sicelo | +1 @thanks | 00:58 |
---|---|---|
sunshavi | there is a question about 'DM_USB_GADGET' on the list. | 03:42 |
mighty17[m] | what browser works in maemo (and pvr) | 08:09 |
sicelo | Almost any, depending on RAM usage, availability and your needs | 09:20 |
sicelo | E.g. on droid 4, firefox works reasonably okay. | 09:20 |
mighty17[m] | sicelo: and matrix client? | 10:44 |
sicelo | Same answer :-p | 10:45 |
sicelo | Nheko, | 10:45 |
sicelo | Quarternion, etc. | 10:46 |
mighty17[m] | Nheko doesnt open virtual keyboard :( | 11:20 |
Wizzup | is it qt5? gtk3? | 11:27 |
Danct12 | mighty17[m] on my n900, i use dillo | 11:33 |
bencoh | (dillo <3) | 11:33 |
Danct12 | i kinda want to go with netsurf but i can't find it in beowulf | 11:33 |
bencoh | Yeah, I was about to say it was removed for some reason | 11:34 |
Wizzup | hm, I thought parazyd packaged it | 11:34 |
mighty17[m] | dillo is less resource intensive :D | 11:34 |
bencoh | (either that, or we used to build it from source, but I vaguely recall it was indeed removed) | 11:34 |
bencoh | https://packages.debian.org/search?suite=all&searchon=names&keywords=netsurf | 11:34 |
parazyd | Wizzup: I packaged surf, not netsurf | 12:50 |
Wizzup | parazyd: ah ok | 13:02 |
Wizzup | parazyd: did you upgrade ke-recv over ssh? | 13:19 |
Wizzup | parazyd: I need your daemon.log info | 13:23 |
parazyd | Wizzup: Yes, I did apt update && apt install ke-recv | 13:24 |
parazyd | The latter does an upgrade of ke-recv only | 13:25 |
parazyd | I'm not at home now so can't access the N900 I did it on atm | 13:25 |
Wizzup | parazyd: do we log warnings in syslog? | 13:25 |
parazyd | What do you need from daemon.log? I pasted already two sprunges when the error happened | 13:25 |
parazyd | syslog and messages | 13:25 |
Wizzup | I mean warning level | 13:25 |
parazyd | I don't know | 13:26 |
Wizzup | ke-recv is *very* verbose about all the things that make it fail | 13:26 |
Wizzup | and I cannot find any message in syslog | 13:26 |
Wizzup | which is weird | 13:26 |
Wizzup | one of the thoughts was that maybe if you upgrade through ssh, then ke-recv and/or dsme are restarted with wrong/different env vars and that causes problems | 13:26 |
Wizzup | https://github.com/maemo-leste/libosso/blob/8dd7d38eae9b075c2b4664a2498b1aa9f5fbc836/src/osso-log.h#L70 | 13:26 |
Wizzup | It would be good to check that those log calls actually work I guess | 13:27 |
parazyd | I upgraded both times through ssh | 13:27 |
parazyd | The full upgrade which caused a reboot and only the ke-recv upgrade which didn't cause a reboot | 13:27 |
Wizzup | yes, but the other case it sounded like you only changed ke-recv and no dsme or other stuff | 13:28 |
Wizzup | I really wonder if we want dsme to 'restart' at all during an upgrade | 13:28 |
Wizzup | I don't think Nokia did that a whole lot, if ever | 13:28 |
Wizzup | freemangordon: ^ any thoughts? | 13:29 |
parazyd | IMHO it shouldn't restart at all | 13:29 |
parazyd | uvos and I discussed this | 13:29 |
Wizzup | but it seems to restart currently | 13:29 |
parazyd | Yeah I mean the lifeguard is useless and gives you no freedom to debug | 13:29 |
parazyd | If a service fails | 13:29 |
Wizzup | uh | 13:29 |
parazyd | And rebooting is not a solution to "fix" the system | 13:29 |
Wizzup | I don't think we are on the same page at all | 13:30 |
Wizzup | I was saying that debian should not restart dsme upon upgrade | 13:30 |
Wizzup | I was not talking about dsme restarting | 13:30 |
Wizzup | (the system) | 13:30 |
parazyd | oh ok | 13:30 |
parazyd | This might work with dh_installinit --no-start | 13:30 |
