**** BEGIN LOGGING AT Fri Sep 15 03:00:02 2017
Sep 15 05:32:31 <Tofe> Herrie: http://build.webos-ports.org/halium-wop-12.1/ :)
Sep 15 05:52:41 <Herrie|TP> Tofe: Morning! OK so that means we build Halium for N4, N5 & Tenderloin now?
Sep 15 05:53:13 <Herrie|TP> I thought we needed some other changes as well to have it actually working?
Sep 15 06:52:39 <Tofe> Herrie|TP: well, it's built, it doesn't mean we use what's built :)
Sep 15 06:53:42 <Tofe> The minimal changes are updating the android-system-image recipe for N4 and N5, and also changing fstab a bit to replace the mount by a bind mount
Sep 15 07:04:02 <Herrie|TP> Tofe: That's what I meant ;-) For N5 this was more or less tested already you said right?
Sep 15 07:06:32 <Tofe> right
Sep 15 07:06:55 <Tofe> N4 I still need to test it, but there are good chances it'll work
Sep 15 07:07:43 <Tofe> We will also need to eliminate all references to my GH repo, to be cleaner
Sep 15 07:08:08 <Tofe> (i.e. in the android additional manifest wop_targets.xml)
Sep 15 07:46:36 <Herrie|Laptop> Tofe: Yeah that would be good to remove those references
Sep 15 07:46:52 <Herrie|Laptop> My N4 is occupied by work Android now, so cannot easily test things
Sep 15 07:51:04 <Tofe> I can test that tonight
Sep 15 08:21:08 <Herrie|Laptop> Tofe: :D
Sep 15 08:32:27 <Tofe> Herrie|Laptop: depending on how things go on N4 & N5 for this Halium thing, we'll include it in the release or not
Sep 15 08:33:23 <Herrie|Laptop> Tofe: If all OK I guess we should include it
Sep 15 08:33:28 <Herrie|Laptop> Tofe: https://git.merproject.org/mer-core/qtscenegraph-adaptation/commit/de6a1d9339d698e47e2e229394031e813d025d12
Sep 15 08:33:35 <Herrie|Laptop> Seems qtscenegraph is upstream now :)
Sep 15 08:35:02 <Tofe> ah, good
Sep 15 08:36:02 <Tofe> One patch to eliminate
Sep 15 08:36:06 <Herrie|Laptop> I'll do that
Sep 15 08:37:51 <Tofe> Thanks :)
Sep 15 08:39:12 <Herrie|Laptop> https://github.com/webOS-ports/meta-webos-ports/commit/1853c68a3b635f416f6a36e5bacc1598bc670138
Sep 15 08:39:30 <Herrie|Laptop> Tofe: Is it correct I don't get graphics yet on latest VirtualBox?
Sep 15 08:42:37 <Tofe> Herrie|Laptop: no, you should get graphics... oh wait, it's that driver conflict again...
Sep 15 08:43:38 <Tofe> Herrie|Laptop: there is a vbox* kernel module configuration file to remove, maybe in /etc/modprobe.d
Sep 15 09:08:48 <Herrie|TP> Tofe: removed the vboxvideo.conf from there but no luck still. There's also some fb something there
Sep 15 09:09:31 <Tofe> you can try to just remove vbox*.conf
Sep 15 09:09:51 <Tofe> the issue being that the virtualbox driver conflicts with uvesa
Sep 15 09:10:40 <Tofe> Herrie|TP: you saw sledges' comment ? looks like there is a missing '#' in the commit
Sep 15 09:11:15 <Herrie|TP> Tofe: Hmmz didn't see that.
Sep 15 09:17:12 <Herrie|TP> I don't see that comment?
Sep 15 09:19:00 <Tofe> https://bpaste.net/show/0ba5eb3e1bf7 in #sailfishos-porters
Sep 15 09:19:38 <Tofe> but I'll provide a fix commit tonight, it'll be easy
Sep 15 10:03:05 <Herrie|TP> Tofe: thnx
Sep 15 10:03:26 <Herrie|TP> Might be I forgot to copy paste something :'(
Sep 15 11:03:29 <Herrie|Laptop> Tofe: https://git.merproject.org/Herrie/qtscenegraph-adaptation/commit/5282b7dc3621f7120a8c23626ac68bb2570d5392
Sep 15 11:08:59 <Herrie|Laptop> I must have been sloppy somehow :S
Sep 15 12:30:24 <Herrie|Laptop> Tofe: I have Sep 15 14:18:39 qemux86 luna-next[484]: FATAL: 14:18:39.393: EGL library doesn't support Emulator extensions in the logs
Sep 15 12:34:49 <JaMa> I was seeing the same back then
Sep 15 12:35:35 <JaMa> 2017-09-08.log:19:32 <+JaMa> Tofe: luna-next fails to start in it, do you see anything obvious? https://pastebin.com/mhR76YDL
Sep 15 12:36:12 <JaMa> 19:33 <+JaMa> might be related to https://bugreports.qt.io/browse/QTBUG-61156
Sep 15 12:41:06 <Herrie|Laptop> Seems it might not be able to find displays in Emulator as per: https://github.com/qt/qtbase/blob/e4c39d5e1e7ee8c2bba273e6d613ec519b7fa9c2/src/plugins/platforms/eglfs/deviceintegration/eglfs_emu/qeglfsemulatorintegration.cpp#L86
Sep 15 13:01:51 <JaMa> is this with qemu or virtualbox?
