[11:05:04] Hi all [11:05:13] Hi [11:05:21] The jQuery Mobile team meeting will start in a minute [11:05:33] hi [11:05:54] hi [11:06:56] hi all [11:06:58] agenda here [11:06:59] https://docs.google.com/document/d/1WtuqYDOWhuQYFuOP1njpRbxK0BpGl3apSFEjHmrwpY8/edit [11:08:07] this might just be gseguin, scottjehl and todd_ (me) [11:08:15] looks like it [11:08:32] ok, then let's go! [11:08:47] first off, let's just check in on what we're focusing on for beta 2 [11:08:57] gseguin - what's on your plate right now? [11:09:11] I was working on the transition branch [11:09:17] not much luck there [11:09:23] no? [11:09:25] no? [11:09:33] I thought it would make an improvement on Android [11:09:33] samesies! [11:09:51] but I tried on Android 3.1 and it still sucks [11:09:52] still the same on android [11:09:56] oh. so you have it implemented, but no noticeable change? [11:10:04] they really need to fix those up [11:10:11] 2.2 looked a little better [11:10:27] we just got a galaxy tab 10.1 today to test honeycomb [11:10:45] well, this will also make it work on Firefox and Opera, right? [11:10:58] scottjehl: yeah I checked in yesterday in the transitions branch [11:11:15] so Opera is very jumpy [11:11:19] oh, ok so we can try it out? [11:11:26] yeah? opera mobile? [11:11:27] I told miketaylr about it [11:11:33] he filed a bug [11:11:57] yeah, bug filed. will advise when i find something out. [11:12:30] so miketaylr, is this just an issue that opera will eventually sort out? [11:12:41] thanks miketaylr :) [11:12:46] if we tweak transitions to work this way, it'll eventually fall into place [11:12:53] yeah, thanks [11:12:55] todd_: not exactly sure what the issue is...but i hope so :) [11:13:01] us too [11:13:26] we'll give it a look, but it seems like this is the way to go in terms of compatibility, no? [11:13:34] yeah [11:14:08] so yeah I was hoping that this would solve 3 of my open bugs but... no luck [11:14:11] did you test on Firefox at all? Desktop of Mobile [11:14:17] heh [11:14:25] I did test on FF desktop and it does work [11:14:29] cool [11:14:38] do you have mobile installed? [11:14:51] the pop.reverse is not working well yet [11:14:56] gseguin: any noticeable decline in smoothness? Drawbacks you've seen so far? [11:15:06] ok [11:15:09] that's really the question. [11:15:18] scottjehl: not really [11:15:28] k [11:15:43] but you guys are probably in a better place to judge of that [11:15:44] if it's not worse, we should just go ahead and land this when ready. the browsers can sort out their issues [11:15:52] i'll do some testing today. [11:16:06] the pop.reverse needs some work [11:16:06] at this point, did you convert all the transitions over in that branch? [11:16:19] and flip won't work, right? [11:16:23] yeah except the flip [11:16:37] are we gonna leave that as-is for now? [11:16:40] flip will need keyframes? [11:16:47] it's a tricky one because it looks so bad in non-iOS [11:16:55] cartwheel city [11:16:56] wait actually [11:17:06] sort of wish we never released that one [11:17:07] * gseguin is double checking [11:17:08] yeah? [11:17:37] it'd be cool to break these into separate css files eventually. Maybe just include slide L-R, pop, and fade by default in our builds [11:17:54] yeah no flip is using transitions too [11:17:57] a good idea anyway [11:18:02] leave out wagonwheel/flip [11:18:05] yeah? [11:18:07] but it'll work only on iOS [11:18:10] make it easy to include if you wan tthough [11:18:18] and not wagonwheel on android? [11:18:34] hmm... gotta check this [11:18:52] * gseguin is checking now [11:18:57] k [11:19:06] dod you say the branch was "transitions" [11:19:27] yes [11:19:50] oops [11:20:09] yeah the flip on android is doing a rotation of the page [11:20:22] don't know if that's what you