Wizzup | if I can make that not happen when upgrading my device, it would help I think | 13:31 |
parazyd | eh, it already is --no-start, so i guess not | 13:31 |
parazyd | but | 13:32 |
parazyd | I don't recall dsme being restarted actually | 13:32 |
Wizzup | I am trying to see if I can find that in the logs now | 13:32 |
Wizzup | in any case ke-recv seems very verbose with it's logging, so there should be messages about how it fails for sure | 13:32 |
Wizzup | ah we also stubbed stopping dsme it seems | 13:33 |
Wizzup | stop() { ebegin "Stopping dsme" eend 0 | 13:33 |
Wizzup | } | 13:33 |
parazyd | There's also dsme-dbus | 13:34 |
Wizzup | https://github.com/maemo-leste/ke-recv/blob/master/src/ke-recv.c#L673 | 13:34 |
Wizzup | basically everything logs to syslog | 13:34 |
Wizzup | https://github.com/maemo-leste/libosso/blob/8dd7d38eae9b075c2b4664a2498b1aa9f5fbc836/src/osso-log.h#L70 | 13:34 |
parazyd | Yeah | 13:34 |
Wizzup | it looks like signal handler failing to install does attempt to log before opening the log | 13:35 |
Wizzup | but that's about it | 13:35 |
Wizzup | I'm wondering how I can debug this when I attempt an upgrade | 13:36 |
parazyd | Jun 17 14:06:27 localhost ke_recv[2130]: ke-recv.c:801: Returned from the main loop | 13:36 |
parazyd | http://sprunge.us/LrWEpG | 13:36 |
Wizzup | I think that is from: | 13:37 |
Wizzup | static void sigterm(int signo) | 13:37 |
Wizzup | { g_main_loop_quit(mainloop); | 13:37 |
Wizzup | } | 13:37 |
Wizzup | + | 13:37 |
Wizzup | g_main_loop_run(mainloop); | 13:37 |
Wizzup | ULOG_DEBUG_L("Returned from the main loop"); | 13:37 |
parazyd | But that looks fine, yeah | 13:37 |
Wizzup | but we don't see any messages from ke-recv at all after that | 13:37 |
Wizzup | which is really weird | 13:37 |
parazyd | Yes | 13:37 |
parazyd | It just fails from dsme | 13:37 |
Wizzup | hmm | 13:38 |
parazyd | Does anything happen if you just send a sigterm to ke-recv? | 13:39 |
dreamer | parazyd: I'm looking bit more at cargo. so in order to create and update tables you need a 'recreatecargodata' permession on the user. sysops have this, but maybe you can add new group with these perms instead? -> https://www.mediawiki.org/wiki/Extension:Cargo/Storing_data#Permissions | 13:39 |
dreamer | permission* | 13:40 |
parazyd | dreamer: Can I give you admin somehow so you deal with it :) | 13:40 |
Wizzup | parazyd: I can try in a bit, we could also change ke-recv startup policy from -t to -f, which does not cause SW reset if it fails to restart, but that mostly makes sense from a debug pov imho | 13:40 |
parazyd | That could be fine for now | 13:40 |
dreamer | parazyd: hehehehehe. root on the machine hosting the wiki? ;) | 13:41 |
Wizzup | it's a lxc container iirc | 13:41 |
dreamer | same difference | 13:41 |
parazyd | No I mean on the mediawiki | 13:41 |
dreamer | or add me to admin group on the wiki | 13:41 |
parazyd | ok I think I did it | 13:42 |
parazyd | Please have a look | 13:42 |
dreamer | https://leste.maemo.org/User:Dreamer << moi | 13:42 |
dreamer | ok, lets see | 13:42 |
dreamer | `Member of groups:Administrators, Autoconfirmed users, Interface administrators, Users` << hopefully good enough yeah :) | 13:42 |
parazyd | ok | 13:43 |
* dreamer will be poking around some more. probably on some temporary pages to get a feel of how cargo works | 13:43 | |
Wizzup | parazyd: I wonder if ke-revc should use DLOG, not ULOG | 13:43 |
parazyd | https://leste.maemo.org/Special:CargoQuery | 13:43 |
Wizzup | #define ULOG_OPEN(X) openlog(X, LOG_PID | LOG_NDELAY, LOG_USER) | 13:43 |
Wizzup | #define DLOG_OPEN(X) openlog(X, LOG_PID | LOG_NDELAY, LOG_DAEMON) | 13:43 |
dreamer | parazyd: have to start by adding a new table before you can query it ;) | 13:43 |
