| tmlind | freemangordon: ok great if dtmf is there too, so really it's only the audio remaining then | 05:53 |
|---|---|---|
| tmlind | hmm so both CLVL and VTS are on n_gsm dlci 2, maybe there's some command similar to CLVL on the ttyUSB4? | 05:56 |
| subbollo | Hi | 06:34 |
| subbollo | N900 gets stuck 2 minutes after hildon-desktop loads | 06:36 |
| subbollo | Even after setting the apt-worker interval to a crazy amount | 06:37 |
| freemangordon | tmlind: even if there is, I can't find it :( | 06:50 |
| subbollo | It maybe wifi related. When it succesfully connects to the internet it just works | 06:50 |
| freemangordon | subbollo: do you have enough free space? | 06:51 |
| freemangordon | tmlind: I see n_gsm in upstream, what does upstream lack in terms of motmdm? | 06:52 |
| subbollo | freemangordon: yes i do. The only thing that seems to fix it is if it finds a known wifi network and it has internet | 06:56 |
| freemangordon | weird. I am not aware of any wifi issue on n900 | 06:57 |
| freemangordon | but I am not using it either | 06:57 |
| freemangordon | lets wait for arno11 or sicelo, then may have ideas | 06:58 |
| sicelo | there's no wifi issue. | 07:05 |
| sicelo | subbollo: the device is just testing your patience :-) | 07:06 |
| subbollo | It maybe wifi related. When it succesfully connects to the internet it just works | 07:09 |
| subbollo | [12:06:09 a.m.] <subbollo> Could it be that apt-worker waits for internet connection before even checking for the interval?Wiki warns about wating til second boot to connect to a network | 07:09 |
| subbollo | Sry lmao | 07:10 |
| sicelo | the thing is ... n900 gets super slow when there's a lot happening simultaneously, such as at boot or when an internet connection is established | 07:11 |
| sicelo | so just be patient | 07:12 |
| subbollo | But ive waited for 20+ minutes when theres no known wifi around and it stays stuck. Not even the kb turns off when slid | 07:12 |
| sicelo | the lag at boot, just after desktop starts showing, is actually there even on Droid 4 ... even Librem5 | 07:13 |
| subbollo | Oh yeah that reminds me | 07:13 |
| freemangordon | sicelo: not here (lag on d4) | 07:14 |
| subbollo | How can i make a battery for the droid 4 | 07:14 |
| * sicelo re-reads the problem statement | 07:14 | |
| freemangordon | subbollo: I think somebody made pictures, but not woken-up enough to remember who | 07:15 |
| freemangordon | Wizzup: ^^^ | 07:15 |
| freemangordon | subbollo: in short - you need a new battery that's compatible with d4 battery | 07:16 |
| freemangordon | then you dis-assemble the original battery, carefully remove the electronics from there and solder it to the cell of the new battery | 07:17 |
| subbollo | I found a lot of d4 batteries on aliexpress but idk if they are even real | 07:17 |
| freemangordon | even if they are, they are so old that it makes no sense | 07:17 |
| sicelo | freemangordon: it's definitely there. boot device, and notice that you cannot launch application list very soon after the desktop is loaded. pressing the icon for it takes a little while to respond. it not a lot, but it's there. | 07:17 |
| subbollo | But the problem is the flimsy connector that kept breaking | 07:17 |
| subbollo | Not the battery itself | 07:18 |
| freemangordon | sicelo: well, if you name half a second delay 'lag', then so be it :) | 07:18 |
| sicelo | it's not there later on. only at boot | 07:19 |
| freemangordon | again, that's not like device not responding for 2 minutes | 07:19 |
| freemangordon | maybe depends on the speed of uSD | 07:20 |
| sicelo | subbollo: so, you say device is unresponsive until you have internet connection? | 07:20 |
| subbollo | Thanks a lot. Please test the issue on the n900 if u can. Take ur n900 away from any wifi it knows and boot it | 07:20 |
