17:00:24 <sledges> #startmeeting Mer-hybris Porters bug triage
17:00:24 <merbot> Meeting started Tue Nov  3 17:00:24 2015 UTC.  The chair is sledges. Information about MeetBot at http://wiki.merproject.org/wiki/Meetings.
17:00:24 <merbot> Useful Commands: #action #agreed #help #info #idea #link #topic.
17:00:35 <sledges> Welcome to the Mer-hybris Porters bug triage!
17:00:42 <sledges> #info We'll start with the bugs: https://bugs.nemomobile.org/buglist.cgi?bug_severity=critical&bug_severity=major&bug_severity=normal&bug_severity=trivial&bug_status=NEEDINFO&bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=TRIAGEDUPSTREAM&bug_status=REOPENED&component=Hybris-ing&email1=need-triage&emailassigned_to1=1&emailtype1=substring&order=bug_id&product=Hardware%20adaptation&query_format=advanced
17:01:06 <sledges> #topic https://bugs.nemomobile.org/show_bug.cgi?id=835 - 2015-10-25 06:14:14
17:01:06 <merbot> Nemo bug 835 in Hybris-ing "[hammerhead][alpha12][cameraplus] crash upon stopping video recording" [Critical,New]
17:02:14 <BluesLee_> two comments on that
17:02:52 <sledges> problem is with nexus5
17:03:08 <sledges> the way encoders work is different
17:03:10 <sledges> than in e.g. jolla tablet
17:03:22 <BluesLee_> the jolla camera app seems to work in general i.e. it can be delivered as the primary camera app
17:03:30 <sledges> and for now we don't have a way to fix it (sailors busy)
17:03:41 <sledges> so all what's left for now is to bisect droidmedia and gst-droid
17:03:43 <mal-> I had similar issue on iyokan when I had incorrect jolla-camera-hw.txt
17:03:45 <sledges> and revert offending commit
17:03:58 <sledges> i am doing that ^^
17:04:22 <sledges> got to the point that it's not gcc4.8 problem
17:04:24 <sledges> which is good
17:04:41 <sledges> so reverting to state where things worked in 1.1.6 is theoretically possible also in 1.179
17:04:44 <sledges> 1.1.9
17:04:45 <sledges> and newer
17:05:03 <sledges> on OBS i could recompile only gst-droid, but not droidmedia due to packaging
17:05:07 <sledges> so now continuing this locally
17:05:12 <sledges> hopefully we'll have good news tonight
17:06:04 <sledges> but that's not long term solution
17:07:18 <BluesLee_> the priority for the bug is okay, should be at the top of our list
17:07:51 <sledges> yep, fact that it's regression is making things sour
17:07:54 <BluesLee_> sledges: maybe you can update the bug after your tests
17:08:30 <BluesLee_> if you need a tester drop a note with instructions
17:10:03 <sledges> just updated with more info: https://bugs.nemomobile.org/show_bug.cgi?id=835#c17
17:10:03 <merbot> Nemo bug 835 in Hybris-ing "[hammerhead][alpha12][cameraplus] crash upon stopping video recording" [Critical,New]
17:10:23 <sledges> let's move on
17:10:48 <sledges> #topic https://bugs.nemomobile.org/show_bug.cgi?id=837 - 2015-10-25 06:15:09
17:10:48 <merbot> Nemo bug 837 in Hybris-ing "[hammerhead][alpha12] Bad battery life caused by sensors" [Critical,New]
17:11:39 <sledges> i remember reverting up to the alpha where sensors got introduced
17:11:48 <sledges> iirc it was eating cpu there too
17:12:06 <sledges> so we never had a fix, just some race conditions where it didn't consume
17:12:36 <BluesLee_> that's correct, i checked that
17:12:44 <sledges> hmmmm
17:14:16 <BluesLee_> don't blame me but i used cm for several days and battery life there was much, much better than sfos. i don't know what we could or should expects
17:15:01 <BluesLee_> but i think that sensors don't eat up much cpu on Jolla phone
17:15:09 <mal-> on other devices the battery life is comparable to cm
17:15:44 <BluesLee_> mal- that means that it can survive two days?
17:16:32 <mal-> BluesLee_: my iyokan can stay on for a week on idle, with cellular but no wifi or mobile data
17:16:41 <sledges> mal-: so sensors.qcom doesn't eat CPU on cm, simple as that?
