| Wizzup | https://github.com/maemo-leste-upstream-forks/ofono/commit/e0af50a500c75369c31e211bfe8c82d6c164a4e8 looks like we did | 00:00 |
|---|---|---|
| uvos__ | Wizzup: yes we did, you need up to date ofono and sphone | 00:04 |
| Wizzup | I think I do | 00:04 |
| Wizzup | I was doing tp testing | 00:04 |
| uvos__ | bencoh: "silly question, but why do we need voiceall-manager on top of telepathy-ring and ofono?" yeah i dont like this either, ofono, ring, v-m and the sphone module are just an insane number of mostly redundant abstraction layers | 00:05 |
| Wizzup | integrating the same code would still give the same bugs, so those need to be solved regardless | 00:06 |
| Wizzup | so for some reason, tp ring + voicecall-manager on vm: - outgoing calls works great, no anonymous bit. -incoming calls aren't even seen | 00:07 |
| Wizzup | on d4: -outgoing calls work but have privacy bit set somehow - incoming calls are seen and work great | 00:07 |
| uvos__ | sphone-ofono path works fine i presume | 00:08 |
| uvos__ | ? | 00:08 |
| Wizzup | will try | 00:09 |
| Wizzup | yes then it's not hidden | 00:10 |
| Wizzup | uvos__: bencoh: fwiw if someone does the work to take the code from voicecall manager and make it into a sphone module I'll be happy to drop what I've been doing | 00:11 |
| Wizzup | but until then I'm just going to stay the course and worry about this later | 00:11 |
| uvos__ | Wizzup: ok v-m doing somethong wierd then | 00:18 |
| Wizzup | uvos__: it's the same modem as pinephone | 00:19 |
| Wizzup | fwiw | 00:19 |
| uvos__ | v-m = voicecall manager | 00:19 |
| uvos__ | not virtual machine | 00:19 |
| Wizzup | oh | 00:19 |
| Wizzup | righ | 00:19 |
| Wizzup | yeah will check | 00:20 |
| uvos__ | well my message was confusing | 00:20 |
| Wizzup | but I'm done with this for today :D | 00:20 |
| Wizzup | still cool that xmpp/sip works as much as it does already | 00:20 |
| Wizzup | like I could type in xmpp or sip user and call them (calls failed in gst or some state change, but hey) | 00:21 |
| Wizzup | (type them in in sphone) | 00:21 |
| Wizzup | tmlind: uvos__: got a reset on 6.1.48: https://paste.villavu.com/raw/gYEEtB2LZEpLpfp0W3y2LjAcsWwTa3Jqagx0porb/ | 00:40 |
| uvos__ | Wizzup: hmm is there no way to import a ics file int qalender at the modment? | 01:21 |
| uvos__ | als what happend to syncevolution it and its package dissapeared | 01:21 |
| uvos__ | for that matter evolution itself seams to have dissapered in this version of devuan | 01:21 |
| Wizzup | uvos__: syncevo should exist | 01:40 |
| Wizzup | but I think there is a problem with the syncing, regression since beowulf | 01:40 |
| uvos__ | yeah it exists | 01:41 |
| uvos__ | but im not trying to sync anyhow# | 01:42 |
| uvos__ | just import ics | 01:42 |
| Wizzup | lmk how it does | 01:48 |
| Wizzup | goes* | 01:48 |
| Wizzup | I need to get some rest :) | 01:48 |
| tmlind | Wizzup: thanks for the oops, will take a look, maybe that's the mystery oops that we've been looking for | 05:59 |
| ac_laptop | hello people | 12:06 |
| Wizzup | hi | 12:09 |
| Wizzup | rafael2k: regarding your kernel question, I tend to agree @ rockchip kernel | 12:09 |
| Wizzup | uvos: btw once this voicecall is working I will try to migrate it to all be inside sphone, maybe even before you merge it, since I do agree it would be better to just have tp in sphone | 12:10 |
| Wizzup | udder: theoretically it might be just 1-2 days of work, but it'll probably be more, I would like some help from your side if I go down that path | 12:10 |
| Wizzup | oops, that was for uvos | 12:10 |
| Wizzup | sorry | 12:10 |
| Wizzup | uvos: but for now just having this testing code is helpful to surface any problems (there aren't that many it looks like), because the migration from voicecall-manager to sphone code should be very simple I think | 12:11 |
| Wizzup | they just have a plugin for tp, so we'd just need to map that plugin interface to their existing dbus interface | 12:11 |