| subbollo | sicelo yes | 07:20 |
| freemangordon | subbollo: did you do any 'improvements'? | 07:21 |
| sicelo | actually most of the time I'm not connected to wifi while using N900 | 07:21 |
| subbollo | Yes i did both the .nomedia and check_interval set to max int value | 07:22 |
| subbollo | Hold up lemme check again real quick brb | 07:22 |
| freemangordon | tmlind: also, I was under the impression that n_gsm patches I made 2 years ago are already upstream :). Shall I send them? | 07:24 |
| sicelo | subbollo: maybe you should see what htop shows | 07:27 |
| sicelo | i | 07:27 |
| sicelo | I.e. without internet, I would expect htop to show you the hyperactive process that's looking for an internet connection | 07:28 |
| freemangordon | tmlind: maybe re-send ngsm servdev patches, I can take over after initial comments if you don't have time | 07:36 |
| arno11 | subbollo: interesting @wifi | 08:31 |
| arno11 | with devel upgrade, hildon desktop is unresponsive for around a minute or so. but a lot more if the usdcard is slow | 08:34 |
| arno11 | *on n900 | 08:34 |
| freemangordon | uvos: do you know what AT$QCTER does? | 08:47 |
| freemangordon | it seems it switched the modem to some weird mode | 09:04 |
| freemangordon | but I cannot get it out of there :) | 09:04 |
| freemangordon | modem AT interface that is | 09:04 |
| freemangordon | I tried all baudrate etc combos, to no use | 09:05 |
| arno11 | hmm, last n900 img is from sept and was half broken iirc | 09:18 |
| arno11 | it could explain subbollo issues | 09:22 |
| Wizzup | freemangordon: maybe some qc testing? | 09:45 |
| freemangordon | Wizzup: can I reflash the modem? | 10:37 |
| Wizzup | probably, but I don't know anything about this | 10:38 |
| freemangordon | also, it is not that it does not respon to commands | 10:38 |
| Wizzup | I have 3-4 more droid4s here at your immediate disposal if that helps unblock | 10:38 |
| freemangordon | *respond | 10:38 |
| freemangordon | but in a weird way | 10:38 |
| freemangordon | its like the baudrate or line discipline is different than default | 10:38 |
| freemangordon | according to what I have found over the inet that command changes the baudrate only | 10:39 |
| dsc_ | freemangordon: do you happen to know how vkeyboard gives the result to the application? | 10:50 |
| dsc_ | on a conceptual level | 10:51 |
| dsc_ | you type "AAAA" in vkb, close vkb, "AAAA" is inserted into the client through what mechanism? | 10:51 |
| Wizzup | maybe HILDON_IM_INSERT_UTF8 or so? | 10:52 |
| dsc_ | yes I saw that one, which is a x11 event, but I'm not receiving HILDON_IM_INSERT_UTF8 when monitoring for it | 10:52 |
| Wizzup | if (event->xclient.message_type == ATOM(_HILDON_IM_INSERT_UTF8) && | 10:54 |
| Wizzup | event->xclient.format == HILDON_IM_INSERT_UTF8_FORMAT) { | 10:54 |
| Wizzup | qHimDebug() << "HIM: x11FilterEvent( HILDON_IM_INSERT_UTF8_FORMAT )"; | 10:54 |
| Wizzup | from qt-x11-maemo/src/gui/inputmethod/qhildoninputcontext_x11.cpp | 10:54 |
| dsc_ | yes | 10:54 |
| Wizzup | so this is indeed not dbus as I said earlier, but it's all x11 client messages | 10:54 |
| Wizzup | in any case - that particular client message type is used to insert arbitrary strings | 10:54 |
| dsc_ | its reasonable to expect such event when closing the keyboard but its not being emitted | 10:55 |
| Wizzup | being emitted by who/what? | 10:55 |
| dsc_ | by him | 10:55 |
| Wizzup | how do you close the keyboard? | 10:55 |
| dsc_ | by clicking outside of the vkb | 10:56 |
| Wizzup | ok, let me check if that is indeed how this ought to be done | 10:56 |
| dsc_ | it works in the terminal + vkb | 10:56 |
| Wizzup | ok | 10:56 |
| dsc_ | https://github.com/maemo-leste/hildon-input-method/blob/33cc6ec1cfe5a6a5c5c576edee586a5082befba4/src/hildon-im-ui.c#L2112 | 10:58 |