17:17:12 <BluesLee_> mal- do you observe the issue on other devices?i guess no
17:17:12 <mal-> sledges: yes
17:17:22 <sledges> s/mal-/BluesLee/ :)
17:17:49 <mal-> BluesLee_: seems to be nexus5 specific
17:17:52 <sledges> BluesLee_: iirc, disabling sensors used to increase batt life on n5
17:18:55 <BluesLee_> sledges: i can check this again (disabling sensors) and comment on the bug but we need help, lpotter?
17:20:40 <BluesLee_> if we fix those two bugs i would call the release beta
17:23:39 <sledges> BluesLee_: we need to get a hold of cases where sensors don't consume CPU - looking from bug that happened
17:24:03 <sledges> just need to be able to identify circumstances, for possible reproduction or at least proof that it's race condition - and what's involved
17:24:22 <mal-> sledges: BluesLee_ have you tried to disable individual sensors from configs to see if't some specific sensor?
17:24:36 <sledges> good thing too
17:24:40 <sledges> we haven't:)
17:25:11 <BluesLee_> no, i have not
17:25:43 <BluesLee_> in which config file we can do that?
17:26:05 <mal-> I think /etc/sensorfw/sensors.conf or something
17:26:15 <mal-> actually primaryuse.conf
17:26:30 <sledges> in this pair:?
17:26:30 <sledges> https://github.com/mer-hybris/droid-config-hammerhead/blob/master/sparse/etc/xdg/QtProject/Sensors.conf
17:26:37 <sledges> https://github.com/mer-hybris/droid-config-hammerhead/blob/master/sparse/etc/sensorfw/primaryuse.conf
17:26:53 <BluesLee_> in general i rotate the phone twice while top is activated in a terminal and it takes up to 10 s to see the outcome
17:27:54 <sledges> that sounds new
17:28:07 <mal-> is the cpu usage constant, so it starts immediately?
17:28:15 <sledges> also we had suspicion that it consumes high CPU only after it had gone to sleep after boot at least once
17:28:33 <sledges> cpu usage is ~12% and it doesn't affect other performance
17:28:46 <mal-> so maybe disable all but one sensor and add new ones if cpu stays ok
17:30:20 <sledges> sounds like a plan
17:30:57 <sledges> let's move on?
17:31:00 <mal-> assuming the sensor left is not the problem :)
17:33:17 <mal-> next bug
17:34:32 <sledges> ok, updated bug with all your suggestions on how to proceed, thanks!
17:34:40 <sledges> #topic https://bugs.nemomobile.org/show_bug.cgi?id=838
17:34:40 <merbot> Nemo bug 838 in Hybris-ing "hybris2-11.0 - the modular-friendly manifest" [Normal,New]
17:35:09 <sledges> closing this one, all done
17:35:46 <sledges> #topic https://bugs.nemomobile.org/show_bug.cgi?id=841
17:35:46 <merbot> Nemo bug 841 in Hybris-ing "wrong owner of /var/lib/ofono (due to wrong package order during mic image create)" [Normal,New]
17:36:25 <sledges> current status: currently hacked inside .ks
17:36:36 <sledges> :(
17:36:51 <sledges> high:major?
17:37:19 <sledges> sadly only critical bugs get painted red
17:37:29 <sledges> but this is quite important for other porters to know
17:39:04 <sledges> we could paint it critical?
17:39:25 <sledges> and it should be renamed to "APN settings not remembered"
17:39:38 <sledges> current name is the cause, not symptom
17:39:50 <mal-> renaming would be a good idea
17:42:48 <sledges> done
17:42:59 <sledges> #topic https://bugs.nemomobile.org/show_bug.cgi?id=845
17:42:59 <merbot> Nemo bug 845 in Hybris-ing "[aosp5] patch "allow nemo to register services"" [Normal,New]
17:43:40 <sledges> Didn't hear any troubles from that porting camp, closing, re-open if becomes an issue
17:45:10 <sledges> #topic https://bugs.nemomobile.org/show_bug.cgi?id=846
17:45:10 <merbot> Nemo bug 846 in Hybris-ing "Devicelock does not work correctly" [Normal,New]
17:46:04 <sledges> is that still relevant? at leat OTA device lock is now working, right BluesLee_ ?
