| rinigus | mal: thank you very much for catching it! As for compass, I don't think so. I used orientation before, as we have by default | 07:21 |
|---|---|---|
| spiiroin | @simonschmeisser: atm suspend and ongoing calls goes something like: if cpu is not needed for audio, it can be suspended -> depending on audiorouting ohm/pulse blocks allows suspend -> mce blocks suspend only in vicinity of call state chages (to allow ui side time to process changes, send dtmf tones, etc) | 09:46 |
| T42 | <adampigg> spiirion: iirc, the keepalive isnt installed by default, let me check ... would that cause the issue? | 09:48 |
| rinigus | something has to be done with https://github.com/mer-hybris/hybris-boot . we have host of PRs submitted in a queue and despite that maintainers seem to merge only for official devices (see last commit) | 10:04 |
| rinigus | maybe all these PRs (oldest from '2014 by piggz) should then be closed and we should forget about having universal hybris-boot with each port keeping its own version | 10:05 |
| piggz | rinigus: except for one PR there! | 10:06 |
| rinigus | which one? | 10:06 |
| rinigus | that from 2014? | 10:06 |
| piggz | rinigus: no, 189 | 10:06 |
| Chengcheng | i am developing a Linux Tablet, anyone who interested in this? details here https://www.reddit.com/r/JingOS | 10:07 |
| rinigus | piggz: yes, maybe. haven't looked into it, but sounds like this will resolve this pileup of PRs that Jolla is not interested in | 10:07 |
| piggz | rinigus: yeah, i think it s reasonable idea .. then all fixup mountpoints can be kept in DCD repos | 10:08 |
| T42 | <ankaos> mal: touchscreen work. thank you | 10:12 |
| rinigus | Chengcheng: hard to say. so far it looks like a branch of reddit and not clear what it is. is it a device or OS? | 10:14 |
| Chengcheng | it is a device, and we customized the OS | 10:16 |
| rinigus | Chengcheng: regarding device - Android hybris based? or Linux kernel mainline? | 10:20 |
| T42 | <ankaos> mal: camera work, wifi work, sound work :D | 10:26 |
| T42 | <simonschmeisser> spiirion: who decides on the "if"'s ? I could hear that the audio connection is still "active"(static low level noise) but no talking arrived anymore. | 10:28 |
| T42 | <simonschmeisser> Just to make sure I understood correctly, it should be possible to suspend the main system and only keep the modem alive during calls? | 10:29 |
| T42 | <simonschmeisser> it's a separate usb modem in the case of pinephone | 10:30 |
| T42 | <ankaos> mal: celcular modem not work | 11:30 |
| T42 | <ankaos> sim card work. sms work. but modem not work | 11:30 |
| mal | @ankaos maybe you need to build dummy_netd (https://github.com/mer-hybris/dummy_netd)? | 12:15 |
| T42 | <ankaos> mal: ı use android 7 | 12:17 |
| T42 | <ankaos> mal: dummy_netd is los-16 | 12:18 |
| mal | ok | 12:33 |
| T42 | <ankaos> mal: call sound not work | 12:34 |
| mal | do you have the pulseaudio hidl plugin? | 12:51 |
| mal | or wait | 12:51 |
| mal | depends on version which one you should have | 12:51 |
| T42 | <ankaos> mal: camera sound work | 12:51 |
| T42 | <ankaos> mal: okey ı looking | 12:52 |
| mal | https://github.com/mer-hybris/audiosystem-passthrough | 12:52 |
| T42 | <ankaos> mal: okey ı looking | 12:53 |
| T42 | <ankaos> mal: install files | 12:55 |
| T42 | <ankaos> mal: now reboot | 12:55 |
| mal | rinigus: the built-in compass in sensorfw seems to work except the output is a bit wobbly, the value is changing a bit all the time but on average it's correct if the value is filtered properly | 13:39 |
| rinigus | mal: nice! I hope that with the different sensorfw options it will now become a norm that compass is working on SFOS devices | 13:59 |
| mal | rinigus: yeah, I'm now checking the code to see if there is something I can do to make this better, I found some rounding issues at least, it just casts to float to int instead of using qRound for example | 14:08 |