| rafael2k | Wizzup, cool! | 12:24 |
| ac_laptop | n900 discharged normally, I've put a fully charged battery in it, it boots fine under fremantle, crashes/shuts down at boot under leste, what should I do ? | 12:25 |
| Wizzup | rafael2k: if we can have it in the same kernel branch as pp ultimately that would be easier, but I would understand if that is not the case | 12:27 |
| Wizzup | rafael2k: then we don't have to do the packaging dance again basically | 12:27 |
| Wizzup | ac_laptop: sorry, I can't really help, please wait for some others | 12:27 |
| sicelo | ac_laptop: i don't know what to say anymore ... we've given you some suggestions, but it is not clear if you've used them | 12:27 |
| sicelo | battery situation is very unlikely to make system crash in any way | 12:28 |
| sicelo | so if you see crashes, you most likely have a bad image, bad SD card, etc. | 12:28 |
| sicelo | the comparison you need to make between fremantle and leste is the voltage each system thinks your battery is at. | 12:29 |
| sicelo | in fremantle, you must use hal for that. don't use the bq27200 script | 12:30 |
| sicelo | so - when leste refuses to boot using the battery, boot fremantle and immediately check the voltage it reports, then let us know | 12:31 |
| sicelo | FYI, the bq27200 script in fremantle does and shows *exactly* the same as the leste kernel ... plus fremantle doesn't directly use the values from the fuel gauge (bq27200). so use hal or bme (not sure if it has a way to expose the voltages it sees) | 12:35 |
| ac_laptop | sicelo: https://paste.debian.net/1299356/ | 12:56 |
| ac_laptop | https://paste.debian.net/1299357/ | 12:58 |
| sicelo | right. so if you boot leste from this state, it crashes/powers off? | 13:04 |
| ac_laptop | sicelo: yes, it boots untils the desktop shows, doesn't even have the time to load the battery applet, Xorg shows a crash message (although I don't know if it's an actual crash of the normal Xorg error message of a shutdown), and shuts down | 13:05 |
| sicelo | then your problem is not battery related. | 13:06 |
| sicelo | the voltage is very high | 13:06 |
| sicelo | try a new sd card or something, redownload and verify the shasums on a new image | 13:07 |
| ac_laptop | ok | 13:08 |
| ac_laptop | but before I'll try to make it boot under leste again | 13:08 |
| sicelo | Wizzup: maybe share with ac_laptop how to boot and get dropped to console instead of h-d | 13:09 |
| sicelo | then we can do further checks on their battery situation | 13:10 |
| ac_laptop | and now it booted fine | 13:11 |
| sicelo | do a cat sys class power_supply bq27200-0 uevent | 13:11 |
| sicelo | anyway, you could also have a look in /var/log to see if any of the log files can tell you what the real problem is | 13:13 |
| sicelo | you could always mount the sd card on a pc to look at those logs | 13:13 |
| ac_laptop | https://paste.debian.net/1299359/ | 13:13 |
| sicelo | voltage/battery seems good here, and is reasonably within the values you saw in Fremantle | 13:14 |
| sicelo | mmm, but something's curious here ... why it shows 0% capacity, when the CI flag is clearly not set | 13:18 |
| ac_laptop | there should be the two boot attempts of yesterday and the successful boot of a few minutes ago https://www.rhizogen.es.eu.org/~ac/bugs/maemo/leste/syslog | 13:21 |
| ac_laptop | batt.txt recorded something this time https://www.rhizogen.es.eu.org/~ac/bugs/maemo/leste/batt.txt | 13:27 |
| sicelo | ac_laptop: go back into Fremantle, then run the bq27200 script without any arguments, and share the output | 13:28 |
| sicelo | do you swap batteries often? | 13:29 |
| sicelo | looks like the fuel gauge data is partly corrupted. frequent swapping of batteries can do this | 13:32 |
| buZz | ac_laptop: seen my battery monitor python tool? | 13:43 |
| buZz | http://space.nurdspace.nl/~buzz/maemo/batmon.py | 13:43 |
