**** BEGIN LOGGING AT Thu Oct 24 02:59:58 2013 Oct 24 05:37:27 LarrySteeze: pong Oct 24 07:38:50 Tofe|Away: ping Oct 24 08:06:39 morphis: pong Oct 24 08:06:53 Tofe: looked a little bit more through maliit Oct 24 08:07:06 and it looks like the nemo keyboard and the maliit one go different ways Oct 24 08:07:16 ah, sh*t Oct 24 08:07:35 the maliit (and the ubuntu one) are resizing the keyboard window to just take the space it needs Oct 24 08:07:42 the nemo one use the whole screen Oct 24 08:07:56 as we saw, then Oct 24 08:08:00 the maliit one also create more than one overlay Oct 24 08:08:13 and also set a position for each overlay window Oct 24 08:08:37 ah, like, for the key preview, it creates a little additional window Oct 24 08:08:55 yes or the magnifier Oct 24 08:09:18 Tofe: https://github.com/maliit/plugins/blob/master/maliit-keyboard/plugin/inputmethod.cpp#L368 Oct 24 08:09:51 Tofe: so I think, we should just take the overlay as it is (not resize it or apply any alpha to it) and render it Oct 24 08:10:09 I am currently reactivating the maliit keyboard to see if it works Oct 24 08:10:30 well, we are still not very good with handling multiple windows, are we Oct 24 08:10:42 multiple overlay windows? Oct 24 08:10:56 for instance, yes Oct 24 08:11:06 I think we have to fix this then Oct 24 08:11:26 ok, agreed. Oct 24 08:12:10 Tofe: you've time to work on this in tonight? Oct 24 08:12:36 mmmh not tonight, I've got a social appointment :) Oct 24 08:13:03 but then tomorrow evening, and most of the week-end if free Oct 24 08:13:17 is* Oct 24 08:13:49 anyway I can begin to sketch a prototype, implementing in QML is quick when you know what to do Oct 24 08:14:09 for example, we must not force the position of maliit's window Oct 24 08:15:08 we may as well not even use a wrapper window Oct 24 08:16:58 Actually, that might be a good first try: not handling the overlay windows. Oct 24 08:18:01 :) Oct 24 08:18:24 Tofe: I am will wrap up a new build today with several components bump'ed and the maliit keyboard back Oct 24 08:18:40 ok, good Oct 24 08:21:15 and if you want to try my little idea, you just have to insert a test in __handleWindowAdded: do nothing if window.windowType==WindowType.Overlay Oct 24 08:23:09 ok Oct 24 08:23:54 I think it might work, if the windows are set correctly on top of others Oct 24 08:24:17 maybe we just have to define an overlayer layer item and set the parent correctly Oct 24 08:24:26 without setting the anchor Oct 24 08:25:42 something like that Oct 24 08:28:58 We have to be careful, maliit already does some parenting between its various surfaces Oct 24 08:30:02 It looks like it would work if we reparent only the surfaces that don't have a parent yet Oct 24 08:33:06 ok Oct 24 08:33:57 but anyway, even without bothering about the z of the surface, I guess it would show up correctly at least the first time Oct 24 08:42:54 lets see Oct 24 08:44:50 Note that in the future, this overlay layer could make a good candidate for some widgets or similar Oct 24 08:45:30 But the completely free placement can also be a bit dangerous, so let's do one this maliit thing first Oct 24 09:13:35 Tofe: for sure Oct 24 09:13:42 Tofe: ok, got the maliit-keyboard working Oct 24 09:13:57 first time I touched the input field in the just type app it showed up Oct 24 09:14:01 on top of the screen Oct 24 09:14:14 used the fix you mentioned for __handleWindowAdded Oct 24 09:14:56 great! Oct 24 09:15:09 the keyboard works well ? Oct 24 09:15:29 default seems to be not the english one Oct 24 09:15:42 can you hide it with swipe down on the keyboard ? Oct 24 09:16:00 no Oct 24 09:16:03 it stays Oct 24 09:16:12 ok Oct 24 09:16:20 Tofe: I will prepare everything for you Oct 24 09:16:30 so you can get it working correctly Oct 24 09:16:51 :) ok thanks Oct 24 09:16:56 btw. the thing why maliit died when using the maliit-keyboard was because of it setting a wrong surface format Oct 24 09:17:07 which wasn't accepted by our EGL driver Oct 24 09:17:19 ah, ok, makes sense Oct 24 09:17:23 so it could never allocate a EGL context which leads qt to segfault Oct 24 09:17:39 a bit rough, but logical Oct 24 09:17:48 yes Oct 24 09:18:04 it was setting a different surface format to make the view transparent Oct 24 09:18:16 makeQuickViewTransparent Oct 24 09:18:18 lets see if its needed for any purpose or if we can just drop it Oct 24 09:18:18 yes Oct 24 09:19:51 