| dsc_ | ill debug this | 10:58 |
| dsc_ | or just the communication in general | 10:59 |
| dsc_ | to see the happy flow | 10:59 |
| dsc_ | happy path* ;p | 11:00 |
| Wizzup | ok, I can try to do some write up of how I understood this to work back when, I can probably say that I know more about x11 client events than dbus, lol | 11:01 |
| dsc_ | the plugin installs an input event handler globally | 11:03 |
| dsc_ | then it interprets incoming xcb/x11 events | 11:04 |
| dsc_ | and determines if they come from vkb or Qt | 11:04 |
| Wizzup | yes, that is correct, that is what I also did, iirc you need to tell h-i-m that you are a 'h-i-m capable' program | 11:06 |
| dsc_ | if the vkb is active, it will eat up events | 11:06 |
| Wizzup | for this you need to set certain X11 atoms | 11:06 |
| dsc_ | right | 11:06 |
| dsc_ | that might actually be my issue yeah | 11:07 |
| Wizzup | although I think if the vkb is raised, you might have already set it - unsure | 11:10 |
| Wizzup | I think you're basically where I got stuck :D | 11:11 |
| Wizzup | basically at the point where I got stuck* | 11:11 |
| dsc_ | yeah I can raise it just fine | 11:12 |
| dsc_ | altho manually | 11:12 |
| dsc_ | not via focus | 11:12 |
| dsc_ | still, regardless of focus I'm expecting some sort of result | 11:12 |
| dsc_ | after dismissing the vkb | 11:12 |
| Wizzup | oh, okay, I see, how do you raise it manually? | 11:13 |
| dsc_ | its just `sendHildonCommand(HILDON_IM_SHOW, w);` | 11:13 |
| dsc_ | where w is QWidget* | 11:13 |
| dsc_ | i dont understand that QWidget* part yet | 11:14 |
| dsc_ | QLineEdit probably inherits from QWidget | 11:14 |
| Wizzup | can you link to the code in question? or tell me where to look? | 11:14 |
| dsc_ | yeahg sec | 11:14 |
| dsc_ | Wizzup: https://github.com/kroketio/vkb/tree/master | 11:17 |
| dsc_ | https://github.com/kroketio/vkb/blob/master/src/lib/qhildonplatforminputcontext.cpp#L51 | 11:17 |
| dsc_ | cmake -Bbuild . | 11:18 |
| dsc_ | make -Cbuild -j4 | 11:18 |
| dsc_ | ./build/bin/hello | 11:18 |
| dsc_ | this is never called: https://github.com/kroketio/vkb/blob/master/src/lib/qhildonplatforminputcontext.cpp#L343 | 11:18 |
| dsc_ | press the btn to open the keyboard :P | 11:19 |
| dsc_ | https://plak.infrapuin.nl/selif/d6m4hj1g.txt | 11:22 |
| dsc_ | probably missing a bunch of setup to massage vkb into working properly | 11:23 |
| dsc_ | so basically you pass vkb a reference to the client? | 11:23 |
| dsc_ | so it knows where to send events to? | 11:24 |
| dsc_ | well, it already knows because we're receiving events | 11:24 |
| Wizzup | is this standalone program a copy of my qt5 wip? | 11:25 |
| Wizzup | just for me to understand | 11:25 |
| dsc_ | its a standalone program with parts from your WIP, but also parts from https://github.com/community-ssu/qt-x11-maemo/ | 11:26 |
| dsc_ | so its not a plugin, but it shouldnt matter I think | 11:27 |
| dsc_ | because we're only testing communication right now | 11:27 |
| dsc_ | the plugin is only so that its applied to any Qt program running system-wide | 11:30 |
| Wizzup | yes | 11:38 |
| subbollo | I just tested it and it really looks like it requires an internet connection in the first 1.5 minutes after hildon-desktop shows the background. I tested with a wifi with no internet and it gets indefinitely stuck too. | 14:31 |
| subbollo | Nvm idk wtf goin on | 14:32 |
| subbollo | But some aint right | 14:33 |
| subbollo | Anyway, the way i got to this stuckment is i installed some image that came with 6.1.x kernel. Then did a normal apt- get upgrade, then found out about the wiki and enabled -devel and did a dist-upgrade to get overclock, but somehow that didnt upgrade kernel so i ended up doing a apt-get upgrade again to fix overclock | 14:35 |
| subbollo | Could it alp be my dumb u1 sdcard? | 14:36 |