| buZz | its still sucking way too much CPU, but its a nice display of current going in&out in 'realtime' | 13:43 |
| sicelo | it will get the same corrupt data though 😋 | 13:44 |
| buZz | its not using the gauges though, just current&voltage | 13:45 |
| ac_laptop | sicelo: https://paste.debian.net/1299362/ | 13:46 |
| ac_laptop | sicelo: now I do, since I have several batteries, but the problem happened before with just one battery | 13:46 |
| ac_laptop | I would keep the device charging all night and the next day the battery would be 100%, the device would be off, and the booting problem would be the same | 13:47 |
| sicelo | try and stick to one battery while we're troubleshooting this, otherwise there's nothing usef | 13:48 |
| ac_laptop | sicelo: ok | 13:49 |
| sicelo | *nothing useful that we will be able to deduce | 13:49 |
| sicelo | in the meantime, clear out the batt.txt, so we have fresh start. how often do you log? i see you have a cronjob for it | 13:51 |
| ac_laptop | sicelo: the one in leste syslog ? it should be every minute | 13:59 |
| sicelo | ok. | 14:02 |
| tmlind | fyi, here are some initial wlcore gnss patches andreas posted: https://lore.kernel.org/linux-kernel/20231126191840.110564-1-andreas@kemnade.info/ | 14:55 |
| rafael2k | Wizzup, I'm not sure we'll be able to share the same kernel fork for PP and PPP at this time... | 15:00 |
| Wizzup | ok | 15:07 |
| Wizzup | tmlind: great | 15:41 |
| sicelo | rafael2k: or, ever :-p | 16:00 |
| sicelo | mobian even seriously considered ending official support for the OG PP | 16:01 |
| Wizzup | that's a little silly | 16:32 |
| Wizzup | kind of in style with gnome though | 16:33 |
| bencoh | does that mean buying a PP isn't worth it these days? | 16:46 |
| Wizzup | rafael2k: sicelo: is the ppp much better battery life wise? | 16:48 |
| sicelo | mobian and gnome actually have nothing in common. | 16:48 |
| Wizzup | sicelo: I thought phosh was gnome based | 16:49 |
| Wizzup | is it not? | 16:49 |
| bencoh | Wizzup: hm, last time I had a look I was under the impression that the ppp seemed worse battery wise | 16:51 |
| bencoh | but that was only based on stuff I read here and there | 16:51 |
| sicelo | the problem with PP is the state of the kernel ... the way it has been developed is too similar to how things are done in the android world, i.e. it's more downstream than mainline | 16:51 |
| sicelo | mobian is debian mobile, so to speak. it ships phosh, which does have gtk roots. it also ships plasma mobile (kde) | 16:53 |
| bencoh | https://xnux.eu/log/081.html | 16:53 |
| sicelo | rafael2k: you'll still maintain both pp & ppp? or you'll focus on ppp nowadays? | 16:58 |
| Wizzup | sicelo: ah, I didn't know it shipped plasma too | 17:02 |
| Wizzup | bencoh: ah that's not much better, weird | 17:02 |
| rafael2k | sicelo, as time permits, I need to update at least pp kernel | 18:06 |
| rafael2k | Wizzup, I did not made many battery tests... I can't tell for now | 18:07 |
| rafael2k | I'm with Mobian on it at the moment, soon to swap to Maemo and start the fun | 18:08 |
| Wizzup | cool | 18:11 |
| arno11 | Wizzup: i've just tried the last N900 img. swap is ok :) first boot is still difficult because nokia-modem causes troubles (2min to load the PIN stuff...and at least 2min to be able to enter the code), and omap_ssi and cmt_speech are still missing on boot (?!) | 19:38 |
| arno11 | few h-d errors in boot log | 19:40 |
| arno11 | concerning background image and priority sched | 19:40 |
| arno11 | (maybe that's why it is so slow to start) | 19:41 |
| arno11 | otherwise the img is ok after several minutes | 19:41 |
| arno11 | after rebooting, it's a bit 'better' but nokia-modem doesn't work | 19:51 |
| arno11 | (will work if we add cmt_speech and omap_ssi on boot probably) | 19:51 |
| Wizzup | hm, I think we had them blacklisted per your earlier comments, but would have to check | 19:51 |
| arno11 | hm, they are not blacklisted | 19:53 |