dreamer | anyway, I'll be poking around. don't expect results asap :P | 13:44 |
parazyd | Wizzup: Maybe it makes sense if it runs as root | 13:44 |
parazyd | dreamer: Sure | 13:44 |
Wizzup | parazyd: I think it also determines where the log messages go | 13:44 |
Wizzup | iiuc | 13:44 |
Wizzup | let me booy my droid and poke around | 13:44 |
parazyd | I think syslog is a concatenation of everything | 13:45 |
* Wizzup doesn't think so | 13:45 | |
Wizzup | but I have to check :) | 13:45 |
parazyd | ok | 13:45 |
parazyd | I'm not sure, but something in my mind tells me that it is :D | 13:45 |
parazyd | Either syslog or messages, don't remember | 13:45 |
Wizzup | messages is just kernel | 13:45 |
Wizzup | syslog I think is not everything either | 13:45 |
Wizzup | I will check | 13:46 |
parazyd | *nod* | 13:48 |
Wizzup | parazyd: daemon.log is not the same as syslog | 13:51 |
parazyd | ok! | 13:51 |
Wizzup | parazyd: ke-recv messages end up in /var/log/user.log lol | 13:54 |
parazyd | ah obviously | 13:54 |
parazyd | ULOG :) | 13:54 |
parazyd | Yeah I didn't even think of checking it | 13:54 |
Wizzup | see /etc/rsyslog.conf | 13:54 |
parazyd | I'll have a look when I have my N900 again | 13:54 |
parazyd | But if you can make the upgrade, you should repro it too | 13:54 |
Wizzup | yeah I understand | 13:56 |
Wizzup | trying to check some other stuff first | 13:56 |
Wizzup | it looks like our rsyslog does not log CRIT to messages at least | 13:57 |
Wizzup | but it does log WARN | 13:57 |
buZz | hmm , apt upgrade just ended with 'Hangup' | 14:05 |
buZz | what a weird msg :P | 14:05 |
Wizzup | buZz: hm, any idea what happened? | 14:07 |
buZz | it was building initrd at the time, it seemed | 14:08 |
buZz | oh huuu | 14:09 |
buZz | it somehow destroyed wifi i think :( | 14:09 |
Wizzup | on the d4, or what? | 14:09 |
buZz | yeah on d4 | 14:09 |
Wizzup | dpkg-reconfigure ? | 14:09 |
buZz | i'll try, just doing another reboot now to be sure | 14:10 |
Wizzup | full system upgrades are difficult, there's a reason nokia only did a few and locked down the system + reboot once it was done :p | 14:11 |
buZz | oh , i remember it updated mce in this upgrade | 14:11 |
buZz | and it failed to shutdown mce | 14:11 |
buZz | what should i dpkg-reconfigure ? | 14:13 |
buZz | what package i mean | 14:13 |
Wizzup | -a ? | 14:13 |
buZz | i tried 'mce' but doesnt seem able to configure it | 14:13 |
Wizzup | well maybe that's too much | 14:13 |
Wizzup | just whatever people do when apt is interrupted | 14:13 |
Wizzup | apt-get install -f? | 14:13 |
* Wizzup 's memory is hazy atm | 14:13 | |
buZz | yeah thats not the issue i think? hmm | 14:14 |
buZz | dpkg-reconfigure linux-image-droid4 now | 14:15 |
buZz | meh, keeps giving 'no connections available' | 14:18 |
buZz | i'm considering just wiping SD and reinstalling ;) | 14:18 |
buZz | or someone maybe has a good idea? :P | 14:18 |
Wizzup | wait | 14:18 |
buZz | i'll wait | 14:18 |
Wizzup | are you on -devel | 14:18 |
Wizzup | parazyd: btw I did apt upgrade just now and didn't get that problem | 14:19 |
buZz | i am | 14:19 |
buZz | i had a 24 package update | 14:19 |
Wizzup | buZz: dpkg -l | grep suppli | 14:19 |
Wizzup | I wonder what the two versions are | 14:19 |
Wizzup | for wpa supplicant and the licdicd-network-wpasupplicant | 14:20 |
buZz | wpa supplicant is 2.9.0-21+2m7.1 | 14:20 |
buZz | libicd-net-wpas is 0.1.7+2m7 | 14:20 |
buZz | wpa supplicant is 2:2.9.0-21+2m7.1 | 14:21 |
Wizzup | that should work, at least that part of the stack | 14:21 |
buZz | lol | 14:21 |
buZz | ok | 14:21 |
buZz | anything else you'd like me to try? | 14:21 |
