| sicelo | the number of incoming voice calls that i miss on the N900 is growing :'( | 16:53 |
|---|---|---|
| sicelo | the incoming call is signaled, then when i try accepting the call, it just keeps vibrating, sphone dies, with vibration going on non-stop | 16:53 |
| sicelo | maybe it's that audio thing you mentioned before arno11? | 16:54 |
| arno11 | hmm, that sounds like an old issue we got long time ago | 16:58 |
| arno11 | something seems really wrong with sphone | 16:58 |
| arno11 | i don't have such issue with sphone 0.9.5 btw | 16:59 |
| arno11 | (at least in silent mode) | 17:00 |
| arno11 | let me double check with vibrator | 17:01 |
| arno11 | no issue (silent mode + vibrator) | 17:03 |
| sicelo | i actually don't even like using the vibrator, but i guess it's on by default | 17:03 |
| sicelo | let me try with silent | 17:03 |
| sicelo | but then what's the point using silent if you want to receive calls :p | 17:03 |
| arno11 | i mean, vibrator +silent avoids the ringtone issue | 17:04 |
| arno11 | at least on my device, but i remember having troubles like yours from the fresh install with wrong meta pkgs / versioning | 17:05 |
| arno11 | have to go, bbl | 17:08 |
| sicelo | arno11: i don't see an `[Out] Earpiece` use case ... did we ever have that? at least sphone seems to want it | 17:13 |
| sicelo | i see "sphone: route-pulseaudio: failure: Set sink to Earpiece No such entity" | 17:14 |
| arno11 | sicelo: @earpeace, i've always seen that failure | 18:08 |
| arno11 | AFAIK, should not affect calls | 18:08 |
| arno11 | *on n900 | 18:09 |
| sicelo | yeah, i've made more tests | 18:12 |
| freemangordon | sicelo: is it possible to calculate mAh from Wh? | 18:12 |
| freemangordon | I guess Wh/U | 18:12 |
| freemangordon | ? | 18:12 |
| crab | sicelo: are you using sip on your n900? | 18:12 |
| sicelo | freemangordon: yeah i think that formula would work. otoh, even upower's mAh is quite inaccurate because it's calculated as well - i've noticed even charge-full changes during runtime. so i guess mAh calculated from it would end up crazy as well | 18:14 |
| freemangordon | hmm, but drivers provide charge_now, no? | 18:14 |
| sicelo | crab: i've just made basic tests. not succeeded to get the media actually working yet | 18:15 |
| crab | i guess you are on daedalus? | 18:15 |
| sicelo | freemangordon: sorry, i meant to say energy-full (which is how upower reports capacity) | 18:16 |
| sicelo | so yes, some drivers might have energy_now and energy_full, but some might not have | 18:16 |
| sicelo | e.g. N900's doesn't have energy_full. my d4 is out of service, so i can't check if it does | 18:16 |
| freemangordon | no, I mean charge_now, in mAh | 18:17 |
| freemangordon | I am sure d4 does and iirc so n900 do | 18:17 |
| sicelo | yes, those exist | 18:17 |
| freemangordon | anyway, I think Wh/U should be correct | 18:18 |
| sicelo | crab: yes. i guess situation would be the same in chimaera though. or what do you mean? | 18:18 |
| sicelo | Wh/U ... U means V? | 18:18 |
| freemangordon | well, U means U in Volts :) | 18:19 |
| freemangordon | unless I have forgotten how voltage is denited | 18:19 |
| freemangordon | *denoted | 18:19 |
| sicelo | :-) | 18:20 |
| sicelo | well yes, looks like the result is correct | 18:20 |
| arno11 | crab: sicelo: SIP works fine on chimaera-devel with rakia/sphone or twinkle AFAIK | 18:21 |
| sicelo | so we can just calculate it inside status-area-applet-battery | 18:21 |
| sicelo | arno11: crab: might be i need to fine-tune my configuration, so things suit the SIP server i'm on | 18:22 |
| sicelo | actually, let me reconfigure to another one (same account) | 18:22 |
| freemangordon | sicelo: exactly | 18:22 |
| freemangordon | also, in SI it is U that is used as symbol for voltage, IIUC | 18:23 |