| rinigus | with the meta packages, is there some trick how to replace "@Jolla Configuration device" with my own root meta package in generated KS? | 14:49 |
| rinigus | or any docs regarding meta packages? | 14:49 |
| mal | rinigus: by using custom .ks which means not being able use build_packages.sh --mic but instead running the mic command manually | 14:54 |
| rinigus | mal: running mic manually means "mic create loop ... " ? | 14:56 |
| rinigus | mal: and where do you generate such custom KS? | 14:56 |
| mal | rinigus: I mean build_packages.sh --mic should generate one and then just edit it and do not use build_packages.sh anymore | 15:00 |
| rinigus | mal: that is problematic. I have KS generated by OBS, as in https://build.merproject.org/package/show/nemo:devel:hw:sony:tama/droid-config-h8324 | 15:06 |
| rinigus | that KS is pulled into CI and used to create images https://gitlab.com/sailfishos-porters-ci/tama-ci | 15:06 |
| rinigus | KS was generated using (nonmerged) https://github.com/mer-hybris/droid-hal-configs/pull/175 and allowed to replace Jolla Configuration Device with jolla-configuration-device package | 15:06 |
| rinigus | as a result, all was handled automatically and I could work on 6 devices in parallel. | 15:07 |
| rinigus | mal: any sane way to reproduce it with the new system? | 15:07 |
| T42 | <ankaos> mal:call sound not work | 15:07 |
| T42 | <ankaos> I dont listen my friend sound | 15:07 |
| rinigus | mal: OK, will try to restore that functionality | 15:31 |
| rinigus | mal, sledges: let's see if it will go through this time https://github.com/mer-hybris/droid-hal-configs/pull/203 | 15:58 |
| T42 | <simonschmeisser> I know it's slightly offtopic, but is there a way to restore/rebuild the partition table on a jolla1? | 15:59 |
| mal | rinigus: I managed to fix the compass issue, it was a problem with the weighted average calculation | 16:02 |
| rinigus | mal: congratulations! | 16:03 |
| T42 | <ankaos> how install whatsapp app? | 16:13 |
| wdehoog | I'd like to build my version of mce on OBS but it gets a version that makes it impossible to install normally. How can I force it's version? | 17:52 |
| spiiroin | @adampigg It might be that pulseaudio-module-keepalive(?) does it and is optional | 17:55 |
| T42 | <adampigg> spiiroin : even with it installed, apprently it still suspends during call | 17:56 |
| T42 | <adampigg> @simonschmeisser could you plat some music in media player and see if it suspends? | 17:56 |
| spiiroin | @simonschmeisser: maybe more like: in android phone ports main cpu can be suspended during (some type of) calls -> there easily could be assumptions based on that here and there | 17:57 |
| spiiroin | I would not be surprised if modem over usb needs to be handled in different manner | 17:58 |
| spiiroin | @adampigg I'm bit unsure about exact details, but IIRC it depends on audio routing, ohm, whatnot ~ once established some routes work without cpu (cellular/dsp -> audio) | 18:00 |
| spiiroin | jusa: would know how wrong I got ^ ;-) | 18:00 |
| T42 | <simonschmeisser> spiiroin: well, for now it would be fine not to suspend anything during calls, can optimize later :) | 18:09 |
| T42 | <simonschmeisser> why is there no "Jolla song" included with the device? Something with pirates n stuff, to test audio? :D | 18:10 |
| T42 | <ankaos> sh-3.2# /usr/bin/anbox-shell.sh | 18:12 |
| T42 | <ankaos> ERROR: Cannot attach to container as it is not running | 18:12 |
| T42 | <ankaos> how fix? | 18:12 |
| T42 | <adampigg> @simonschmeisser my latest builds for pro1 and volla included some stock photos and videos .. so, there must be a packge that includes them somewhere | 18:15 |
