**** BEGIN LOGGING AT Mon Mar 09 02:59:58 2015 Mar 09 07:38:59 morning Mar 09 07:39:17 I can confirm that latest image doesn't seem to boot on Tenderloin Mar 09 07:39:22 morphis: ping Mar 09 08:12:24 morning Mar 09 08:14:41 btw: MoboReader works quite good in latest testing builds. Even spritzing, which was a bit slow in October or November. Did we change something there? It uses a canvas and draws text on that very often. Mar 09 08:27:17 Garfonso I think nizovn fixed some performance bug in Luna Next Cardshell Mar 09 08:27:34 Specifically in the system menu Mar 09 08:28:32 Which was eating CPU especially noticeable in the emulator I think Mar 09 08:29:31 Herrie|Veer: no, i failed to fix it. Mar 09 08:29:33 morning Mar 09 08:30:43 that should be more on the side of hardware acceleration in the web app rendering stuff or something like that. Maybe my last tries were even longer ago. Not sure. ;) Mar 09 08:31:21 but when I last tried, the canvas updates were very laggy and worse than on a N4 than on my Veer, which was a bit sad. ;) Not it works quite good. Mar 09 08:32:06 The system menu was eating cpu cycles even when not visible and there were a few others as well that were fixed Mar 09 08:33:22 morning Mar 09 08:33:39 Herrie|Veer, Garfonso: qt improved its rendering performance with 5.4 a lot Mar 09 08:34:00 that is why you're seeing such a better performing UI in contrast to how it was perform Mar 09 08:34:07 cardhsell doesn't have much influence here Mar 09 08:34:21 that makes much more sense Mar 09 08:35:20 Ah ok :) Mar 09 08:35:33 morphis any idea what broken Tenderloin build? Mar 09 08:36:02 Changes that were done don't look like they should cause this that obviously? Mar 09 08:38:02 Herrie|Veer: as I said, its my fault after updating the header files Mar 09 08:41:53 now I just have to figure out sometime, why MoboReader has one line of white pixels below its bottom toolbar on LuneOS, but not on webOS. :-/ Mar 09 08:42:15 Garfonso: possible that this comes from the cardshell Mar 09 08:42:28 but by default the webview has a white background so propably not Mar 09 08:42:45 it is visible also in card mode. Mar 09 08:43:53 also what confuses me is that Preware 2 (just one example for enyo2 app with bottom toolbar) does not have this.. Mar 09 09:02:36 morphis: Ah must have missed the msg about header Mar 09 09:03:47 any qml devs awake? :-) Mar 09 09:04:10 Me but wouldn't call myself a QML dev :P Mar 09 09:04:16 Tofe is way better LOL :P Mar 09 09:04:37 morphis: Can't find any msg about header in the logs :P Mar 09 09:04:46 I guess it never made it into the channel somehow Mar 09 09:05:05 i need a quick review from someone who can tell me what boneheaded mistake i'm making :-) Mar 09 09:05:23 Well I can try :P Mar 09 09:05:39 it's probably not good to read code on a Veer. I'd like you to remain sighted. :-) Mar 09 09:05:58 I have access to a laptop :P Mar 09 09:09:05 morphis: We'll skip this month's release? So we can get Alan to get a note out for that? Mar 09 09:18:39 Hm.. somehow I can't select html-nodes in webkit inspector right now. If I hover it, the node gets highlighted on device. But if I click, nothing happens. Body always stays selected. Clicking on other stuff works. Not sure what could be wrong. Mar 09 09:20:46 ah.. there is a js exception in treeoutline.js, line 1063, "undefined is not a function". Hm. Mar 09 14:49:39 Herrie: it's in meta-smartphone Mar 09 14:49:41 Herrie: and yes, lets skip the release for this months Mar 09 14:49:53 Garfonso: for me navigating with arrow-keys works Mar 09 14:58:11 nizovn: which build did you test that with? Mar 09 14:59:15 reboot on TP? 20150307210040-testing-0-114 Mar 09 15:02:11 yeah Mar 09 15:08:35 nizovn: triggered a new round of builds which should be fixed and run on TP now Mar 09 15:08:51 great Mar 09 15:08:59 nizovn: https://github.com/webOS-ports/webos-ports-setup/commit/abfecce1cade39fe1520d716b1d6c8002dc3c928 Mar 09 15:09:02 that one was missing Mar 09 15:09:37 JaMa: btw. anything from your side against switching to HEAD for meta-smartphone in testing? Mar 09 15:09:59 nizovn: btw. you've found what is wrong with ALS and display management? Mar 09 15:12:49 no, i waited for new build, as it works good on my aged, and i broke my build environment Mar 09 15:18:01 :) Mar 09 15:18:11 sorry for that Mar 09 15:22:45 morphis: no Mar 09 15:23:37 JaMa: ok, then let me switch that soon Mar 09 16:51:09 morphis: thanks, that works. Any idea what is the root issue? Maybe some version mismatch between webkit-renderer and inspector? Mar 09 16:56:57 possible Mar 09 16:57:08 but qtwebkit is mostly out-of-maintenance now Mar 09 