| sicelo | i'll work on it this evening | 18:24 |
| freemangordon | cool, thanks | 18:24 |
| arno11 | sicelo: btw @calls/sphone crash, is it working again after reboot or is the issue permanent now? | 18:27 |
| sicelo | haven't rebooted. but started sphone from terminal to see what's happening. seems to behave for now | 18:28 |
| Wizzup | I will merge the sphone mr for the contacts fix | 18:37 |
| Wizzup | also, I have seen on my d4 that mpv cannot connect to PA | 18:37 |
| Wizzup | I wonder if either PA crashes or there is some issue accessing it | 18:38 |
| arno11 | ah, interesting | 18:39 |
| arno11 | btw i checked audacious, smplayer, twinkle, emulators and no troubles with sound/PA | 18:40 |
| arno11 | but ringtone and rakia/sphone crashes PA | 18:42 |
| sicelo | are you sure it's PA that crashes? | 18:43 |
| arno11 | not 100% sure ofc | 18:45 |
| sicelo | ok | 18:46 |
| sicelo | i do see "2025-04-21T17:36:03.429651+02:00 devuan-n900 pulseaudio[7288]: [alsa-source-AIC34 tlv320aic3x-hifi-0] alsa-source.c: Resume failed, couldn't restore original sample settings" ... so you could be right :-) | 18:47 |
| sicelo | btw Wizzup (or anyone else) - does it also happen to you that if are connected to a wireless network, then go out of range, the icon will blink indefinitely until you manually choose Disconnect? | 18:49 |
| Wizzup | sicelo: I will have to test that, it didn't use it, but it might with some more recent changes by fmg | 18:57 |
| sicelo | freemangordon: Wizzup: here goes the mAh MR: https://git.maemo.org/leste/status-area-applet-battery/pulls/10 | 18:59 |
| Wizzup | building new sphone | 19:15 |
| Wizzup | sicelo: the mr looks ok to me, but I'll leave it to others to check as well | 19:32 |
| sicelo | sure | 20:05 |
| freemangordon | sicelo: sorry, just got home after ~900km drive, will not look now :) | 20:08 |
| sicelo | yeah, no worries | 20:09 |
| sicelo | Wizzup: i was thinking we should really consider setting up the CI so an MR automatically gets built, allowing testers to immediately play with the artifacts even while the MR is undergoing review. their testing might reveal issues (or benefits) that could affect the review | 20:11 |
| sicelo | arno11: rakia+sphone is really unhappy, even with a different SIP server (both belong to the same ISP [which I work for], but they're configured differently) | 20:14 |
| sicelo | let me try twinkle | 20:14 |
| arno11 | yeah no surprised @rakia on daedalus, but twinkle should be fine for sure | 20:15 |
| arno11 | but config could be tricky depending of isp | 20:16 |
| sicelo | i have a working config for twinkle on desktop, so i can just replicate that on N900, i hope | 20:16 |
| arno11 | ah cool | 20:17 |
| uvos | Wizzup: did you build sphone or do i need to tag a release? | 20:17 |
| sicelo | actually i maintain those SIP servers, and twinkle (on desktop) is how i do all testing ;-) | 20:17 |
| sicelo | uvos: since the change is limited to debian/*, iirc there's no need to tag anew. at least that's what i seem to have read on wiki yesterday (while looking for some other information) | 20:28 |
| sicelo | btw, gentle ping regarding the mce MR | 20:28 |
| Wizzup | uvos: I built it, since it was just a deps change, there was no need for a new release | 20:29 |
| Wizzup | sicelo: yes @ CI, I think we will | 20:29 |
| Guest7102 | just rebooted with sphone update, no contacts issue | 21:03 |
| Wizzup | great | 21:03 |
| sicelo | :-) | 21:06 |
| arno11 | btw we really need to hide sphone sms icons by default to avoid confusion | 21:17 |
| arno11 | *with conversations | 21:18 |
| sicelo | arno11: btw, did you find that the N900's wlan IP address is more stable now than before? | 21:18 |
| sicelo | yes @ those icons | 21:18 |
| arno11 | @IP i didn't really check/notice | 21:19 |