| rinigus | while trying to install droid-config-device (up to date dcd submodule), I am getting an error: file /usr/lib/systemd/system/droid-hal-init.service conflicts between attempted installs of droid-config-h8324-0.3.8+master.20201128120434.1.g60c4416-1.21.1.jolla.armv7hl and droid-config-h8324-0.3.8+master.20201128120434.1.g60c4416-1.21.1.jolla.armv7hl | 18:18 |
| rinigus | was fine with 3.3.0; error appears with 3.4.0 update | 18:19 |
| mal | which command did you use? | 18:19 |
| T42 | <ankaos> https://paste.ubuntu.com/p/B33dp8XPcx/ -> /usr/libexec/droid-hybris/system/bin/logcat -b radio output | 18:23 |
| T42 | <ankaos> call sound not work. | 18:23 |
| rinigus | mal: that is a part of the build for droid-hal-version, paste at https://dpaste.org/61W2 | 18:26 |
| T42 | <simonschmeisser> @adampigg I wish it had suspended, this testfile I found is horrible (and volume controls don't switch mode yet) | 18:31 |
| T42 | <elros34> rinigus: strange error , maybe it's because you still use old /lib/system instead /usr/lib/systemd for droid-hal-init.service | 18:32 |
| T42 | <adampigg> @simonschmeisser so, the keepalive module works for medi playback? | 18:33 |
| mal | rinigus: I think it's because of the submodule change, the location moved, try installing it manually first | 18:34 |
| mal | with --force if needed | 18:34 |
| rinigus | mal: I cannot, it is installed during OBS build by OBS. I suspect that I have something mixed up in config, maybe something needs more to be updated. | 18:35 |
| T42 | <elros34> just try to move sparse/lib/systemd to sparse/usr/lib/systemd | 18:36 |
| mal | rinigus: can you show your config repo | 18:37 |
| mal | rinigus: found the problem | 18:39 |
| mal | rinigus: you have custom droid-hal-init in /lib when submodule has it in /usr/lib | 18:39 |
| mal | rinigus: so move the whole https://github.com/sailfishos-sony-tama/droid-config-sony-tama-pie/tree/master/sparse/lib to sparse/usr/lib | 18:40 |
| rinigus | mal: thank you very much! will do | 18:41 |
| mal | the reason why it fails with those is that there is a symlink in systemd which makes those path end up being the same so it tried to install file to same place twice | 18:45 |
| mal | rinigus: some odd issues with compass still, the raw data in debug output is correct but app shows wrong value | 18:46 |
| rinigus | mal: waiting for the build to complete | 18:47 |
| rinigus | mal, compass: really strange. I wounder whether qt applies some extra filtering as well | 18:47 |
| mal | yeah, need to debug | 18:48 |
| mal | rinigus: for example now I saw in debug log heading of 180 and app shows 230 | 18:49 |
| mal | also there is an issue that the compass filter remains active even after compass app is stopped | 18:50 |
| rinigus | mal: config issue with the port looks to be fixed, thanks! | 18:53 |
| rinigus | re compass - didn't have to debug in such detail with orientation. was just happy to see it finally pointing in non-random direction, but something that is reasonable. | 18:55 |
| rinigus | re filter: if there is access to time, then maybe drop the filter after longer delay in updates? that's what Osmo did in pure maps for calculation of movement direction and it seems to be reasonable | 18:56 |
| rinigus | ... i.e flush filter "old points" | 18:56 |
| T42 | <adampigg> rinigus: did you switch to new style pattern too? | 19:00 |
| T42 | <simonschmeisser> @adampigg and spiiroin: yes exactly, pulseaudio-module-keepalive inhibited suspend successfully but during call it still dropped | 19:00 |
| T42 | <adampigg> ok | 19:00 |
| T42 | <adampigg> good to know | 19:00 |
| rinigus | @adampigg: kinda. will preserve jolla-configuration-devID and use that in KS. now looking through other dependencies that I am expected to pull in | 19:01 |
| rinigus | (comparing it with older SFOS) | 19:01 |
| rinigus | is patterns-sailfish-device-configuration-devname expected to have something which is device-specific? I have only 2 requires in it: patterns-sailfish-applications & patterns-sailfish-ui | 19:06 |