| Wizzup | subbollo: what device, what image did you start with | 14:40 |
| arno11 | Wizzup: last time i did a fresh install, upgrade to devel was 'broken' by gconf2 and dbus errors. reconfiguring them solved the upgrade but device was unstable with input methods not working and extrem slowness. | 14:44 |
| arno11 | another guy reported the same symptoms on TMO | 14:44 |
| Wizzup | I see, and what device is this? n900? | 14:45 |
| Wizzup | oh, sorry, diff. user I suspect | 14:45 |
| arno11 | on n900 that is | 14:45 |
| Wizzup | yeah, for you I know :D | 14:46 |
| arno11 | :) i mean the 2 other users are on n900 | 14:46 |
| arno11 | subbollo: i recommend you to wait a new img. overclock is cool, but better have a stable and smooth device before using it | 14:48 |
| arno11 | i mean it is smooth enough with no overclock actually (when the install is ok ofc) | 14:50 |
| freemangordon | Wizzup: re my modems(s) | 15:29 |
| freemangordon | they behave very weird, it,s like 2 devices are connected in parallel or something | 15:29 |
| subbollo | N900 and i cant remember what image but it couldve been the april one cause i flashed it in june | 15:30 |
| arno11 | ah ok. so what did you do exactly ? fresh install, no wifi on first boot, expand card, reboot, upgrade ? (you should directly dist-upgrade to devel btw) | 15:34 |
| arno11 | and reboot before tweaking anything | 15:35 |
| arno11 | ah and you should check if swap is activated btw | 15:37 |
| subbollo | Yes fresh install and expand on first boot. Yes i did reboot and left it that way after finding out it was slow. And i just picked it up a couple days ago and started tweaking, and did the .nomedia and apt-worker check_interval fix. Then later on i upgraded kernel to 6.6 in order to get overclock to work, cause the boost file wouldnt appear on 6.1. | 15:43 |
| subbollo | I really noticed this problem after the dist-upgrade i did that didnt upgrade the kernel. Basically it was like this: flashed "april" image, expanded card, connected to wifi, installed emacs, left it that way for like 5 months, apt update then apt upgrade, some packages didnt configure well, did a dpkg --reconfigure -a, enabled -devel, dist-upgrade | 15:43 |
| subbollo | that didnt upgrade kernel, apt update and apt upgrade. | 15:43 |
| Wizzup | yeah, the dpkg configure problem with gconf is annoying | 15:43 |
| Wizzup | this should not happen on later images at least | 15:43 |
| subbollo | Its either something wifi related at boot time, my u1 card that somehow gets it stuck forever or idk wtf goin on | 15:46 |
| arno11 | Wizzup: it happened to me with the last img from sept iirc | 15:47 |
| arno11 | subbollo: what kind of u1 btw ? | 15:48 |
| subbollo | The adata 32gb pink and orange one | 15:48 |
| arno11 | probably too slow unfortunately, specially when updating | 15:51 |
| arno11 | but anyway that's not the main issue actually | 15:52 |
| tmlind | hmm so are the usb qmi kernel and ofono in some devel repo? i have deb https://maedevu.maemo.org/leste chimaera-devel main contrib non-free droid4 | 16:09 |
| tmlind | freemangordon: did a reboot clear the modem? you can always reflash the modem if nothing else helps | 16:09 |
| tmlind | freemangordon: feel free to take over the ngsm stuff | 16:12 |
| subbollo | How can i get openmediaplayer to scan my music? | 18:14 |
| Wizzup | it think it should do that automatically, but it won't scan everything, just some directories | 18:15 |
| subbollo | What are those directories? | 18:21 |
| subbollo | Did the .nomedia thing prevents it from scanning? | 18:21 |
| Wizzup | let me pull up gsettings and check | 18:26 |
| Wizzup | $ gsettings get org.freedesktop.Tracker.Miner.Files index-recursive-directories | 18:28 |
| Wizzup | ['&DESKTOP', '&DOCUMENTS', '&MUSIC', '&PICTURES', '&VIDEOS'] | 18:28 |