| arno11 | just missing on boot | 19:53 |
| arno11 | these 2 ones are essential | 19:54 |
| arno11 | anyway that's just missing modules, slowness is far more problematic | 19:55 |
| Wizzup | I don't understand how they would be missing on boot | 19:57 |
| arno11 | there were missing from the beginning iirc | 19:58 |
| arno11 | i didn't remember this issue has not been resolved | 19:59 |
| Wizzup | what does missing mean, sorry for being pedanctic | 20:00 |
| Wizzup | pedantic | 20:00 |
| Wizzup | do they not exist in /lib/modules ? | 20:00 |
| Wizzup | or are they just not loaded on boot | 20:00 |
| arno11 | they are just not loaded on boot | 20:02 |
| Wizzup | and without doing anything, they are loaded at the next boot? | 20:07 |
| arno11 | nope, still not loaded | 20:08 |
| arno11 | i think that's quite the same issue than nokia-modem module: probably need to add them in etc/module-load.d/ or /etc/modules | 20:11 |
| arno11 | personally i use /etc/modules to load them | 20:12 |
| Wizzup | ok | 20:15 |
| arno11 | oh, maybe they are only in devel lol, have to check | 20:15 |
| arno11 | (fresh install is 6.1.9 and devel still 6.1.48) | 20:16 |
| arno11 | not in devel, and options nokia-modem pm=1 is still in use. | 20:19 |
| Wizzup | what is 'not in devel' ? | 20:21 |
| arno11 | ok so many things are wrong with nokia-modem, i'll have a look later and let you know (and send you an email) | 20:23 |
| arno11 | ('not in devel' <-- omap_ssi and cmt_speech) | 20:29 |
| Wizzup | not loaded, or not present in /lib/modules | 20:29 |
| arno11 | sorry, not loaded | 20:29 |
| Wizzup | ok | 20:31 |
| Wizzup | we can do module-load.d in leste-config | 20:31 |
| arno11 | yep | 20:32 |
| arno11 | and we need to remove 'options nokia-modem pm=1' in modprobe.d | 20:34 |
| Wizzup | hm, why? | 20:35 |
| arno11 | because it's buggy now when loaded on boot (since 6.1.30 iirc) | 20:36 |
| Wizzup | hmmmmmmmmmmmm | 20:42 |
| Wizzup | do you know how it is buggy? | 20:42 |
| Wizzup | I mean we can disable it for now but it really shouldn't be buggy - as in we ought to fix that | 20:42 |
| Wizzup | arno11: do you know what makes it buggy? | 20:48 |
| arno11 | i can't give you a quick answer, currently checking my notes (can't remember what was the issue with it) | 20:51 |
| Wizzup | ok :) | 20:55 |
| sicelo | actually i meant to ask arno11 ... i don't experience the module issue you mention | 21:06 |
| sicelo | not in Leste, nor in pmOS. they load normally every single time | 21:06 |
| sicelo | but from what you said just now, maybe the difference is that i don't use SIM PIN, while it sounds like you do? | 21:07 |
| sicelo | so I'm beginning to wonder if the problem is really modules, or rather there's ofono problem or Leste stack related to unlocking the SIM perhaps? | 21:08 |
| arno11 | sicelo: you probably add them following the process from emails | 21:13 |
| sicelo | no | 21:13 |
| arno11 | ok | 21:13 |
| sicelo | it's completely ootb | 21:13 |
| sicelo | maybe try to disable the SIM PIN and see if behavior is different | 21:13 |
| arno11 | i think there is a misunderstanding, let me reboot the fresh install (need 10 min lol) | 21:18 |
| arno11 | sicelo: that's ok i found them with lsmod :) | 21:26 |
| arno11 | so the main issue is just, again, slowness... | 21:26 |
| sicelo | right :-p | 21:26 |
| arno11 | a fresh install is really difficult to use and more difficult to troubleshoot lol | 21:26 |
| Wizzup | the slowness goes away after a few minutes right? I don't think we'll able to combat that completely | 21:27 |
| arno11 | for the slowness i found hildon-desktop/home errors on boot logs | 21:29 |
| arno11 | maybe it can help...or not | 21:30 |
| arno11 | Wizzup: sicelo: personally, i removed nokia-modem on boot in my main config: this way, no troubles and faster to start Leste. and i just modprobe it when needed | 21:36 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!