| T42 | hanasveta was added by: hanasveta | 19:07 |
| T42 | <adampigg> rinigus: would be the same as old jolla-configuration-devname | 19:17 |
| T42 | <adampigg> https://github.com/sailfish-on-fxtecpro1/droid-config-t5/blob/master/patterns/patterns-sailfish-device-configuration-t5.inc | 19:17 |
| rinigus | @adampigg: ok, I see. then I can ignore it as I have other package for the same purpose. hence the redundant default config in my case | 19:19 |
| T42 | <adampigg> sure, i recall your package structure | 19:20 |
| rinigus | @adampigg: looks like you kept gstreamer1.0-plugins-good and friends. on xperia they seem to have dropped those (https://github.com/mer-hybris/droid-config-sony-ganges-pie/blob/master/patterns/patterns-sailfish-device-adaptation-kirin.inc#L42) | 19:24 |
| rinigus | or are these gstreamer packages pulled separately? | 19:24 |
| mal | rinigus: those are now included in common patterns | 19:52 |
| rinigus | mal: thanks! | 19:52 |
| rinigus | mal: which common pattern is that? sorry, tried to find but couldn't | 19:58 |
| mal | rinigus: the closed internal meta-packages actually | 19:58 |
| mal | patterns-sailfish-* | 19:59 |
| rinigus | mal: sounds like a major secret has been spilled. :) | 19:59 |
| mal | not really | 19:59 |
| rinigus | OK, will keep the requirement for the port, can always remove it later | 19:59 |
| mal | it's quite obvious those are now coming some other way than device patterns | 19:59 |
| mal | rinigus: usually it's quite safe to follow the way x10 or xa2 are done, of course some very latest changes might need unreleased sfos version | 20:01 |
| rinigus | mal: I tend to do so, comparing with x10. as it is the same base as tama, at least for now | 20:02 |
| mal | rinigus: in this case you can remove the gstreamer1.0-plugins* things from patterns if you want, just keep the gst-droid and gmp-droid | 20:03 |
| rinigus | mal: as I am using slightly different config, I am not 100% sure I included exactly the same packages. will make devel build and then can test after that | 20:04 |
| mal | rinigus: you seem to have some custom meta-package things in your repo | 20:07 |
| rinigus | mal: I do. it was from almost the beginning and I am reluctant to change it. main advantage - it supports 6 devices | 20:08 |
| rinigus | ... with autogeneration of KS | 20:08 |
| T42 | <ankaos> mal: ttps://github.com/mer-hybris/audiosystem-passthrough -> where ı looking? | 20:43 |
| mal | rinigus: I found why there was an offset | 20:49 |
| rinigus | mal: and it was... :) | 21:33 |
| mal | rinigus: sensorfw has declination configuration possibility and that comes from geoclue (parsed from nmea data) and it had some ridiculous value | 22:00 |
| mal | that should usually just couple of degrees to compensate for the location but it was 50 degrees on my x10, on my other devices it's between 0 and 6 | 22:01 |
| vknecht | wdehoog: generally you can set the version you want in OBS by setting branch and revision you want in the package's _service file | 22:02 |
| wdehoog | @vkencht thanks. I had a branch named with only numbers (1.106.3) and then it looked like tar_git decided on a quit different the version 1.101.xxxx . but it could also have been caused by my git incompetence | 22:09 |
| rinigus | mal: I was assuming that compass does not compensate for it. And have compensation in pure maps (as it knows location) | 22:10 |
| rinigus | Helsinki is not 50 degrees :) | 22:10 |
| mal | rinigus: that is the magnetic declination (also known as magnetic variation) value, not longitude or latitude | 22:21 |
| mal | https://en.wikipedia.org/wiki/Magnetic_declination | 22:22 |
| mal | rinigus: how do you calculate the correction in pure maps? | 22:24 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!