**** BEGIN LOGGING AT Tue Jan 28 02:59:59 2014
Jan 28 08:13:08 <Garfonso> morning
Jan 28 10:09:14 <morphis> Herrie: I already bumped the version of the pdf, filemanager, ... apps
Jan 28 10:09:27 <morphis> so they should arrive as a recent version in a build soon
Jan 28 10:18:59 <Herrie|Veer> morphis: Thanks, sent some more suggestions to Roshan's way after test driving it for a bit :)
Jan 28 10:19:09 <Herrie|Veer> That's for PDF app :)
Jan 28 10:19:18 <morphis> ok
Jan 28 10:19:23 <Herrie|Veer> Back on the Veer :)
Jan 28 10:20:00 <Herrie|Veer> Couldn't stand laggy music player & buggy wIRC on Pre 3.
Jan 28 10:25:20 <Herrie|Veer> Hopefully will have some time in the weekend to play with Ares and mockup some bits for a phone app :)
Jan 28 10:41:39 <morphis> Herrie|Veer: I am not really if we shouldn't create the phone in QML
Jan 28 10:41:45 <morphis> just for speed reasons
Jan 28 10:45:37 <Herrie|Veer> Not sure, it works fine on legacy in legacy Enyo. I'm sure speed should be no issue in Enyo 2 seeing hardware specs are better too.
Jan 28 10:46:02 <Herrie|Veer> Phone app on my 800 mhz Veer in Enyo runs fine as long as my call history doesn't get too long.
Jan 28 10:46:36 <Herrie|Veer> Speed of Enyo 2 is a lot better compared to legacy Enyo for what I've seen so far.
Jan 28 10:49:07 <Herrie|Veer> Advantage of Enyo apps is that patching is easier for patchers and devving might be easier too?
Jan 28 10:54:13 <Herrie|Veer> When we get some of the old dev's back on board they might be happier to work on Enyo2 instead of QML due to familiarity?
Jan 28 11:15:48 <bhuey> Why's Enyo2 faster ?
Jan 28 11:21:41 <Herrie|Veer> Minified and better optimized it seems :)
Jan 28 11:22:23 <Herrie|Veer> All Enyo 2 apps I tested so far run smooth like butter on legacy webOS where legacy Enyo is laggy sometimes
Jan 28 11:26:23 <morphis> Herrie|Veer: hm
Jan 28 11:26:42 <morphis> when someone gets a enyo based version up we can start with that
Jan 28 11:28:22 <Herrie|Veer> You have any reasons for QML? Besides being native?
Jan 28 11:54:00 <morphis> speed :)
Jan 28 11:54:24 <morphis> currently we have the problem web apps using a lot memory and taking some time to launch
Jan 28 11:59:29 <Herrie|Veer> Is that due to the ARM JIT bug in WebKit? Or something else
Jan 28 11:59:32 <Herrie|Veer> ?
Jan 28 12:22:58 <Tofe> morning
Jan 28 12:23:30 <morphis> Herrie|Veer: no, it's a general problem with starting a huge process :)
Jan 28 12:23:38 <Tofe> morphis: I fear I didn't fully understand what should be done with the service file in the webos-keyboard source tree; should I modify the CMakeLists ?
Jan 28 12:24:03 <morphis> Tofe: morning
Jan 28 12:24:16 <morphis> Tofe: no, just put it into service/ folder
Jan 28 12:24:26 <morphis> the webos_system_bus.bbclass will pick it from there
Jan 28 12:24:26 <Tofe> that's done, I updated the PR
Jan 28 12:24:29 <morphis> thanks
Jan 28 12:24:42 <morphis> Herrie|Veer: libQtWebKit.so.5.2.0 is 25 MB huge
Jan 28 12:25:04 <morphis> when you start webapp-launcher it works off a new process where the processing will happen (the WebProcess)
Jan 28 12:25:11 <morphis> all this needs time and memory
Jan 28 12:25:42 <Herrie|Veer> Morphis: OK but that shouldn't be much of a problem? Or we need to start 1 instance for each app?
Jan 28 12:25:59 <morphis> Herrie: thats what we're currently doing
Jan 28 12:26:05 <morphis> but I fear that was the wrong decision
Jan 28 12:27:07 <Tofe> morphis: for the luna-init PR, can I just do a PR of your commits ? :p
Jan 28 12:27:25 <Herrie|Veer> That gets costly in resources quickly... might need reworking in future? I guess the model Palm had was to avoid this?
Jan 28 12:27:25 <morphis> yes
Jan 28 12:27:35 <morphis> yes
Jan 28 12:27:46 <Herrie|Veer> With the browser-server etc?
Jan 28 12:29:15 <morphis> no
Jan 28 12:29:24 <morphis> classic webOS uses webkit1
Jan 28 12:29:31 <morphis> which doesn't have a multi process modell
Jan 28 12:29:42 <morphis> there is just one webappmanager process doing all the web work
Jan 28 12:34:57 <Tofe> morphis: uh... but wait, the commits are already there for luna-init/webOS-ports/master
Jan 28 12:35:31 <morphis> Tofe: hm, looks like I pushed them already :)
Jan 28 12:35:43 <Tofe> ok, well, that's done !
