| Wizzup | sicelo: re: iio-sensor-proxy not starting on boot, maybe this is why the pinephone doesn't rotate in daedalus image? | 12:11 |
|---|---|---|
| sicelo | yes | 12:11 |
| sicelo | even d4 .. everything | 12:11 |
| Wizzup | ok | 12:11 |
| Wizzup | seems like a daedalus blocker | 12:12 |
| sicelo | yes, that's why i added it to the milestone | 12:12 |
| Wizzup | ah, just saw that | 12:12 |
| Wizzup | great | 12:12 |
| Wizzup | so do we add it to maemo system services, or what? | 12:29 |
| sicelo | what does maemo-system services do? | 12:35 |
| sicelo | oh i see | 12:36 |
| sicelo | we could add the init script there, yes. | 12:36 |
| sicelo | of course i would like us to ship latest upstream iio-sp (3.7) so we can also get N900 proximity | 12:37 |
| sicelo | and face-down/up stuff | 12:38 |
| sicelo | i already did the porting, just surprised i didn't push somewhere. will find where it's stashed then open PR | 12:39 |
| sicelo | https://github.com/sicelo/iio-sensor-proxy/commits/leste/daedalus/ ... but i think i want to redo this so it's a bit cleaner | 12:40 |
| Wizzup | yeah, so do we add the init script to maemo system services? | 13:03 |
| Wizzup | then we can figure out the upstream-fork | 13:03 |
| sicelo | sounds like a plan :-) | 13:07 |
| Wizzup | ok | 13:11 |
| Wizzup | I think I will leave the pine64 kernel upgrade to 6.12 for now, the sunxi64_defconfig disappeared in favour of some 'all config' style | 13:11 |
| Wizzup | which means the kernel build would take too long and be very big | 13:11 |
| Wizzup | sicelo: wait if we're forking it any way we could just put it in there | 13:12 |
| Wizzup | I'll just add it for now | 13:16 |
| Wizzup | sicelo: ok that's sorted | 13:26 |
| Wizzup | interestingly enough the daedalus pinephone image (upgraded to daedalus-devel) does not detect charging | 13:36 |
| Wizzup | oh, it does detect some other cable | 13:41 |
| inky | > I think I will leave the pine64 kernel upgrade to 6.12 for now, the sunxi64_defconfig disappeared in favour of some 'all config' style | 13:45 |
| inky | i have a working config on pinephone. | 13:45 |
| inky | i built it under gentoo and it works. | 13:45 |
| inky | then i rebuilt with 6.13, so now i run 6.13 on pinephone gentoo. | 13:45 |
| inky | minute. | 13:45 |
| Wizzup | I did the rebase on 6.12 | 13:45 |
| inky | and everything works well. | 13:45 |
| Wizzup | but I need to patch the builddeb script | 13:45 |
| Wizzup | and then also rebase out config patches | 13:46 |
| inky | ah so you don't need the config? i can give you. | 13:46 |
| Wizzup | no, I don't need the config | 13:46 |
| Wizzup | their config has a bunch of thngs that do not work for us | 13:46 |
| inky | good, thank you for doing that. | 13:46 |
| Wizzup | like making EXT4 a module | 13:46 |
| inky | okay. with my config i have tested that xorg works, and bluetooth works. i have enabled xfs as [*] because i run xfs on all my gentoos, even on pinephone. anyway, know i have it, if needed. | 13:47 |
| inky | in 6.13 bt works better than in 6.12. much better. | 13:48 |
| inky | i used megi's orange_pi branch because apparently that's what megi suggested to use, it is the main branch where megi puts the changes for pinephone. | 13:49 |
| Wizzup | in any case | 13:50 |
| Wizzup | there is a pinephone daedalus image now | 13:50 |
| Wizzup | we'll get a kernel upgrade shipped some time soon | 13:51 |
| inky | wow! | 13:51 |
| inky | i'll dd it right now. | 13:51 |
| inky | so if i backup /home/user that's enough for all my settings? also wifi passwords? | 13:51 |
| Wizzup | not sure | 13:53 |
| Wizzup | I don't know where icd2 gconf is stored | 13:54 |
| inky | ok at least i hope it won't harm to backup the home. | 13:54 |
| Wizzup | note that there are a few fixes in daedalus-devel (like rotation) | 13:54 |