call wagonwheel [11:20:41] but it's not looking very nice [11:21:51] that's it [11:21:59] such a bad way to handle it [11:22:11] we're checking this out now [11:22:17] that's true in master too though. no big deal [11:22:30] yeah [11:22:38] so i'll do some testing today [11:22:40] so you have all the transitions switched over to css transitions now then? [11:22:42] maybe using translate3d would solve it [11:22:46] and if you can fix up any loose ends [11:22:52] we can probably push this in [11:23:07] not the spin [11:23:23] k [11:23:43] that could always move to a gif/data uri [11:23:54] yeah I need a CSS expert (read you or scott) to take a look at the pop.reverse [11:23:58] Kin was saying it's a tad costly to be animating that thing during transitions anyway [11:24:16] oh, for the loading spinner? [11:24:20] ya [11:24:24] we only use spin for that [11:24:31] gotcha [11:24:42] we should make a ticket for that [11:24:52] ok, so let's all test this branch [11:25:03] and gseguin will button this up [11:25:31] although that spin thing may work on android using transitions [11:25:42] should OTransitionEnd be oTransitionEnd? [11:25:47] worth a shot for now [11:25:49] but yeah we can move it to an animated gif [11:26:11] ...I could be off base [11:26:12] :) [11:26:13] miketaylr? [11:26:14] scottjehl: I picked that up from somewhere [11:27:18] http://www.opera.com/docs/specs/presto23/css/transitions/ [11:27:26] * miketaylr looks [11:27:47] yeah oTransitionEnd [11:27:49] I had this in something I was working on, but maybe mine is the one that's incorrect: webkitTransitionEnd oTransitionEnd transitionend [11:27:51] not sure if that is the right ref. [11:28:02] so lowercase? [11:28:05] k. so does Moz use non-prefixed? [11:28:07] OTransitionEvent is the interface that isn't exposed (by us) [11:28:15] todd_: yeah lowercase o [11:28:32] good you're here mike! [11:28:35] alright I'll change that [11:28:38] thanks miketaylr [11:28:43] ok, so we're good then [11:28:47] anything else gseguin? [11:29:05] nothing else that's in progress [11:29:06] ok [11:29:17] I can pick up bugs from my list after that [11:29:19] so scott - you've been working on a few things... [11:29:25] scottjehl that is [11:29:54] yeah so I've added a couple branches [11:29:55] let's see [11:30:35] this branch removes ajax-loaded pages after hide https://github.com/jquery/jquery-mobile/commit/2265330677ca4ce1b3808112a7c5952cd555ca5e [11:30:39] (looks like window.otransitionend even works http://quirksmode.org/opera/transitionevent.html) [11:31:07] the commit note pretty much covers it [11:31:43] nested listviews are treated as a set, so all nested pages are removed when the parent page is hidden [11:32:01] other than that, it's just an option that keeps the DOM to usually 2 pages at any given time [11:32:05] or 1 [11:32:19] cool [11:32:23] it relies on the assumption that the browser cache can be trusted [11:32:27] for re-requests [11:32:35] so there has been talk about just relying on the browser's cache [11:32:45] is that for later? [11:32:49] depending on device memory constraints, it may be flaky [11:32:52] no that's this [11:32:54] or i guess this woudl leverage that [11:32:57] yep [11:32:57] wouldn't it? [11:33:16] so if we nuke a page, the browser could pull the original out of cache if re-requested [11:33:17] so it adds an option to the page plugin: [11:33:29] ajaxDomCaching: false [11:33:45] that is the default? false = remove on hide [11:33:48] false means remove an ajax-loaded page from DOM after hide [11:33:52] right [11:33:53] yep [11:33:58] true = the way it works today [11:34:14] yep. We could recommend only enabling it on a per-page basis [11:34:35] ie "this page is super heavy to build and we don't want to do it over and over" [11:34:36] or