| sicelo | ok. might not matter if you don't use the device in certain ways ;-) | 21:19 |
| arno11 | :) | 21:20 |
| arno11 | sicelo: btw, the issue you got with missed call was with general or silent profile ? | 21:28 |
| arno11 | *calls | 21:28 |
| sicelo | general | 21:28 |
| arno11 | ah ok i see | 21:29 |
| sicelo | but does sphone even respect those? or it uses its own `sphone -c options` ? | 21:29 |
| arno11 | so it is the ringtone issue | 21:29 |
| arno11 | (yep, i think it respects those) | 21:29 |
| arno11 | i have quite same troubles if i use general profile | 21:30 |
| arno11 | that's why i use silent + vibra | 21:31 |
| arno11 | no troubles with calls so far with this profile | 21:42 |
| arno11 | i mean, for daily usage, no other workaround atm | 21:43 |
| arno11 | Wizzup: sicelo: oh, we don't have tonegend in daedalus | 21:59 |
| Wizzup | ah... | 22:01 |
| Wizzup | we do need this I suspect? | 22:01 |
| Wizzup | if we do, it shojuld be in some dep / meta pkg | 22:01 |
| arno11 | iirc we need it for tones and dtmf | 22:06 |
| sicelo | yes, afaik it's not related to this issue | 22:09 |
| sicelo | arno11: i'll use the silent + vibra for a while ... but i do need a ringtone, because i'll miss many calls | 22:23 |
| arno11 | oh..sorry | 22:24 |
| arno11 | so the *main* question is: it is n900 specific ? because we have cmtspeech, cmt_pulse, ofono, vcm, telepathy, and sphone involved... | 22:30 |
| arno11 | and PA ofc | 22:30 |
| sicelo | could be. cmtspeech is a PA client just like sphone, and we know sphone's PA code has a real bug. cmtspeech most likely does too -> disaster :p | 22:32 |
| arno11 | maybe :P | 22:34 |
| arno11 | but why we have issues with only ringtone and telepathy rakia atm with quite same PA errors. sounds like cmtspeech is not involved | 22:41 |
| arno11 | and cmtspeech/cmt_pulse have no control on ringtone afaik | 22:41 |
| arno11 | and with no ringtone, no call issue | 22:43 |
| Wizzup | https://nlnet.nl/project/MaemoLeste-AppArmour/ :) | 23:02 |
| sicelo | awesome!! | 23:14 |
| sicelo | arno11: btw there was some shm stuff at some point. did it make it to daedalus? | 23:19 |
| arno11 | sicelo: @hsm, yeah we have it. | 23:48 |
| arno11 | i tried to decrease/increase it, no impact, still the same issue | 23:48 |
| sicelo | arno11: uvos: https://git.maemo.org/leste/bugtracker/issues/773 ... tried to document the sphone bug | 23:49 |
| arno11 | ok cool | 23:54 |
| uvos | hmm | 23:54 |
| uvos | never herd of this bug | 23:54 |
| uvos | and its quite strange | 23:54 |
| uvos | as sphone only calls that function on exit and sphone only has one single thread, the main thread | 23:55 |
| uvos | so it cant really call it from the wrong thread | 23:55 |
| sicelo | uvos: https://freedesktop.org/software/pulseaudio/doxygen/threaded_mainloop.html - there's a new thread created, "The added feature in the threaded main loop is that it spawns a new thread that runs the real main loop" | 23:57 |
| uvos | yeah i know | 23:57 |
| sicelo | in sphone's case, the stop is called from that thread, because of the callback | 23:57 |
| uvos | no | 23:57 |
| sicelo | ok :-) | 23:58 |
| uvos | its called in sphone_pa_distroy_interface which is called in sphone_module_exit | 23:58 |
| uvos | which is called in sphones main thread | 23:58 |
| uvos | i think | 23:58 |
| uvos | at leas | 23:58 |
| uvos | t | 23:58 |
| sicelo | no, that's not the only time it's called. it's called two more times via callback | 23:58 |
| uvos | ah yeah its allso called in sphone_pa_state_callback | 23:58 |
| uvos | your right | 23:59 |
| uvos | ok | 23:59 |
| sicelo | sphone_pa_state_callback | 23:59 |
| sicelo | yeah, and that's where the issue comes from | 23:59 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!