Wizzup | does wpa_cli work? | 14:21 |
Wizzup | 'scan', 'scan_results', etc | 14:21 |
buZz | yes | 14:22 |
Wizzup | do you see networks? | 14:22 |
buZz | yep, all the same as normal | 14:22 |
Wizzup | ok, but not from ui? | 14:22 |
buZz | indeed | 14:22 |
Wizzup | is icd2 running? | 14:22 |
Wizzup | pidof icd2 | 14:23 |
buZz | it is | 14:23 |
Wizzup | hm | 14:23 |
buZz | '/usr/sbin/icd2 -l2' | 14:23 |
Wizzup | and apt reports nothing else to fix/do? | 14:23 |
buZz | as far as apt is concerned , everything is ok, it seems | 14:23 |
buZz | apt -f install doesnt want to execute anything | 14:24 |
Wizzup | can you reconfigure icd2 and those other packages? | 14:24 |
buZz | ehm | 14:24 |
buZz | mayyybe, lets see | 14:24 |
buZz | :P | 14:24 |
parazyd | Wizzup: hmm ok | 14:26 |
buZz | oh huh | 14:26 |
buZz | i have linux-image-droid4 5.11.2 installed | 14:27 |
buZz | but uname -a tells me i'm running 5.11.22 | 14:27 |
parazyd | uname != package version | 14:27 |
buZz | oh right, ok | 14:27 |
bencoh | that's pretty confusing thoug | 14:27 |
Wizzup | parazyd: random q, how can I enter charging mode from kexecboot on d4? I don't see the menu entry | 14:27 |
parazyd | Yeah, I'm thinking of setting it up to always keep the pkg version at 5.11.0 | 14:27 |
parazyd | Debian does it this way too | 14:27 |
buZz | i've dpkg-reconfigure'd everything with networking in my last apt upgrade | 14:28 |
bencoh | yeah, but that's confusing as well tbh | 14:28 |
buZz | but nope, nothing changed | 14:28 |
buZz | :( | 14:28 |
parazyd | Wizzup: We didn't push that update yet | 14:28 |
Wizzup | parazyd: to -devel ? ok | 14:28 |
Wizzup | and emergency shell should work, right? | 14:28 |
parazyd | You can put softlevel=charge-mode in boot.cfg cmdline after you install charge-mode | 14:28 |
parazyd | Yes | 14:28 |
Wizzup | I'm booting it now for the first time :) | 14:28 |
Wizzup | I think it's booting normally though | 14:28 |
Wizzup | maybe I don't know how to use kexecboot keys :D | 14:29 |
parazyd | Well, you'll still see a tty | 14:29 |
Wizzup | I see X and everything, h-d | 14:29 |
parazyd | ah then it passed through | 14:29 |
parazyd | uvos said this happens sometimes | 14:29 |
Wizzup | in any case nvm for now, I'm not too interested in it atm | 14:29 |
parazyd | And ofc you need to be charging for it to actually show up | 14:30 |
parazyd | But he said sometimes the current draw is too high(?) so it skips | 14:30 |
Wizzup | aha | 14:30 |
parazyd | Only a mapphone problem | 14:30 |
bencoh | silly question: is https://maedevu.maemo.org/pkgweb/ supposed to show up as application/octet-stream? | 14:34 |
bencoh | firefox suggests to download it, instead of displaying the page | 14:34 |
Wizzup | seems like missing index.html default or some | 14:40 |
parazyd | oh sry | 14:41 |
parazyd | misconfig | 14:41 |
parazyd | Fixed | 14:41 |
parazyd | bencoh: Try now | 14:43 |
bencoh | yay, works! thanks :) | 14:44 |
parazyd | :) | 14:44 |
bencoh | it's lovely and pretty useful, by the way | 14:44 |
bencoh | did you reuse code from the devuan infra? | 14:44 |
bencoh | (oh and, you might want to display the package arch in the listing) | 14:45 |
parazyd | The architecture is in the URL | 14:47 |
parazyd | And if you open a package's page you also see it listed | 14:47 |
parazyd | https://maedevu.maemo.org/pkgweb/beowulf/main/binary-amd64/hildon-desktop.html | 14:47 |
parazyd | We reused a Go program written for Devuan and modified it a bit for our usecase: https://github.com/maemo-leste/mpi/ | 14:47 |
parazyd | I wrote the search in a very barebones, filesystem-as-db way though | 14:48 |