whatever [11:35:24] so this can be turned on via a data-attr on the page container? [11:36:52] data-ajax-dom-caching=”true” on the page container [11:38:00] yep [11:38:10] what's need to land this? [11:38:11] k. there are more branches [11:38:13] needed [11:38:18] umm UNIT TESTS [11:38:30] anyone want to help write tests?? [11:38:42] for? [11:38:52] this one will be simple to test [11:38:58] some of these new features scottjehl is working on [11:39:06] probably 2-3 tests [11:39:27] I can help with that once I'm done with the transitions [11:39:28] I can handle that one [11:39:34] ok [11:39:38] so next one... [11:39:39] here's another branch: https://github.com/jquery/jquery-mobile/commits/prefetch [11:40:13] this one's pretty simple. add a data-prefetch attr to any anchor element and jQM will load it [11:40:17] but not change to it [11:40:39] it's basically a markup shortcut for $.mobile.loadPage [11:40:58] I'd wanted to implement on top of native features for this, but they're scattered [11:41:12] so far, Firefox has had link[prefetch] for a long time [11:41:33] Chrome apparently just implemented a different way altogether [11:42:00] so this is jQM-specific. pretty harmless, but hopefully it'll wean people off of multi-page documentes [11:42:05] documents [11:42:54] that just needs tests too? [11:42:57] so, again this needs tests ya [11:43:02] there's also the auto-init branch, which basically allows devs to delay the auto-init call until whenever they want to call it: https://github.com/jquery/jquery-mobile/commits/autoinit-option [11:43:50] this is often requested by people who are serving up an empty body tag and js templating the dom. Not recommended by any means, but this option lets people do whatever they want [11:44:09] this also would bring the ability enhance a chunk of content, like page does [11:44:17] no [11:44:22] that's the decoupled widgets branch [11:44:23] oh [11:44:24] :) [11:44:27] last one [11:44:28] heh [11:44:36] too many branches [11:44:37] :) [11:44:38] yep [11:44:53] ok, getting ahead of myself [11:44:55] they're all fairly up-to-date with master [11:45:00] so that needs tests too [11:45:26] decoupled widgets is one that isn't fully baked yet. Tough to get the order right with theme inheritance [11:46:12] url for that one? [11:46:14] I've also got a secret branch locally that enables dragging between pages, previous and next. Hope to push sometimes soon, but not pull into master until post-1.0 [11:46:24] https://github.com/jquery/jquery-mobile/commits/decoupled-widgets [11:46:30] was the other one [11:47:13] should we shoot for the first 3 in for beta 2 in 2 weeks? [11:47:19] decoupled would be later [11:47:21] sure. oh there's pushstate too [11:47:24] but for 1.0 [11:47:26] heh [11:47:27] heh [11:47:30] geez [11:47:35] wow [11:47:40] yeah. That one's working well [11:47:47] now in Opera too! [11:47:49] :) [11:47:55] cool [11:47:56] url? [11:48:03] sorry, taking notes [11:48:55] https://github.com/jquery/jquery-mobile/commits/pushstate-integration [11:49:04] this one actually has a problem in iOS [11:49:14] that's the outstanding issue to work out [11:49:20] address bar starts to hang [11:50:41] that's why we need test help [11:50:49] we can let scott bang out the features [11:51:21] well, that's a big lis [11:51:23] list [11:51:38] can you automate that kind of testing? [11:51:48] not sure [11:51:50] which kind? [11:51:54] I'm afraid not [11:52:05] the address bar thing on iOS [11:52:09] when _nickel is back, it will flow [11:52:11] the addr bar? Oh it's a history issue. I'll fix it [11:52:17] oh ok [11:52:17] ah, gotcha [11:52:25] a few more items i want to cover... [11:52:26] mala opened this issue June 29, 2011 [11:52:26] XSS with XHR level2 cross domain request [11:52:26] https://github.com/jquery/jquery-mobile/issues/1990 [11:52:27] it's actually breaking history in some situations in ios [11:52:30] oh ya [11:52:40] scottjehl: gotcha [11:52:43] gseguin: you worked out a fix right? [11:52:43] i want to land this this week if we can [11:52:46] todd_: I closed it [11:52:46] seems bad [11:52:53] it [11:52:55] as in, pulled it into master? [11:52:59] it's landed [11:53:02] sweet [11:53:05] ah, awesome! [11:53:28] k, next [11:53:31] IE9: Border-Radius of Lists/Buttons [11:53:32] https://github.com/jquery/jquery-mobile/issues/1933 [11:53:45] might need to just blog about this [11:54:12] the gist: WP7 doesn't display our gradients now, but would if we added some tweaks [11:54:21] are we pulling -ms gradient out? [11:55:47] and Mango (IE9) has serious rendering issues [11:55:52] with corners and filter gradients [11:56:10] so current WP7 doesn't show gradients, Mango chokes on them [11:56:13] ...and they won't fix it [11:56:23] so I think we shoudl strip all the -filter gradients out [11:56:52] agreed. unfortunately. aim for gradients in IE10. [11:57:41] IE 10 will support standard css gradients [11:57:49] so i want to blog about this [11:57:52] first [11:58:14] but the end result won't look much different except IE( will have clean corners [11:58:24] gseguim - whatcha think? [11:58:58] so you want to remove the filter: [11:59:01] gseguin [11:59:13] yep [11:59:22] that will break gradients in IE9, but will work on IE10 [11:59:25] right? [11:59:26] since it doens't actually do anything right now but cause rendering issues [11:59:40] IE10 will pick up the non-prefixed gradient CSS syntax [11:59:42] that sounds good [11:59:45] ok, cool [11:59:53] i can do this after i blog about it this week [11:59:58] ok [12:00:05] ok, question about this issue: [12:00:06] cannot feed variable into $.mobile.path.parseUrl [12:00:06] https://github.com/jquery/jquery-mobile/issues/2016 [12:00:46] never mind, bender closed this with a fix [12:00:48] nice [12:01:02] fix issue 1851 as suggested - Checkbox list with same name do not allow multiple selection [12:01:02] https://github.com/jquery/jquery-mobile/pull/1911 [12:01:47] I think this one was in good shape [12:01:53] we can look into merging it [12:01:54] "This pull request can not be automatically merged." [12:01:56] ok, so can you pull in? [12:01:59] fun [12:02:26] next [12:02:27] #1921 listview filtering/search config [12:02:27] https://github.com/jquery/jquery-mobile/pull/2006 [12:03:25] idea of having a data-filter-type attribute with possible values being start, contains, end, and regex -- start would be the default if the data-filter-type attribute is not explicitly defined in the markup. [12:04:28] this seems like a decent idea [12:04:32] ya [12:04:57] * gseguin still reading [12:05:48] yep [12:05:48] ignore the first stuff...it's way too narrow and i said "no" [12:06:20] sound ok? i just replied and said "code it if you want" [12:06:23] next [12:06:23] fix: swipe event fired during vertically scrolling [12:06:23] https://github.com/jquery/jquery-mobile/pull/2031 [12:07:11] https://github.com/wtetsu/jquery-mobile/commit/f95f1202af1b3eebc0d1c032a56ee5d6cea914cc [12:07:47] document.body.scrollTop will need some normalizing and qualifying for one [12:08:10] I can take a look. I'm working on swipe related stuff anyway [12:08:48] k [12:09:27] next [12:09:27] Updated nested list pages to show back button [12:09:27] https://github.com/jquery/jquery-mobile/pull/2030 [12:10:46] yeah that one I tried his first version and didn't like it and then he closed it [12:11:31] hmm i don't think this is a back button issue [12:11:42] it's more like nested lists need a custom breadcrumb nav [12:11:45] or something [12:12:04] not sure. nested lists are such