as far as I understand, it tries to get an ARGB view, which we already provide anyway Oct 24 09:20:54 ok, but we seem to be on a good track with that keyboard Oct 24 09:22:26 :) Oct 24 09:24:07 if not english layout, what would it be ? Oct 24 09:40:14 soem cyrilic Oct 24 09:42:47 well, let's learn russian... Oct 24 09:43:21 ... or change the default layout. Oct 24 10:09:00 My wife would be happy with Russian LOL ;) Oct 24 12:52:23 Tofe: ok, new builds are queued Oct 24 12:53:34 yepee Oct 24 12:55:03 I can see the task 291, I'll download the image when it'll be finished Oct 24 13:03:52 make sure you mention when it's done Oct 24 13:03:55 so i can grab it too :) Oct 24 13:26:10 HaDAk: for sure :) Oct 24 13:26:33 yesterday morning's build looked pretty sexy. i'm hoping this build includes the keyboard. Oct 24 13:27:10 Tofe: https://github.com/webOS-ports/luna-next-cardshell/pull/33 Oct 24 13:27:21 HaDAk: it does, but not really usable Oct 24 13:28:17 HaDAk: and I guess it doesn't include the mini-fix I proposed to show maliit-keyboard overlay Oct 24 13:28:48 as long as a keyboard pops up, i'm happ(y/ier) Oct 24 13:29:18 morphis: I'll see if I can move all those "utility" QML components in one place, so that main.qml stays a bit consistent Oct 24 13:37:39 morphis: the task failed for libmaliit-keyboard-plugin Oct 24 13:38:21 A tiny little missing Q_UNUSED(view); I'd say Oct 24 14:16:41 Tofe: no, the build failed cause of something different Oct 24 14:16:49 ah, ok Oct 24 14:18:44 morphis: I didn't see it when I merged, but is it normal that in Preferences.qml, in setPreference, you call subscribe ? For setting a preference value ? Oct 24 14:23:40 yes Oct 24 14:23:46 ah stop Oct 24 14:23:57 ah no, thats wrong Oct 24 14:25:14 Tofe: https://github.com/webOS-ports/luna-next-cardshell/pull/34 Oct 24 14:27:47 morphis: for maliit, I guess the keyboard appears on top of the gesture area ? Oct 24 14:28:02 yes Oct 24 14:28:28 ok Oct 24 14:48:09 morphis: is it me, or maliit always makes directly reference to the screen for positioning itself ? Oct 24 15:47:24 Tofe|Away: what do you mean with "directly references to the screen"? Oct 24 15:54:51 http://www.theverge.com/2013/10/24/5023912/hp-looking-to-sell-webos-and-other-mobile-patents-says-bloomberg Oct 24 16:03:44 Yup read that earlier today... Well HP doesn't use it, so why not sell it..... Wondering who's interested (if anyone).... Oct 24 16:17:43 HaDAk: new emulator build has finished Oct 24 16:18:06 but afaik not yet sync'ed to build.webos-ports.org Oct 24 16:28:13 Herrie, :/ Oct 24 16:46:15 morphis: oh, emulator? i thought it was grouper Oct 24 16:50:21 HaDAk: grouper is still building Oct 24 16:50:49 kk Oct 24 17:00:03 HaDAk: btw. can you add your install script to https://github.com/webOS-ports/webos-ports-sdk? Oct 24 17:04:53 morphis: not until it's been tested and approved :) Oct 24 17:05:18 i mean, it's "good" right now, but i need other people to test it and make sure it works for them Oct 24 17:05:24 nobody has done that yet Oct 24 17:05:32 ok Oct 24 17:10:38 this is the latest version. just made a small, untested change. shouldn't be a problem, though. http://pastebin.mozilla.org/3327003 Oct 24 17:10:56 maybe run it through a few times? test with downloading both kernel and image, then just kernel, then just image? Oct 24 17:17:18 Like your ASCII art LOL ;) Oct 24 17:25:04 :) Oct 24 17:29:04 anyway. test it. try to break it. let me know if/how you do. Oct 24 18:23:58 anybody have a chance to test it? Oct 24 18:25:47 HaDAk: sorry, not yet Oct 24 18:25:55 no worries. Oct 24 18:32:07 JaMa|Off: ping Oct 24 18:47:51 HaDAk, Tofe|Away: new images are sync'ed now Oct 24 18:48:01 sweet. Oct 24 18:48:05 kernel update, or no? Oct 24 18:48:10 i'm guessing not Oct 24 18:48:34 no Oct 24 18:48:40 awesome. Oct 24 18:49:01 imma upgrade now Oct 24 19:15:34 hum. this is taking too long. it should be done by now. Oct 24 19:15:37 i think it failed. Oct 24 19:20:40 HaDAk: it really failed? Oct 24 19:20:57 yeah. i'm going to try again. Oct 24 19:21:13 no worries. it *may* have been the script. Oct 24 19:21:18 i found a little bug i'm sorting out right now anyway Oct 24 19:21:53 ok Oct 24 19:21:56 I am off Oct 24 19:21:56 gn8 Oct 24 19:34:43 aw, no keyboard Oct 24 19:34:44 sad day Oct 25 00:51:42 morphis: I think I need to move to europe to talk to you. **** ENDING LOGGING AT Fri Oct 25 02:59:58 2013