parazyd | So it's not the fastest thing, but it does the trick | 14:49 |
bencoh | neat | 14:49 |
bencoh | (and I yeah, I was only referring to the listing itself, regarding the arch) | 14:49 |
buZz | so no ideas how i can get icd2 working again? :P then i'll just reinstall from scratch, nothing unique on my install atm | 15:18 |
bencoh | did you try purging/reinstalling? | 15:18 |
buZz | which packages? the libicd- one ? | 15:20 |
bencoh | yeah | 15:20 |
buZz | wpa_cli is working fine , btw | 15:20 |
bencoh | err, actually I dont know which of those package hosts the config files | 15:20 |
buZz | i bet icd2-settings-default and icd2-network-wpasupplicant-config | 15:21 |
buZz | 1.11+2m7 and 0.1.7+2m7 , are the ones i have | 15:21 |
bencoh | ah, you're on devel | 15:22 |
buZz | i am | 15:25 |
Wizzup | parazyd: so the provider will create a second entry in the UI for every network it maches | 15:42 |
Wizzup | parazyd: this gets messy quickly with lots of wifi APs | 15:42 |
Wizzup | since you now get double the amount of connections | 15:43 |
_uvos_ | Wizzup: so on mapphones the charging state in sysfs is set by the kernel simply to "charging" or not depending on current_now. ofc that is very noizy so this value flickers back and fourth all the time. some times when charging_mode reads the state it will happen to chance upon the value while it not charging and simply boot to desktop | 15:46 |
_uvos_ | we can either fix this somehow, or enter charing mode depending on usb connected state instead of charging state or both | 15:47 |
Wizzup | right | 16:02 |
parazyd | https://twitter.com/vm_call/status/1405937492642123782 | 16:32 |
parazyd | Wizzup: ok | 16:33 |
parazyd | Does this apply to existing networks that have a saved connection, or it comes up on scan for all of them? | 16:33 |
parazyd | In any case, we could add some state and tweak connui | 16:34 |
Wizzup | parazyd: all of them | 16:55 |
parazyd | ok | 17:18 |
parazyd | Wizzup: So could the code be modified/copied to work like one-for-all networks | 17:19 |
parazyd | Perhaps modifying is better since nothing else ever used this code | 17:24 |
parazyd | I suppose the way it works now is allowing you to set a provider per-network | 17:27 |
parazyd | Perhaps there is just a way to have it show up when you're connected to a network | 17:28 |
parazyd | That sounds ideal | 17:28 |
parazyd | And could be a simple connui tweak | 17:28 |
Wizzup | maybe it could be a dialog + pref | 19:38 |
buZz | dbus starts with 'unknown username 'geoclue'' | 19:53 |
buZz | on 6-20 image on droid4 | 19:54 |
buZz | dont think its anything serious , seems to boot fine | 19:54 |
buZz | at least wifi works now \o | 19:55 |
buZz | :) | 19:55 |
buZz | the new font on the theming is nice! | 19:56 |
freemangordon | Pali: well, at least Tom looks like a sane guy | 21:02 |
Pali | :-) | 21:16 |
sicelo | parazyd: by the way, I have permanent Location button on my status menu. Not sure if that's expected | 21:28 |
freemangordon | I guess not | 21:29 |
parazyd | Yeah it shows always, but it's easy to hide it | 21:33 |
parazyd | We can do that if it's necessary | 21:34 |
parazyd | lmk | 21:34 |
freemangordon | parazyd: well, in fremantle that button is only visible when there is active GPS session | 21:36 |
sicelo | I'm not used to seeing it all the time, so yes :-) | 21:36 |
parazyd | ok | 21:37 |
parazyd | I'll add that | 21:37 |
freemangordon | parazyd: lots of status menu items are like that, which kinda makes sense, otherwise status menu will easily get crowded | 21:37 |
sicelo | Thanks | 21:37 |
parazyd | freemangordon: Yeah, makes sense. | 21:38 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!