Sep 15 13:02:04 <JaMa> it worked for me and Tofe in virtualbox later
**** BEGIN LOGGING AT Fri Sep 15 16:05:51 2017
Sep 15 17:08:19 <Herrie|Pre3> Tofe: I also fixed it and sent to sledges earlier
Sep 15 17:08:50 <Herrie|Pre3> So I guess our branches are equal now :P Seems I indeed forgot a # and a patch for a whole .cpp
Sep 15 17:09:36 <Tofe> Herrie|Pre3: are you to do MRs ?
Sep 15 17:09:49 <Herrie|Pre3> Tofe: Nope still 500 error
Sep 15 17:10:02 <Herrie|Pre3> But pasted link to sledges in #sailfishos-porters
Sep 15 17:10:14 <Tofe> ok, let's see who he prefers :p
Sep 15 17:10:23 <Herrie|Pre3> Mine wasn't tested (seems builder went offline at home) but should be straight forward enough
Sep 15 17:10:35 <Tofe> mine was ! hehe
Sep 15 17:11:06 <Tofe> I'm currently building mako with halium, and I get this android-header issue... annoying
Sep 15 17:14:35 <Tofe> there is a dependency of mediaindex on android-headers ?!
Sep 15 17:14:42 <Tofe> mediaindexer
Sep 15 17:15:22 <Herrie|Pre3> Tofe: Well my original commit missed a # (sloppy c&p) in context.cpp and I forgot a whole but in pluginmain.cpp
Sep 15 17:15:36 <Herrie|Pre3> So I guess your fix might not be required
Sep 15 17:16:21 <Tofe> Herrie|Pre3: in this kind of situation, I would recommend having both git remotes and do just a cherry-pick
Sep 15 17:17:48 <Herrie|Pre3> Tofe: Yeah I guess so
Sep 15 17:18:00 <Herrie|Pre3> Will learn for next time
Sep 15 17:18:29 <Herrie|Pre3> Tofe: There might be a dependency but would be weird I'd say
Sep 15 17:18:46 <Herrie|Pre3> mediaindexer just scans files with node-taglib
Sep 15 17:18:58 <Herrie|Pre3> Not sure why it would need android headers there really
Sep 15 17:19:02 <Tofe> Herrie|Pre3: I think bb processes all the dependencies recursively, and there seems to be qt5 in the middle
Sep 15 17:19:24 <Tofe> so it installs headers even if it isn't a direct depend, which is imho a mistake
Sep 15 17:22:00 <Tofe> the depends.dot output of bitbake is... surprising
Sep 15 17:23:03 <Tofe> here's an extract of all the entries beginning with "mediaindexer", i.e. its direct dependencies: https://bpaste.net/show/dd2d307b055a
Sep 15 17:23:57 <Herrie|Pre3> Weird indeed
Sep 15 17:24:08 <Herrie|Pre3> Shouldn't be so many imho
Sep 15 17:24:44 <Tofe> ah, I think I got why: there is "DEPENDS += " at the beginning, which is suspicious, and I think should be simply  "DEPENDS ="
Sep 15 17:25:36 <Tofe> ... nope, not enough
Sep 15 17:32:42 <Tofe> ok, it's via the qtbase depends, or via the luna-sys-common depends: it is recursive.
Sep 15 17:33:57 <Tofe> it just just doesn't make sense to install all there .pc files when we should just need the direct depends
Sep 15 17:34:13 <Tofe> s/there/these/
Sep 15 17:36:05 <Tofe> ah, or via the headers, there might be a chain of #include ...
Sep 15 17:38:11 <Tofe> I think OE is missing a concept of "exposed dependencies"
Sep 15 17:45:59 <Tofe> also, why only android-header has this issue...
Sep 15 17:47:49 <Tofe> are we missing a "COMPATIBLE_MACHINE" in android-header-* ? no idea...
Sep 15 18:02:27 <Tofe> Herrie|Pre3: for Halium on N4, I see a log of "init: cannot execve('/system/bin/something'): Permission denied"
Sep 15 18:02:31 <Tofe> lot*
Sep 15 18:02:51 <Tofe> we get UI though, and touch, and wifi
Sep 15 18:28:43 <Herrie|Pre3> Tofe: Not a bad start on UI. Other bits seems something small with permissions?
Sep 15 18:35:31 <Tofe> yup
**** ENDING LOGGING AT Sat Sep 16 03:00:01 2017