| Wizzup | $ gsettings get org.freedesktop.Tracker.Miner.Files index-single-directories | 18:28 |
| Wizzup | ['$HOME', '&DOWNLOAD'] | 18:28 |
| Wizzup | those are XDG dirs | 18:28 |
| Wizzup | you can add to this and it will get honoured | 18:28 |
| sixwheeledbeast | is tracker config extras package available in leste? | 18:53 |
| freemangordon | tmlind: I reflashed it (I think so), didn't help | 20:44 |
| freemangordon | flash-droid-4-fw.sh, I think that's made by you | 20:45 |
| freemangordon | tmlind: I think kernel/ofon are still not in ther repos | 20:46 |
| freemangordon | Wizzup: ^^^ | 20:46 |
| freemangordon | will you build kernel | 20:47 |
| freemangordon | tmlind: is doing 'fastboot flash radio radio.img' enough to flash the modem? | 20:49 |
| freemangordon | ofono you should build from https://github.com/maemo-leste-upstream-forks/ofono/tree/upstream-wip | 20:54 |
| Wizzup | did you flash radio.img? | 20:54 |
| Wizzup | or whatever it is | 20:55 |
| freemangordon | Wizzup: yes, see 2 lines ^^^ | 20:58 |
| freemangordon | actually I did flash-droid-4-fw.sh | 20:59 |
| freemangordon | part of it is flashing radio.img | 20:59 |
| freemangordon | no use | 20:59 |
| freemangordon | seems some nvram is corrupted or dunno | 20:59 |
| Wizzup | oh, maybe dump nvram | 21:00 |
| freemangordon | how? | 21:00 |
| Wizzup | https://github.com/tmlind/tcmdrw | 21:01 |
| Wizzup | maybe compare | 21:01 |
| Wizzup | that's all I know | 21:01 |
| Wizzup | I wonder if radio is maybe wifi | 21:02 |
| freemangordon | hmm, dunno | 21:05 |
| freemangordon | I guess no, I think I saw some lte/gprs strings relatot to it | 21:05 |
| freemangordon | *related | 21:06 |
| Wizzup | ok | 21:06 |
| freemangordon | will try nvram thing | 21:06 |
| Wizzup | here is how we use it https://leste.maemo.org/Motorola_Droid_Bionic#Unlocking_the_modem_for_usage_in_Europe | 21:06 |
| Wizzup | but it can dump whole regions | 21:07 |
| freemangordon | it really looks like whole config is garbage | 21:08 |
| freemangordon | and it looks like it types commands to itself | 21:08 |
| freemangordon | it is really weird | 21:08 |
| freemangordon | as soon as I attach minicom and issue ATI, I get response to ATI, but then several more OK and ERROR messages | 21:09 |
| freemangordon | like if it thinks its own response is commands from the user | 21:09 |
| tmlind | freemangordon: yeah flashing radio.img should be enough | 21:30 |
| tmlind | maybe you put it into some echo mode like dlci 7 is on ngsm? | 21:30 |
| tmlind | i think flashing radio.img just copies it to device, you need to let it reboot to flasher that runs the update-binary 3 1 radio.zip on the file extracted from the radio.img while modem is in uart mode | 21:35 |
| tmlind | if you do that from linux you need to restart the modem 2 or 3 times after flashing | 21:36 |
| tmlind | to boot modem into flash mode, you need to modprobe phy-mapphone-mdm6600 flash_mode=3 with the modem flash patch i posted several years ago, need to find it | 21:38 |
| subbollo | Excuse me but does the tracker follow sym links? I changed xdg_music_dir to point to a symlink and openmediaplayer still dont show none | 22:14 |
| subbollo | Bro i cant even get irc think working on. Maybe i just need to stop and get an iphone | 22:26 |
| gnarface | does it really not follow symlinks? | 22:30 |
| gnarface | you sure you're not just missing some other XDG_* variable? | 22:30 |
| gnarface | i'm not entirely clear on it, but i was under the impression that you usually have to set either XDG_RUNTIME_DIR or several others at once | 22:31 |
| arno11 | subbollo: please wait for a new img. don't waste your time installing stuff on an half broken install | 22:51 |
| arno11 | (specially if your device is slow) | 22:52 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!