| Wizzup | those will make it to then next image build I guess | 13:54 |
| inky | yay. thank you. | 13:54 |
| inky | maemo is really useful on pp. | 13:55 |
| dsc_ | ah, my VM is again in this state where it thinks there is no internet connection | 14:19 |
| dsc_ | arno11: I cannot test the addressbook stuff, sorry ^^ | 14:19 |
| dsc_ | maybe something installs during pkg install `purple-telegram` that puts it in this state, i dont know | 14:21 |
| dsc_ | or maybe related to adding telegram via mc-tool | 14:21 |
| dsc_ | https://plak.infrapuin.nl/selif/dqwndovi.png i see this | 14:22 |
| Wizzup | network error means that the plugin did not manage to connect | 14:23 |
| Wizzup | could be for all kinds of reasons | 14:23 |
| Wizzup | dbus will probably give more info, or you can run the connection manager manually | 14:23 |
| inky | $ pwd | 14:27 |
| inky | /home/user/.gconf | 14:27 |
| inky | $ grep -ir uncertain * | 14:27 |
| inky | system/osso/connectivity/IAP/b0cff3ac-7cd4-4652-ba09-72f703513945/%gconf.xml: <stringvalue>uncertain areas</stringvalue> | 14:27 |
| inky | i think it's in home. hope it'll work after dd. | 14:27 |
| inky | uncertain areas is my wifi name. | 14:27 |
| inky | essid* | 14:27 |
| Wizzup | ok | 14:29 |
| Wizzup | keep in mind that blindly rsyncing things back and forth can cause loads of trouble | 14:29 |
| Wizzup | but you can give it a try I guess | 14:29 |
| Wizzup | inky: any idea about pinephone pro? is it the same kernel tree/branch? | 15:23 |
| inky | i think megi said that all the newest changes go to that orangepi branch. | 15:49 |
| inky | https://codeberg.org/megi/linux/commits/branch/orange-pi-6.15 here i see | 15:51 |
| inky | Merge branch 'rk3399-typec-6.15' into orange-pi-6.15 | 15:51 |
| inky | that's a pinebook pro and pinephone pro chipset i think. | 15:51 |
| inky | yes this lists pinephone pro | 15:52 |
| inky | https://codeberg.org/megi/linux/commit/f9936b9f7dc01ad55c530937305c88d6632b6866 | 15:52 |
| arno11 | dsc_: @network error, did you install pidgin first and add tg account through it ? (to be able to receive code). | 16:01 |
| arno11 | if not, you can't log your account through mctool or tg account plugin | 16:02 |
| dsc_ | arno11: i located the error, it was because apparmor was active | 16:02 |
| arno11 | ah | 16:03 |
| arno11 | dsc_: btw @chat wrong title issue: in fact, it always shows the name of the last one who send a msg, which is a bit confusing sometimes | 16:33 |
| dsc_ | arno11: will look later, I need to invest some time into getting a good development environment | 16:33 |
| dsc_ | the VMs are not suitable atm | 16:34 |
| arno11 | ok, i see. np | 16:34 |
| arno11 | it is already usable anyway | 16:35 |
| dsc_ | yesterday I (again) lost $x amount of time, it stops now, I will create my own image :P | 16:35 |
| arno11 | ok :) | 16:35 |
| dsc_ | yes apart from this, latest conversation should be OK | 16:35 |
| arno11 | yeah | 16:36 |
| dsc_ | as in, no drastic bugs :P | 16:36 |
| arno11 | sure | 16:37 |
| arno11 | Wizzup: btw, lost offline sms (cellulard) is maybe a blocker (could be critical for users) | 16:53 |
| arno11 | i already try to start cellulard as late as possible through Xsession.post or from init.d (with some additional deps), no success unfortunately | 16:55 |
| arno11 | i mean, it is never late enough | 16:55 |
| sicelo | arno11: i wouldn't say it's a blocker, because it's not a regression ... we've always had this :-) | 16:58 |
| sicelo | but yes, it's important to fix | 16:58 |
| freemangordon | sorry, what? | 16:59 |
| freemangordon | how is cellulard related to lost sms-es? | 16:59 |
| arno11 | sicelo: yes indeed, not a regression | 17:01 |
| sicelo | freemangordon: if an sms is sent while the device is off, it gets queued in the air. when the modem goes online, the network will deliver it | 17:01 |
