? [16:42:51] so far yuvipanda yes :) [16:43:01] heh :) [16:43:03] ios 5.1 doesn't use iscroll right? [16:43:06] jdlrobson: nope [16:43:10] no iScroll [16:43:53] jdlrobson: this 'fix' was 'discovered' when I found that about page doesn't scroll, while everything else does. Looked for what differs, and found that changig the order in which things were hidden and shown 'fixes' things [16:44:13] yeh but I want a logical reason why this is happening [16:44:17] this is happening too much :) [16:45:01] jdlrobson: yeah, true that. [16:45:24] jdlrobson: the 'root cause' might be somewhere deep in the bowels of webkit :P [16:45:44] if(navigator.userAgent.match(/OS 4/)) { [16:45:57] that matches iOS 5 right? [16:46:02] jdlrobson: no? [16:46:03] iphone 5\ [16:46:14] there is no iPhone 5 [16:46:22] lola getting confused again [16:46:30] so the problem is on iOS 5.1 right? [16:46:34] jdlrobson: yes [16:46:45] chrome.setupScrolling only gets added on iOS 4 [16:46:55] so therefore iOS 5.1 doesn't use iscroll? [16:47:06] jdlrobson: yes [16:47:10] it doesn't use iScroll [16:47:13] iscroll is only 4.x [16:47:14] ok [16:47:41] so how does scrolling on the #content element work in iOS 5.1 ? [16:48:05] jdlrobson: -webkit-overflow-scrolling: touch [16:48:34] jdlrobson: see app.css, styles for .scroller [16:50:07] so yuvipanda do you have a physical iOS 5.1 device? [16:50:19] jdlrobson: yup! [16:52:09] jdlrobson: the problem might also be in 5.x - I don't have a physical 5.x device to test on [16:53:08] k [16:53:18] hopefully I'll get a chance today to look at it [16:53:41] jdlrobson: any chance you'll let this get merged so it can go to market today? :) [16:56:22] yuvipanda: do we use css visibility anywhere? [16:58:29] jdlrobson: yes - it is being used to 'hide' the 'clearSearch' button [16:58:29] yuvipanda: MaxSem https://plus.google.com/hangouts/_/50e1e4a2c3c9d683e3d1c1885a65846c05ac0fd7?authuser=1&hl=en [16:58:31] should be taken out [16:58:44] jdlrobson: but that is inside the searchbar, think it might be causing this? [16:58:55] so yuvipanda http://www.webarnes.ca/2012/04/mobile-safari-bugs-so-far/ reports this problem " It also does not work if the scrolling DIV is contained in an element that was previously visibility-hidden" [16:59:54] another thought that yuvipanda is just to use iscroll on ios 5.1 .. [16:59:55] jdlrobson: hmm, #content never is set visiblity (just grepped) [17:00:05] jdlrobson: iScroll is slower... [17:00:10] does jQuery show or hide visibility? [17:00:11] plus we *do* have a mysterious working workaround [17:00:23] jdlrobson: yeah, looking at jquery source now for that... [17:00:34] yuvipanda: MaxSem --^ [17:00:51] but the workaround yuvipanda is just to never display:none #content right? [17:00:57] brb standup [17:01:16] yuvipanda: joining us ? [17:01:28] yup! [17:02:10] my default browser instance still hates this, so incognito + logging in [17:08:37] eww, the sound from SF was flaky today [17:09:10] it had interruptions and periods of reduced volume [17:11:18] MaxSem: interesting it was the same set-up we typically use [17:11:48] contact issues? [17:26:27] so yuvipanda find anything interesting in the jQuery source/ [17:27:55] jdlrobson: yep [17:27:56] uses display [17:27:58] not visibility [17:27:58] http://james.padolsey.com/jquery/#v=1.6.2&fn=jQuery.fn.hide [17:28:05] (nice website, that one) [17:28:43] so I have another theory [17:29:01] out of interest can you try adding a really high z-index after hiding the element [17:29:03] philinje: http://en.wikipedia.org/wiki/Special:ApiSandbox#action=query&prop=revisions&format=json&rvprop=user&rvlimit=10&titles=Petrified_Forest_National_Park [17:29:17] I wonder if some element is over it that prevents touch