15:30:02 #startmeeting Fedora Mobility SIG 15:30:02 Meeting started Mon Oct 11 15:30:02 2021 UTC. 15:30:02 This meeting is logged and archived in a public location. 15:30:02 The chair is nirik. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions. 15:30:02 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:30:02 The meeting name has been set to 'fedora_mobility_sig' 15:30:02 #meeting_name fedora_mobility_sig 15:30:02 #chair torbuntu pbrobinson _Yoda_ 15:30:02 #topic init process 15:30:02 Current chairs: _Yoda_ nirik pbrobinson torbuntu 15:30:18 morning 15:30:21 👋 hello 15:31:05 o/ 15:31:16 * nirik will wait a few for folks to wander in. 15:32:52 alrighty... 15:32:57 #topic remix status 15:33:10 so, how are thinking looking right now? still sound issues? 15:33:41 Still sound issues yes. The latest seems to be that pipewire-pulse and callaudiod are having trouble with getting proper state on boot. 15:33:58 Also, the latest image on github is missing alsa-ucm altogether 🤦‍♂️ 15:34:10 oh 15:34:26 :( 15:34:28 At one point I could have sworn alsa-ucm was pulled in with alsa-ucm-utils, but maybe no longer? 15:35:15 I was trying to build a new image, but none have been successful with audio so I didn't post any to github. Someone was trying to set up a CI pipeline to generate them though through github 15:35:42 cool. 15:36:02 can it be callaudiod that needs changing ? 15:36:16 so fedora decided to stick with wayplumber for f35... so we should stick with it and file bugs on it. 15:36:29 The reason I don't point to callaudiod is because it wasn't the package that was changing. 15:36:41 👍️ cool, wireplumber it is then. 15:37:09 https://gitlab.com/mobian1/callaudiod/-/issues/16 <- maybe related callaudiod bug 15:37:26 they did also drop a new version friday that has a lot of fixes. 15:37:34 Not sure if you have tried that one yet 15:37:40 wireplumber? 15:38:10 I tried 0.4.3 as soon as it was available in koji but that didn't seem to change anything. 15:39:11 yes, sorry, now I am doing it. ;( 15:39:16 I see there was a patch relating to a segfault so I'll try out 0.4.3-2 this week 15:40:39 so whats the symptom? just doesn't switch from/to calls? or no audio at all? 15:40:54 For me it doesn't auto switch profiles 15:41:03 same here 15:41:09 if I kill callaudiod and it restarts, I may get one switch in one direction, but never again after 15:42:17 :( ok. I think the wireplumber maintainer is pretty responsive, so might be worth filing a bug and see if he can figure it out... 15:42:43 The thing is I'm not sure if it is wireplumber or pipewire-pulse 15:43:15 The more annoying thing is it does work on my main phone now 15:43:16 (using pipewire-media-session) 15:43:36 that does make it sound like a wireplumber thing then? 15:44:13 FYI, pipewire-media-session was sort of a demo/proof of concept I think, and wireplumber is the longer term 'real' implementation... 15:45:31 https://gitlab.com/mobian1/callaudiod/-/blob/master/src/cad-pulse.c#L842 so this is the error I'm seeing in callaudiod, which I think makes it point to something pulse related. Would that still now be handled in wireplumber? 15:46:19 well, it could be, or could be in pipewire-pulse as you say. ;( 15:46:27 I think there's another issue with pipewire-pulse 15:46:49 perhaps #pipewire could isolate where it is at least? (or whatever matrix room they hang out in ) 15:46:50 I uninstalled pinephone-helpers and rebooted and pipewire-pulse was still not running on startup 15:47:39 `systemctl --user status pipewire-pulse` after startup shows that the service was started but the process isn't running anymore 15:47:50 I have to do `systemctl --user restart pipewire-pulse` 15:48:09 yeah, thats normal... 15:48:23 it's got a socket unit... so when something requests pulse it starts one 15:48:32 systemctl --user status pipewire-pulse.socket 15:48:53 but pavucontrol hangs trying to connect to PulseAudio until I do `systemctl --user restart pipewire-pulse` 15:49:17 Thats weird. The socket unit should start one... is the socket showing disabled? 