| freemangordon | and? ofono should act on that and queue it | 17:02 |
| sicelo | the problem is - as soon as ofono sets the modem online, the sms is delivered, and disappears, unless there's 'someone' looking out for it on dbus | 17:02 |
| freemangordon | no, ofono queues it (or at least it should) | 17:02 |
| sicelo | it doesn't | 17:02 |
| sicelo | ofono doesn't queue anything | 17:03 |
| freemangordon | who is that 'someone'? tp-ring? | 17:03 |
| sicelo | yes | 17:03 |
| freemangordon | so, we shall start tp-ring before ofono, no? | 17:03 |
| sicelo | yes | 17:03 |
| freemangordon | ok | 17:03 |
| freemangordon | btw, why do you think ofono does not queue sms-es? | 17:04 |
| freemangordon | see https://git.maemo.org/leste-upstream-forks/ofono/src/branch/maemo-ofono/doc/messagemanager-api.txt | 17:06 |
| freemangordon | GetMessages() | 17:06 |
| sicelo | i know it doesn't | 17:06 |
| freemangordon | how do you know? | 17:07 |
| sicelo | i've experienced it lots of times, even outside of maemo. | 17:07 |
| freemangordon | like, did you try to call GetMessages() to see it returns empty list? | 17:07 |
| freemangordon | if that's the case, then we have issue in ofofno, no? | 17:07 |
| sicelo | yes :-) | 17:07 |
| freemangordon | *ofono | 17:07 |
| freemangordon | is that n900 specific? | 17:08 |
| sicelo | well i don't know how ofono side really meant for that method to work. i've tried it before and at least at the time i knew nothing about ofono code, and concluded it's a non-working method | 17:08 |
| sicelo | not n900 ... it's just general ofono | 17:08 |
| freemangordon | ok, could you please re-test that method | 17:08 |
| freemangordon | or wait, I can do on d4 I guess | 17:09 |
| freemangordon | have to reboot my PC (to clean-up a fan which is a bit noisy :) ), will look at it later on | 17:09 |
| sicelo | you can try, but assuming you don't have the issue with unACKed SMS ... because in that case, you will get false results | 17:09 |
| freemangordon | I have usb modems as well; | 17:10 |
| sicelo | and, welcome back | 17:10 |
| freemangordon | well, I was an a holiday :) | 17:10 |
| freemangordon | not sure how 'bak' I am, given the RL work load though :( | 17:10 |
| freemangordon | *back | 17:10 |
| freemangordon | brb | 17:10 |
| sicelo | freemangordon: so, i might be wrong, but, GetMessages() runs this function. https://elixir.bootlin.com/ofono/2.17/source/src/sms.c#L1026 | 18:25 |
| freemangordon | yes, seems it lists pending tx messages | 18:26 |
| sicelo | i think that *data is NULL. question would be, where would real data come from :-) | 18:26 |
| freemangordon | however, I don;t see incoming messages being received by the modem after it is powered on | 18:26 |
| freemangordon | this is some ec25 I have here | 18:27 |
| sicelo | plus, what would 'queued' mean, since, afaict, ofono doesn't know/care that there are any clients connected. it just blindly pushes stuff to dbus. tldr; this method doesn't really work, and can't work until ofono can keep track of 'Getted' messages, or 'Getting' clients | 18:28 |
| freemangordon | ofono keeps track of 'outgoing' queue | 18:28 |
| sicelo | but yeah, i'll give this actual testing in a couple of hours on N900, L5, and my laptop (Ericsson modem) | 18:28 |
| Wizzup | arno11: I think the sms thing is there on chimaera too | 18:29 |
| freemangordon | so yes, GetMessages is not what we need | 18:29 |
| Wizzup | freemangordon: btw, we discussed this while you were gone and wondered if cellulard could decide not to online until tp-ring appears on the bus | 18:29 |
| freemangordon | but, first I want to be sure that messages are queued by the network and send to the modem once it is back online | 18:30 |
| freemangordon | Wizzup: lemme first make sure sms-es are delivered at all | 18:30 |