a pain [12:12:08] so what he wants is if you're deeplinked into a nested list, have a way to go back to the parent, right? [12:12:17] ya [12:12:19] we shoudl support that [12:12:31] but not worry about showing a back button [12:12:39] unless you turn that feature on globally [12:12:54] if you deep link to a node in a nested list, we should bounce you to the top, right? [12:13:03] might be simplest [12:13:22] we can't know if there is a place for a button in the toolbar [12:13:24] what's the point of deeplinking then? [12:13:26] might not even be a toolbar [12:13:41] not following [12:13:42] in this special case, there isn't [12:13:51] this is just to recover [12:14:01] break all deep links to nested lists? [12:14:02] so you don't end up in a bad place [12:14:20] they are all part of the same "page" [12:14:42] feel like every time we talk about nested lists, things get ugly [12:14:46] let's skip thi sone [12:14:55] i just don't care about 'em [12:14:56] heh. they're so abused [12:15:00] yup [12:15:11] maybe it we don't support them well, people will move away from it [12:15:24] next [12:15:26] Enable changing of loading indicator message on runtime [12:15:26] https://github.com/jquery/jquery-mobile/pull/1974 [12:15:47] I think we just need to tie up a loose end on getting back to top, then decouple nested lists from the listview plugin and make them an optional download [12:15:56] yeah [12:15:59] but later [12:15:59] and advocate not using them [12:16:07] so low in the priority for me [12:16:19] so do we want to support custom loading msgs? [12:16:22] seems pretty clean [12:17:04] tho this seems to repeat the markup twice [12:17:09] not dry [12:17:12] where is his customMessage var coming from? [12:17:17] but as a feature? [12:17:29] feature sounds like a fine idea to me [12:17:36] same here [12:17:39] k, so as an implementation? [12:17:41] should be doable with minimal code weight additions [12:17:52] this doesn't look quite like what we'd want [12:18:01] maybe a new private function that spits out the markup [12:18:06] 1 arg for the message text [12:18:23] anyway, the idea's good [12:18:40] ok [12:18:50] shoudl we say, good idea, but we'll re-write this [12:18:58] is there an issue for it? [12:19:22] don't think so [12:19:31] * gseguin creating one [12:20:16] hey, found another pull for the same feature! [12:20:16] https://github.com/jquery/jquery-mobile/pull/1947 [12:20:32] https://github.com/dgeb/jquery-mobile/commit/a6de24b15ac8078673fd8c10825c576a00a16c33 [12:21:00] woah, seems over engineered [12:21:27] delay, fadetime, wrapper class, [12:21:44] theme might be a good idea [12:21:48] swatch that is [12:22:02] ultimately, we need a message overlay plugin that the loader will use [12:22:20] right [12:22:53] created https://github.com/jquery/jquery-mobile/issues/2044 [12:22:58] so for now, does it make sense to make a small tweak to be able to passin a different message [12:23:10] cool thx [12:23:13] or re-work this to be a message overlay that we use for the loader [12:23:21] go for the more generic thing now [12:23:22] small tweak [12:23:24] k [12:23:30] tweak it is [12:23:41] plugin will turn into the dialog everyone uses [12:23:57] and then we'll need to get plugins working in it, etc [12:24:03] so, ya, simple for now [12:24:04] :) [12:24:10] yeah [12:24:21] ok, gseguin post the link to the issue when done [12:24:25] oops, did it [12:24:40] :) [12:25:08] ok, closed both and linked to the issue [12:25:10] go team [12:25:26] ok, last item [12:25:27] Feature Request: "pagecreate"-Event on body [12:25:27] https://github.com/jquery/jquery-mobile/issues/1972 [12:26:03] not a pull, but looking for an opinion on this [12:26:03] umm [12:26:08] there's already an event [12:26:13] called pagecreate [12:26:25]