**** BEGIN LOGGING AT Wed Feb 08 03:00:03 2017 Feb 08 04:21:48 I suppose for this I could update the temp code. Would just be nice to figure out this little issue. Feb 08 04:21:50 Thanks Feb 08 04:22:17 It seems like it copies code over. It must just be one of the definitions that is borked... Feb 08 06:08:20 Tofe: Nice what was the issue? Feb 08 06:08:24 Is this with 12.1 or 13.0? Feb 08 06:08:33 dkirker: What you're working on? Feb 08 06:44:26 morning btw ;) Feb 08 07:05:48 Herrie, right now? compiling a version of libmojocore that won't crash so mmsservice on the Pre3 doesn't blow up Feb 08 08:32:01 Ok... It took me long enough, but I think I figured it out! I specified a branch=master in the SRCURI, and then I made sure that S and B were NOT the same, because if they were, bb cleared out B.... Feb 08 08:32:25 Morning Feb 08 08:32:32 morning Tofe Feb 08 08:32:55 dkirker: I'm sorry, I looked at your issue, but I didn't find out what was the problem exactly Feb 08 08:33:00 * dkirker is bitbaking, or yoctoing, or whatever it is Feb 08 08:33:19 Tofe, no worries! I managed to find a clue that lead me to the answer! Feb 08 08:33:26 but thanks for looking! Feb 08 08:33:33 You could also directly go in the temp git directory, add a remote source, and work like that, without modifying S_* Feb 08 08:34:17 Yeah, I thought about that. That is what Herrie suggested. I dunno though, it felt kinda dirty. :P But using local-builds.inc seems to be useful. Feb 08 08:34:22 (just be careful to push your changes regularly, a do_unpack would erase the previous temp directory...) Feb 08 08:34:31 ok Feb 08 08:35:41 Herrie: CM12.1, but sensors were also working on CM13 Feb 08 08:36:00 Herrie: the issue is still service timing, mostly Feb 08 08:36:54 luna-next is started quite early, and triggers the qtsensors plugin, while sensorfwd isn't ready yet, or lxc container hasn't even started yet Feb 08 08:37:12 and it fails to recover when the backend gets in place Feb 08 08:37:46 It's similar to the ofono issue, actually Feb 08 08:39:19 Ah, and there was also an issue with the driver itself; the sailfishos guys managed to fetch a very similar driver from another hardware, which works better for hammerhead Feb 08 08:40:30 Tofe: Ah OK. So seems we might need to tweak systemd order a bit Feb 08 08:43:08 Herrie|Pre3: yes, at least start sensorfwd after android-system Feb 08 08:44:02 There's also a strange behavior of luna-next I'd like to investigate: even with the boot logo, it already asks for ofono status, which should only be done by the status bar Feb 08 08:44:28 Tofe: I guess some things changes a bit between CM 10.1/11.0 and 12.1/13.0 :P Feb 08 08:48:59 Seems there are some updates to sensorfw as well we might want to pick up Feb 08 08:53:34 yes, but they shouldn't change the situation I think Feb 08 08:54:17 anyway, it's nearly fixed. Now there is still bluetooth to fix, and I think my gridUnit numbers are not perfect Feb 08 08:59:08 Tofe: OK Feb 08 09:01:05 Yeah the changes shouldn't have much difference. **** BEGIN LOGGING AT Wed Feb 08 09:36:32 2017 Feb 08 09:37:13 FYI, this little hack would very probably also work for luneos: https://talk.maemo.org/showthread.php?t=98882 Feb 08 10:38:56 Tofe: What does this offer that's interesting? Newer Chromium? Feb 08 10:48:53 Herrie|Pre3: more like a debian chroot, and desktop apps Feb 08 10:49:16 Mostly useless on phones, might have some use case on tablets Feb 08 12:21:34 Tofe: OK Feb 08 21:32:30 Tofe: Would it be helpful if I make a systemd bootup start order chart from 10.1/11.0? Feb 08 21:32:40 To see where we go wrong on 12.1/13.0 ? **** ENDING LOGGING AT Thu Feb 09 03:00:02 2017