| sicelo | ok, do test, but yes, they are. if you already have a dbus listener before ofono goes online, you get them. at least i have, and i think arno did too | 18:30 |
| freemangordon | ok, ok | 18:31 |
| freemangordon | lemme see what tp-ring does in terms of bringing the modem online | 18:32 |
| Wizzup | tp-ring does not online the modem | 18:37 |
| Wizzup | only cellulard does | 18:37 |
| freemangordon | mhm | 18:50 |
| freemangordon | but the issue is that tp-ring runs as user | 18:50 |
| freemangordon | while cellulard runs as root | 18:50 |
| freemangordon | I am not sure how would that work in multi-user environment. also, how would cellulard know that tp-ring is started?\ | 18:51 |
| Wizzup | It can listen on dbus :) | 18:59 |
| Wizzup | but if you have another idea, I'm all ears | 18:59 |
| Wizzup | The tricky thing is that all tp things are auto started, so we don't really control when they are started, unless we start them explicitly | 19:00 |
| sicelo | I don't see why cellulard should be run as root, but this is not related to this particular issue. afaict, what it does can all be done from a basic user | 19:09 |
| freemangordon | sicelo: because starting it as user means there should be user logged in | 19:14 |
| freemangordon | also, what if we have 2 users? shall we have to cellulard instances? | 19:14 |
| freemangordon | *two | 19:14 |
| freemangordon | Wizzup: I don;t have another idea, but I will try to come-up with some | 19:15 |
| sicelo | ok @ user logged in | 19:15 |
| freemangordon | my gut feeling tells me ofono API is at fault here | 19:17 |
| sicelo | yes, ideally ofono should keep messages until it gets an ack that they have been read | 19:18 |
| freemangordon | mhm | 19:18 |
| sicelo | @ user logged in, if root (cellulard) enables the modem, nothing can use it anyway. e.g. say emergency call comes in ... tp-ring must handle that, and it's running as user. Anyway, not to derail us from main issue | 19:21 |
| freemangordon | you can't have incoming emergency call :) | 19:26 |
| freemangordon | however, you can have auto gprs connection | 19:27 |
| freemangordon | that's handled by icd2 | 19:27 |
| freemangordon | and you don't need user for that | 19:27 |
| freemangordon | sicelo: it is not derailing from the main issue, I would like a discussion to have all arguments | 19:28 |
| freemangordon | also, why tp-ring? | 19:29 |
| freemangordon | what if we don';t want to use TP? | 19:29 |
| arno11 | Wizzup: @sms/chimaera, yep | 22:33 |
| arno11 | sicelo: we are now sure there is no lost sms if ofono goes online after tp ring, but what would happen if tp ring crash/restart for some reason ? lost sms in the meantime because of no listener ? or ? | 22:41 |
| sicelo | absolutely, yes :-) | 22:43 |
| sicelo | if you're not on dbus at the right time, smses just whizz by | 22:44 |
| arno11 | so...i'm agree that ofono is at fault | 22:45 |
| sicelo | yes, i agree totally. i wonder what SFOS does | 22:47 |
| arno11 | could be interesting to know | 22:48 |
| sicelo | and yes, it's not directly about tp-ring ... it's about 'anything' that is responsible for picking up the sms, which in our case is tp-ring. a different platform could just do direct dbus, and have exactly the same problem (if it starts later) | 22:49 |
| arno11 | mhm | 22:54 |
| sicelo | Wizzup: not sure if it's really a blocker, but the inverted battery/current_now sign drives me crazy on the Droid 4, and I won't have time soon to test a kernel patch | 22:56 |
| arno11 | @sms, wonder if we can use some kind of redundancy to workaround the problem | 22:57 |
| sicelo | ah, i can more easily fix the issue in status-menu-applet-battery | 23:24 |
| dsc_ | sicelo: did you ever create an image via image-builder? | 23:59 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