16:57:15 so not sure if upstream has a fix for that Mar 09 16:58:06 morphis, good call to wait on next release Mar 09 16:58:26 Hm.. it seems my MoboReader's panel is one pixel too short in height dimension. I can shift it by one pixel up/down then the white line shows on top/bottom. Issue is: Height is set to 100%. Mar 09 16:58:37 ka6sox: yeah Mar 09 16:58:45 Herrie: can you work with Alan to put out a note? Mar 09 16:59:12 morphis, Alan is working on this now Mar 09 16:59:19 ah great Mar 09 17:00:06 talking with him now about this Mar 09 17:01:14 panels object height is identical to window.innerHeight. Is it possible that there is a one off error there somewhere? Or do we have no control about all this? Mar 09 17:03:45 "one off error"? Mar 09 17:04:02 morphis: Yeah Alan is writing it already ;) Mar 09 17:07:49 it seams to me (with my limited knowledge about CSS and Browser stuff in general) that window.innerHeight is one pixel to few and that's where the white line comes from at the bottom. Because window.innerHeight is 577, element height also is 577, but there clearly is one row of pixels at the bottom that is not painted. Mar 09 17:08:52 hm Mar 09 17:09:01 maybe that is a problem from the QML side Mar 09 17:09:13 so I'd say window.innerHeight should be 578px. Mar 09 17:09:51 as I said, I can shift the element one pixel down, then the white line appears at the top and is gone at the bottom Mar 09 17:16:20 hm, but if that is a general problem we should fix that the general way Mar 09 17:16:27 can you have a look at luna-webappmanager? Mar 09 17:16:34 maybe you see something wrong there Mar 09 17:51:09 morning Mar 09 17:51:28 EricBlade: still need some help? Mar 09 18:10:20 btw. what are you guys thinking about starting a small app competion? Mar 09 18:10:58 mainly to get let us find out what we need to work on regarding to specific features apps need Mar 09 19:01:25 good Idea Mar 09 19:14:03 Tofe: Benjamin asked if it's possible to edit QML live, so without rebuilding & running for the LuneOS Components Mar 09 19:14:46 So on the fly basically and see the changes like you can with some CSS tools in the browser Mar 09 19:18:41 Herrie|2: for the theme ? Mar 09 19:20:05 For QML in general, apart when the resources are built in the final application (which we don't do iirc), only re-starting the app is needed. But a real live modification, no, that won't be possible Mar 09 19:21:04 Herrie|2: btw. please tell that it is absolutely fine to push minor changes Mar 09 19:51:54 Tofe: Yeah for theme, just that it would be quicker that way ;) But if not possible, then not possible ;) Mar 09 20:16:03 DougReeder: Sent the request for the icon to Benjamin Mar 09 20:16:52 Tofe: Benjamin pointed at: https://github.com/quiverproject/quiver Mar 09 20:16:57 You've ever toyed with that? Mar 09 20:17:24 Herrie|2: just tried https://github.com/penk/qml-livereload and it seems to work Mar 09 20:22:33 Ah our friend Penk ;) Mar 09 20:22:41 He used to hang out here too ;) Mar 09 20:22:49 He made a Open webOS Live CD sometime ;) Mar 09 20:23:10 wait, what? Mar 09 20:23:14 I wanted one of those Mar 09 20:23:20 I didn't realize anyone made one Mar 09 20:23:23 GodGinrai: He did, let me look it up ;) Mar 09 20:23:34 http://penkia.blogspot.nl/2012/10/unofficial-open-webos-livecdusb.html Mar 09 20:23:52 oh man Mar 09 20:23:55 neat Mar 09 20:24:24 It's old and still using old LunaSysMgr etc Mar 09 20:24:36 still, exciting :) Mar 09 20:24:53 thanks for the link Mar 09 20:41:12 nizovn: How to get that working exactly? Mar 09 20:48:42 Herrie|2: the GH app pointed by nizovn is a little app that will reset the QML view and reload the whole QML model in memory. Could be handy for testing, I guess. Mar 09 20:51:02 Herrie|2: The only issue with that is that the QML side is actually restarted. That would be like surveying the qml files of an app, and send a restart event to our qml-launcher when they are modified. Mar 09 20:53:11 Herrie|2: and quiver is exactly the same thing, only you can modify the qml path to load in the view. Mar 09 20:53:50 But eventually, integrating such inotify events of the launched qml app could be fairly easily integrated into our qml-launcher. Mar 09 20:53:58 Herrie, thanks. Mar 09 20:58:17 Quiver's way of doing that seems smarter Mar 09 21:16:02 Herrie|2: build it in qtcreator, put main.qml in build directory, execute livereload. but you need to replace applicationwindow for example with rectangle. as Tofe wrote, quiver may be better and seems doesn't need such editing, but livereload is a lot simpler and even i can understand how it works. **** ENDING LOGGING AT Tue Mar 10 02:59:58 2015