Jan 28 12:35:54 <Tofe> just meta-wop is left
Jan 28 12:36:14 <morphis> ok
Jan 28 12:36:26 <morphis> Tofe: what do you think about reworking the preference key structure?
Jan 28 12:36:41 <morphis> currently its a huge object you have to always full provide
Jan 28 12:37:02 <morphis> using something like "keyboard.enabledLanguages" would be better imho
Jan 28 12:37:45 <Tofe> morphis: yes, that would be convenient, and could even prove itself useful for some usecase
Jan 28 12:38:04 <Tofe> (concurrent access scenarios)
Jan 28 12:38:19 <morphis> right
Jan 28 12:38:38 <morphis> and from what I remember if we just provide the keyboard object with one child all others get overriden
Jan 28 12:48:34 <Herrie|Veer> Ah ok. WebKit2 has multi process :)
Jan 28 12:50:38 <morphis> right
Jan 28 12:50:56 <morphis> I think we have to switch back to the same model soon
Jan 28 14:04:26 <Garfonso> hm.. but shouldn't the library be shared, i.e. memory consumed only once?
Jan 28 14:09:13 <Tofe> yes, most of it is shared, hopefully.
Jan 28 14:25:51 <morphis> Garfonso: right
Jan 28 15:25:02 <JaMa> morphis: I'm pushing new tags and -next with qtmultimedia_5.1.1 and db8 upgrade (not sure if it will fix my qtwayland build issue, but it should rebuild it again at least)
Jan 28 15:27:02 <morphis> ok
Jan 28 15:27:42 <morphis> JaMa: with the patch for webos-systemd-services included?
Jan 28 15:29:12 <JaMa> y
Jan 28 15:29:26 <JaMa> I've pushed it to master-next and will bump SRCREV
Jan 28 15:29:44 <morphis> ok
Jan 28 18:33:30 <morphis> Garfonso: do you had success with setting up the build env for the mediaindexer?
Jan 28 19:11:29 <Garfonso> MACHINE=qemux86 bitbake -c compile -f mediaindexer runs without error.. but I could not find out where results are created. ;)
Jan 28 19:13:14 <Tofe> look in tmp-work
Jan 28 19:15:11 <Tofe> I have a poor memory tonight , but should be in work/qemux86/mediaindexer-....
Jan 28 19:15:50 <morphis> Garfonso: look into the source folder you added in conf/local.conf
Jan 28 19:16:13 <morphis> I have:
Jan 28 19:16:20 <morphis> S_pn-mediaindexer = "/home/morphis/work/wop/work/mediaindexer"
Jan 28 19:16:43 <morphis> in that folder the binary is BUILD-armv7a-vfp-neon/mediaindexer
Jan 28 19:16:56 <morphis> for qemux86 it should be BUILD-i586/mediaindexer
Jan 28 19:21:35 <Garfonso> ah, ok. Makes sense.
Jan 28 19:23:55 <Garfonso> I see a folder, but it is empty
Jan 28 19:52:20 <Garfonso> ok, now it builds the mediaindexer binary.. not sure what was the issue.
Jan 28 19:52:21 <Garfonso> :)
Jan 28 20:04:03 <morphis> Garfonso: https://github.com/webOS-ports/mediaindexer/commit/8d5d9c21f4cf13f97b989d0ef08cb4317bc5fb67
Jan 28 20:04:08 <morphis> some code I was playing with today
Jan 28 20:04:20 <morphis> the QFileSystemWatcher has some disadvantages
Jan 28 20:04:42 <morphis> we only get a notification when something in a directory has changed but nothing more
Jan 28 20:04:49 <morphis> so we have to do a full rescan
Jan 28 20:05:08 <Tofe> morphis: the notification doesn't contain any information on the change ?
Jan 28 20:07:39 <morphis> no
Jan 28 20:07:51 <morphis> it's just a signal directoryChanged(path)
Jan 28 20:07:59 <morphis> and I don't want to register a watch per file
Jan 28 20:08:27 <morphis> as there is a maximum of watches for inotify
Jan 28 20:09:09 <morphis> but a rescan per directory should be ok for now
Jan 28 20:09:22 <morphis> we just need to make sure it happens only once in a specific interval
Jan 28 20:10:43 <Tofe> yep, and you can always register a watch per directory, recursiverly
Jan 28 20:11:06 <Tofe> (I guess that's what you do already anyway)
Jan 28 20:15:06 <Tofe> oooh c++11 !
Jan 28 20:19:34 <Tofe> morphis: https://github.com/webOS-ports/meta-webos-ports/pull/18 <-- I hope I didn't forget anything ?
Jan 28 20:42:54 <JaMa> morphis: fwiw: new build from scratch failed in qtwayland the same as before
Jan 28 20:43:07 <JaMa> morphis: don't you have newer/older wayland in your builds?
Jan 28 20:43:25 <JaMa> I've already checked morphis/work branch and it's doesn't seem so
**** ENDING LOGGING AT Wed Jan 29 02:59:58 2014