15:49:33 but anyhow, we could debug that more back in the main channel... 15:49:54 any other items on remix status? or just sound ? :) 15:50:04 https://gitlab.freedesktop.org/pipewire/pipewire/-/issues/1553 <- pipewire issue for historical 15:50:34 I can't think of anything new and exciting on that front. 15:50:39 my point is, I think there's another problem with pipewire-pulse besides the ALSA UCM thing specific to the PinePhone's audio hardware because I am encountering an issue just using USB audio 15:51:51 What about Plasma Mobile packaging? 15:51:58 could be sure... I wasn't seeing that a while back, but I haven't looked on my phone recently. 15:52:09 will have that in the next topic. :) 15:52:17 #topic Fedora Mainline status 15:52:33 So, sounds like we have a list of plasma mobile packages and have made some progress on them? 15:52:34 Still kernel stuff mostly. 15:53:17 and yes, kernel is still it's normal blocker. Hopefully pbrobinson can make some progress with getting the modem working with fedora kernel. 15:53:36 I updated the ticket with the packages and status 15:54:08 cool. 15:54:17 I know there's some folks that know cmake pretty well... 15:54:41 https://pagure.io/fedora-mobility/issue/10 15:54:59 I managed to pack kalk and phone-settings 15:55:14 nice 15:55:34 but the main package phone-components fail - and I just know too little about this stuff 15:56:00 folks in the fedora kde room may be able to help... 15:56:06 The others was basicly just edited spec files from Suse :-) 15:56:06 #kde:fedoraproject.org 15:56:52 Yoda[m]1: can you elaborate how it failed? 15:56:58 * nirik is looking forward to playing with it. 15:57:16 FWIW Plasma Mobile is working towards switching from oPhono to ModemManager. 15:57:18 it's not quite done yet 15:57:43 https://invent.kde.org/teams/plasma-mobile/issues/-/issues/62 15:57:50 nirik: I poke occasionally but it's just so little time of late, one day I will be able to just live in my house 15:58:01 yes - but that should just make it easier for us 15:58:01 glad they are tho... ophono died already once. ;) 15:58:17 pbrobinson: yeah, I hear ya... :) 15:58:37 any other fedora mainline news? 15:58:47 so we can make images with both phosh and plamo in it :) 15:59:55 yep. 16:00:10 you mean both in one image? That would be cool 16:00:24 yes 16:00:29 it would be nice to be able to switch and see which you like best... and they are both small. 16:00:44 anyhow... 16:00:47 #topic all other business 16:00:53 I had a few random things... 16:00:56 that was not possible when they used modemmanager/ofono 16:01:43 Fairphone 4 was released (in eu only) not long ago... and postmarketos folks already submitted some upstream kernel patches. I wonder if it could be a interesting target for us... 16:01:50 the non us is sad for me tho. ;( 16:02:07 that would be cool, but also, I'm in the US, so :/ 16:03:03 I'm not really sure how much of it's mainline, etc. 16:03:21 Well, I have no plans for buying more phones ... for it is interesting of course 16:03:26 https://wiki.postmarketos.org/wiki/Fairphone_4_(fairphone-fp4) 16:03:46 https://www.phoronix.com/scan.php?page=news_item&px=Fairphone-4-Linux-Patches 16:04:22 anyhow, also FWIW pine64 has some mysterious announcement later this week. 16:04:30 https://twitter.com/thepine64/status/1446506900586176515 16:05:08 I haven't heard any news about pinephone keyboard going on sale yet. 16:05:12 I'm going to lose my mind on the 15th 16:05:14 yeah, hard to guess... 16:05:36 * nirik suspects risc-v something, but who knows. 16:05:53 Yoda: can you post a link to the build log of the failed plasma-phone-components build? Maybe a few more people looking at it can figure out the problem 16:06:12 I'll try 16:06:28 ok, anyone have anything else? 16:06:41 I don't this time. 16:07:54 ok, thanks for coming everyone! 16:07:58 #endmeeting