17:46:47 <BluesLee_> good question
17:47:37 <BluesLee_> its iyokan device
17:47:55 <sledges> oyes :")
17:48:14 <sledges> setting needinffo
17:48:26 <mal-> I didn't know that bug was for iyokan, I need to do some testing
17:49:05 <mal-> adding the codename to bug title would be nice
17:49:06 <sledges> mal-: can i assign it to you then, or to oh1jt?
17:49:10 <sledges> already added;)
17:49:21 <mal-> I can take that
17:49:50 <sledges> thanks
17:49:54 <sledges> setting needinfo to you
17:50:22 <mal-> somehow missed the bug
17:50:41 <sledges> usually there always are more things we get busy with than devicelocks:)
17:50:54 <mal-> I have never even tried devicelock
17:50:55 <sledges> #topic https://bugs.nemomobile.org/show_bug.cgi?id=848
17:50:55 <merbot> Nemo bug 848 in Hybris-ing "[hammerhead] unable to change volume when in call" [Normal,New]
17:52:47 <sledges> i suggest we rename this to "low volume heard at other end of the call, nexus5 mic needs amp boost"
17:52:57 <sledges> as we already have bug 815
17:52:57 <merbot> Mer bug 815 in QtCreator Mer Plugin ""Unknown publisher" message installing 20131021 on Windows" [Normal,Reopened] https://bugs.merproject.org/show_bug.cgi?id=815
17:53:13 <sledges> merh:)
17:54:05 <BluesLee_> i was not aware of this bug, have to check it
17:55:49 <mal-> related to previous devicelock bug, I couldn't reproduce that on 1.1.9.30
17:56:14 <sledges> mal-: feel free to close it
17:56:24 <mal-> I'll make further testing to make sure
17:56:30 <sledges> ok thanks
17:56:38 <sledges> #topic https://bugs.nemomobile.org/show_bug.cgi?id=849
17:56:38 <merbot> Nemo bug 849 in Hybris-ing "[hammerhead][alpha 12] empty call log" [Normal,New]
17:57:08 <BluesLee_> not appeared again, please close it
17:57:46 <sledges> thanks marked
17:57:49 <sledges> #topic https://bugs.nemomobile.org/show_bug.cgi?id=850
17:57:49 <merbot> Nemo bug 850 in Hybris-ing "[hammerhead][1.1.7.25] HighVolumeNotification reappears on headphones" [Trivial,New]
17:58:46 <sledges> hmm, a bug for TJC? i somehow remember similar wierdnesses on jolla phone too
17:58:52 <sledges> can't elaborate more atm
17:59:11 <BluesLee_> no bug at all i would say
17:59:48 <BluesLee_> not sure but i think it's the same on the Jolla
18:00:35 <BluesLee_> you get the notification as Jolla does not want to be blamed and this should survive a reboot;-)
18:02:06 <BluesLee_> or is the bug about unsaved headphone volume after rebooting?
18:02:28 <sledges> it says it's saved
18:02:37 <sledges> i think you're right first time, commented and closed:)
18:02:45 <sledges> ok, we've spent a nice hour, half-way through bugs
18:03:00 <sledges> if you want to triage any remaining bug important to you, please do now https://bugs.nemomobile.org/buglist.cgi?bug_severity=critical&bug_severity=major&bug_severity=normal&bug_severity=trivial&bug_status=NEEDINFO&bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=TRIAGEDUPSTREAM&bug_status=REOPENED&component=Hybris-ing&email1=need-triage&emailassigned_to1=1&emailtype1=substring&list_id=451&order=bug_id&product=Hardware%20ada
18:03:07 <sledges> otherwise we leave remainder for next week
18:03:35 <mal-> there is a fix for this https://bugs.nemomobile.org/show_bug.cgi?id=861
18:03:35 <merbot> Nemo bug 861 in Hybris-ing "[hammerhead][alpha12] Microphone not working for apps like recorder, saera etc" [Normal,New]
18:04:14 <sledges> mal-: yep, i noticed your comment. we should close it after adding that package to patterns?
18:05:00 <mal-> yes, it should work on al devices, I think someone tried on n5?
18:05:55 <sledges> we'll see if we receive comments
18:05:58 <sledges> let's regroup next week
18:06:05 <BluesLee_> cool, thank you
18:06:11 <sledges> thank you all!
18:06:14 <sledges> #endmeeting