events from registering [17:29:32] jdlrobson: hmm, let me test that. [17:30:43] note we have some ridiculously high z-indexs in app.css :) [17:31:02] jdlrobson: we do :D [17:31:03] jdlrobson: testing [17:34:58] also try adjusting the rule in app.css for .scroller > * { to use -webkit-transform: translate3d(0, 0, 0); instead of -webkit-transform: translateZ(0px); [17:35:06] and remove .scroller -webkit-transform: translate3d(0, 0, 0); [17:35:20] found a report where someone said applying the transform to the scroller itself can break it [17:35:32] preilly: thanks [17:40:02] jdlrobson: z-index has no effect [17:40:14] investigating translate [17:52:26] it's times like this that i love not having central notice on mobile [17:52:45] for this just would make me screem http://en.wikipedia.org/wiki/Main_Page?usebanner=B12_061522_GreenButtonTall [17:53:11] tfinc: I don't see anything? [17:53:30] yuvipanda: you don't see a giant blue banner at the top of your screen? [17:53:40] if you've dismissed banners within their time window then it might not show [17:53:44] be happy that you don't [17:53:55] i tried incognito, doesn't show [17:53:59] although the use banner should avoid the cookie [17:54:27] yuvipanda: are you logged it? its anon only [17:54:36] tfinc: not logged in on incognito [17:54:54] * yuvipanda tries different browser [17:55:07] nothing on safari either [17:55:11] tfinc: geotarded perhaps? [17:55:14] then i'm out of ideas. its an obnoxious blue banner that loads above your page [17:55:15] yes [17:55:16] USA [17:55:27] explains it :) [17:55:39] http://en.wikipedia.org/wiki/Main_Page?usebanner=B12_061522_GreenButtonTall&country=us [17:56:20] wow [17:56:26] it is green, rather than blue [17:56:27] but still [17:56:39] ah now it is blue [17:57:25] tfinc: fundraiser started already? [17:58:59] yuvipanda: this is part of their normal testing [17:59:07] we just happened to see it as its in the us [17:59:31] heh [18:00:18] as ben put it. testing is good. lets test better things [18:13:44] jdlrobson: and the translate3D had no effect either [18:20:49] preilly: what do we need to do besides scheduling ashers time for next mobile default sibling move ? [18:23:23] tfinc: nothing [18:23:27] preilly: i'm reading through http://www.mediawiki.org/wiki/MobileFrontend/Dynamic_Sections and want to compare its gains to your blog post. where can i see your draft ? [18:23:28] preilly: k [18:23:34] tfinc: besides testing [18:23:44] preilly: will we need to do another one box test? [18:23:52] tfinc: yes [18:24:06] i want arthur to shepherd this one but he's not back till july 2nd [18:25:05] is the test just a request to ops or do we have to do anything ? [18:25:08] tfinc: I just have the this written thus far http://blog.wikimedia.org/wp-admin/post.php?post=15426&action=edit [18:25:34] tfinc: yes using curl to test for 302 responses is all [18:26:02] tfinc: the one box test is ops installing the redirector on one squid [18:26:10] preilly: thats an ok start but it needs our real world success stories [18:26:40] is that email or rt to get them to set it up ? [18:27:49] tfinc: I've been just talking to Asher directly about it [18:28:05] so a heads up email should be enough [18:28:11] tfinc: I understand that it needs our real world success stories [18:28:21] tfinc: yes a heads up email should be enough [18:31:22] jdlrobson: http://www.mediawiki.org/w/index.php?title=MobileFrontend%2FDynamic_Sections&diff=553766&oldid=553748 [18:31:41] philinje: --^ this gets us the best of both worlds [18:32:27] tfinc: is your internet connection slow today? [18:32:37] working just fine [18:32:48] tfinc: hmm mine is painfully slow [18:33:01] preilly: are you connected to a different access point ? [18:33:08] one that isn't local to our part of the floor [18:33:19] tfinc: I've a half-finished implementation for the app that does *exactly* what you mentioned in that diff. Let me see if I can put some time into that during the weekend [18:34:19] yuvipanda: excellent. when we get both fast initial load and async section load users will be really happy [18:34:31] the wont notice it but pages will load exceptionally fast [18:35:20] tfinc: true that! I had done this a few weeks back, and we hit a problem with the references not being available. MaxSem added an API feature to make that available, so no technical blockers. [18:35:23] hey jrobson [18:35:51] jrobson: neither z=index nor translate had anything [18:36:36] booo [18:36:47] sorry I thought irc was open but it wasn't [18:36:54] mm [18:37:49] oak try something else.. after $("#content").show(); do .css("-webkit-overflow-scroll", "touch") [18:38:00] ie. reapply it [18:39:53] trying [18:41:40] jdlrobson: no effects [18:43:11] mm yuvipanda mm [18:43:27] tfinc: any luck re: testers? [18:44:35] ok yuvipanda try putting position: relative on #main [18:47:36] yuvipanda: I wonder if this is related - https://lists.webkit.org/pipermail/webkit-unassigned/2011-August/359730.html (although this specifies an iframe) [18:48:39] jdlrobson: trying [18:48:58] jdlrobson: also, we were using iFrames in the app when I first started on it. Horrible level of horribleness... [18:49:27] http://forrst.com/posts/iOS_scrolling_issue_solved-rgX [18:49:31] jdlrobson: position: relative kills scrolling everywhere :D [18:49:34] use scrollTop ! [18:50:02] wtf [18:50:25] jdlrobson: position: relative on #main has no effect [18:50:44] k I'm interesting in the scrollTop thing [18:50:48] *interested [18:50:56] yeah, am trying it out [18:51:11] and also another suggestion I've seen is after $("#content").show(); do .css({"-webkit-overflow-scroll", "touch", overflow: "scroll" }) [18:51:15] e.g. set overflow as well [18:51:58] jdlrobson: scrollTop does the trick :D [18:59:12] haha [18:59:17] so yuvipanda [18:59:43] let's create a showContent event instead of doing $("#content").show [18:59:55] and lets link to that in the code so we have a reference point [19:00:04] *showContent function [19:00:31] yuvipanda: and also a corresponding hideContent [19:00:38] we might even want to patch jQuery show / hide [19:00:42] could be a query bug there [19:00:53] jdlrobson: ? [19:00:57] jQuery bug sorry [19:01:04] jdlrobson: jQuery show/hide does only display [19:01:15] jdlrobson: this is a bug in webkit. [19:01:41] jdlrobson: and why 'showContent' event? There's already a 'showContent' function - the problem here is that in this particular case, we want to show the main header but not the content [19:01:50] I meant showContent function sorry [19:01:54] jdlrobson: pages using showContent() didn't have this probelm [19:01:56] *problem [19:02:00] mm [19:02:19] chrome has showContent [19:02:28] this is the only place where we area ctually doing this 'raw' [19:02:36] Yeh I'm not sure what the criteria is for jQuery to cover a glitch like this - show/hide is obviously broken for elements with -webkit-overflow-scroll - would be good to create a MTC [19:03:14] jdlrobson: ah, yes! [19:03:44] but i'm unsure if this is worth the additional checking effort for every element that isn't -webkit-overflow-scroll [19:03:47] *doesn't have [19:04:10] yeh sure but would be good to create a bug in jquery and webkit so at least other people have a reference if they encounter this problem [19:04:20] so lets fix up the app and get it out and then create a MTC [19:04:40] I'm assuming the issue will also be present in a browser [19:05:57] jdlrobson: https://github.com/wikimedia/WikipediaMobile/pull/260 [19:06:49] jdlrobson: how did you hit upon http://forrst.com/posts/iOS_scrolling_issue_solved-rgX? A matter of having google-fu? [19:06:54] New review: preilly; "(no comment)" [mediawiki/extensions/MobileFrontend] (master); V: 1 C: 2; - https://gerrit.wikimedia.org/r/12270 [19:06:54] Change merged: preilly; [mediawiki/extensions/MobileFrontend] (master) - https://gerrit.wikimedia.org/r/12270 [19:14:16] yuvipanda: feel free to merge that [19:14:22] I'd prefer not to context switch [19:14:25] I trust you on this :) [19:14:47] do we use $("#content").show(); anywhere else in the code base? [19:15:38] jdlrobson: grep says no [19:16:13] yuvipanda: I searched for "-webkit-overflow-scrolling: touch" "when hiding" [19:16:16] apparently [19:16:33] [WikipediaMobile] yuvipanda pushed 2 new commits to master: http://git.io/bccKTQ [19:16:33] [WikipediaMobile/master] Make pages scroll properly on iOS 5.x when viewed from nearby - YuviPanda [19:16:33] [WikipediaMobile/master] Merge pull request #260 from yuvipanda/fix-no-scrolling-from-nearby - Yuvi Panda [19:21:51] tfinc: any luck finding willing testers? [19:24:00] i'd hopefully push this out for review at least today [19:27:20] i'm asking around [19:28:06] tfinc: thanks! [19:29:23] peter genres is in. sending him the test list now [19:30:17] sweet [19:45:42] heh, upgrade testing's quite a bore [20:04:31] yuvipanda: feedback sent to you [20:08:22] tfinc: yup, looks good [20:10:00] yuvipanda: james alexander is testing as well [20:11:29] tfinc: sweet! [20:11:41] i'll run through 5.1 for 10 mins more, then do a few rounds on 4.2 [20:11:45] k [20:40:35] [WikipediaMobile] yuvipanda pushed 1 new commit to master: http://git.io/E4fk0A [20:40:35] [WikipediaMobile/master] Fix intermittent search scrolling issue in iOS 5.x - YuviPanda [20:49:38] preilly: https://rt.wikimedia.org/Ticket/Display.html?id=2996 is the rt ticket for the new .m domains [20:49:40] still open [21:01:09] tfinc: should we ping CT about it? [21:01:21] preilly: i'm already talking to him about it [21:01:34] tfinc: okay [21:01:55] philinje: if elka isn't available can we find someone else to showcase the app? [21:01:59] perhaps maarten ? [21:09:02] maarten is harder to reach than Elke [21:09:29] IIRC multichill is sortof on vacation? [21:09:49] really? i'll just set the time and date - 11am on June 28, and propose that [21:10:14] he mentioned that in email a while back, might have been done [21:11:21] yuvi, does that time work for you? [21:12:30] philinje: yup! [21:12:53] cool [22:09:56] tfinc: send the app out to apple now? [22:24:37] yuvipanda: you can do it now ;) [22:24:49] alrighty! :) [22:31:22] wah that process looks daunting [22:31:29] with the amount of paperwork [22:31:36] (code-signing work, that is) [22:32:38] make sure you sign it with the right key [22:34:58] yes, RTFMing [22:35:11] In triplicate, in blood... [22:38:10] Amgine: facing the west, exactly at dawn, sacrificing your pet iguana... [22:40:33] yuvipanda: exactly at dawn *in Cupertino, CA, USA* [22:42:25] Amgine: while a particularly shaped shadow falls upon a particular table? [22:42:43] ooo, I dunno... try it both ways. [22:43:29] tfinc: I apparently 'do not have access' to the iOS Dev Center? [22:43:58] oh [22:44:00] nevermind [22:44:06] redirect bug on apple's side [22:46:30] tfinc: is iOS Team Provisioning Profile the one you use when submitting apps to the app store? [22:46:37] or something else? [22:47:37] apparently I need to have a profile that has 'app store' as distribution method, [22:47:43] but I don't see that info anywhere [22:52:05] and I don't have rights to create provisioning certificates on iOS Dev Center [22:52:17] * yuvipanda pokes tfinc  [22:53:51] no [22:53:56] we use the distro cert [22:54:00] do you see it? [22:54:04] if not i can send it to you [22:54:12] tfinc: i don't see it [22:54:12] yuvipanda: --^ [22:54:28] I see 'dev' and 'team provisioning' [22:54:38] sending it to you now [22:55:37] okay! [22:56:32] yuvipanda: sent [22:58:07] tfinc: only the private key 'wikimedia foundation' can sign it. [22:58:10] i don't have perms for that [23:04:08] tfinc: any luck? [23:04:22] ugh ok [23:04:29] well .. why don't i just sent it then [23:04:40] send* [23:04:43] tfinc: okay! [23:04:45] :( [23:04:49] that'll save you from having to deal with this pain [23:04:54] [WikipediaMobile] yuvipanda pushed 1 new commit to master: http://git.io/arGgiA [23:04:54] [WikipediaMobile/master] Remove animation for scroll to top in iOS 4.x - YuviPanda [23:05:03] yuvipanda: which commit should i pull up to ? [23:05:22] tfinc: bc1786145afb93d8f4728d020a72657467f7b3e3 [23:06:10] yuvipanda: got it [23:06:20] and since i was the one that caught the breakage … it'll be quick to know wether its still there [23:06:26] plus its great to get 1.7 out [23:06:26] :D [23:06:30] linSmith: yup! [23:06:31] err [23:06:34] tfinc: yup! [23:06:36] i'll submit it later today after the visual editor celebration [23:06:45] tfinc: alright. I might not be around. [23:07:00] thats fine. you've done a great job this week and should enjoy your time [23:07:03] tfinc: plus I think we can push out a 1.3 alpha to android on Monday! [23:07:05] i'll mail you if anything comes up [23:07:12] tfinc: sure! [23:07:17] yuvipanda: is 1.3 anything more then 1.7 phone gap ? [23:07:33] yuvipanda: how are we looking for thew WLM showcase? [23:07:36] tfinc: I've volunteered to run WLM in India with a bunch of people, so weekend will be spent on wiki :) [23:07:42] did you get all the graphic assets you needed ? [23:07:46] tfinc: linSmith sent me the styleguide, so am working on that now [23:07:48] tfinc: yup, did! [23:08:04] yuvipanda: does that mean you'll be beta testing our app in India for WLM ? [23:08:07] tfinc: she didn't update the design to figure out where to put monument search though, so that's one blocker (functionality has landed) [23:08:17] tfinc: well, not yet - there are lists to compile, etc before that. [23:08:20] philinje: --^ [23:08:22] tfinc: but yes, eventually I hope to! :D [23:08:25] woot [23:08:30] nothing better then eating our own dog food [23:08:38] true that! :) [23:08:40] yuvipanda: btw .. i still can't believe that your going back to scool [23:08:46] say it aint so [23:08:46] ;) [23:08:57] tfinc: 1.3 would be 1.7 and dynamic sections. [23:09:04] yuvipanda: nice [23:09:16] " Technically nicer looking, but actually does slow us down." ? [23:09:18] tfinc: I think I'm blocking it out of my head for now :P Wait for the second day of college for me to start cribbing :P [23:09:31] tfinc: it was animating the 'scroll to top' [23:09:39] tfinc: which was slowing things down by about 500ms on iOS [23:09:51] tfinc: and it was jittery-ish, so turned it off [23:10:23] so were including a fix that slows us down ? [23:10:45] tfinc: no, the animation was technically nicer looking, but actually does slow us down, so the commit killed it [23:10:59] the comment was about the animation, rather than th commit [23:11:01] ahh [23:11:10] that makes more sense. it wasn't clear from the message [23:11:11] got it [23:12:15] ok … time to congratulate the visual editor team. yuvipanda let me know if there is anything else you need before you crash [23:12:16] tfinc: i'll be in college for a week, will then join back up for WikiMania anyway :) [23:12:24] good [23:12:27] we'll hack there [23:12:27] :) [23:12:30] true [23:12:44] tfinc: pass on backslaps from me too! awesome job, etc :)