norayr | https://www.openwall.com/lists/oss-security/2022/10/13/2 | 01:24 |
---|---|---|
norayr | wlan stack issue, allows remote execution of code. | 01:25 |
freemangordon | Wizzup: what is this 'lima' var supposed to do? | 08:04 |
freemangordon | like, don;t we use the same mesa on all devices? | 08:04 |
Wizzup | I think it includes the lima component for mesa | 08:04 |
Wizzup | we have it for pinephone so I added it | 08:04 |
freemangordon | but isn;t it already there? | 08:04 |
freemangordon | I mean - in mesa for n900, for example? | 08:05 |
Wizzup | deb https://maedevu.maemo.org/${codename} ${release} main contrib non-free ${lima} ${devname} | 08:05 |
freemangordon | isn;t that obsolete? | 08:05 |
Wizzup | freemangordon: probably, I just did it anyway :) | 08:05 |
freemangordon | that's my point | 08:05 |
freemangordon | ah | 08:05 |
freemangordon | ok | 08:05 |
freemangordon | ok, I found SD card on my desk, and building u-boot is ok | 08:05 |
freemangordon | Wizzup: the issue is with dependencies | 08:05 |
freemangordon | they say we need some 'multilib' gcc | 08:06 |
freemangordon | but later on it fails to fing hardfp gcc | 08:06 |
freemangordon | *find | 08:06 |
freemangordon | now I only need a kernel | 08:06 |
freemangordon | Wizzup: will arm-sdk build DTBs for us? | 08:13 |
freemangordon | hmm, looks like yes | 08:14 |
Wizzup | freemangordon: it might, but it usually fails | 08:22 |
Wizzup | freemangordon: I was thinking of using the domains as layed out here for the talk https://wiki.maemo.org/Documentation/Maemo_5_Developer_Guide/Architecture/Top_Level_Architecture | 08:38 |
Wizzup | seems sensible, yeah? | 08:38 |
Wizzup | of course it would need to be updated | 08:38 |
freemangordon | do you have time to update that now? | 08:44 |
Wizzup | not the whole diagram no :) | 08:45 |
freemangordon | Wizzup: where shall I change the dependency so gcc-8-arm-linux-gnueabihf to be pulled instead of gcc-multilib in our image builder? | 08:46 |
freemangordon | that allows u-boot to be build | 08:46 |
Wizzup | I am not sure, it looks like various builds don't build u-boot or kernel because of missing packages | 08:47 |
Wizzup | I thought it was the resolv.conf problem that prevented it from being installed | 08:48 |
Wizzup | but that doesn't seem to be the case | 08:48 |
freemangordon | not, it is missing cross-compiler | 08:48 |
freemangordon | at least that's whaty happened here, in leste VM I am using for build | 08:48 |
Wizzup | yeah, that seems relevant, but I don't get really know where this is set up in arm-sdk or if it's even in arm-sdk | 08:48 |
freemangordon | no, it is in the VM we use | 08:49 |
freemangordon | we just need to install that there | 08:49 |
Wizzup | I can do that | 08:49 |
freemangordon | I guess | 08:49 |
freemangordon | please do and re-enable u-boot | 08:49 |
freemangordon | now I am trying to compile the kernel | 08:49 |
freemangordon | 5.18.y | 08:49 |
Wizzup | so I just install gcc-8-arm-linux-gnueabihf on the host? | 08:49 |
freemangordon | yes | 08:49 |
Wizzup | on the 32 bit image builder I guess | 08:49 |
freemangordon | this will remove gcc-multilib | 08:49 |
freemangordon | yes | 08:49 |
Wizzup | on the 64 bit one we also need a way to build arm64 kernel(s) | 08:50 |
Wizzup | since the raspi4 ones also fail | 08:50 |
Wizzup | ok, it's installed | 08:50 |
freemangordon | do we have failure log? | 08:50 |
Wizzup | yes | 08:50 |
freemangordon | ok, I will look at it later on | 08:50 |
Wizzup | https://phoenix.maemo.org/view/Images/job/leste-image-raspi4-64bit/117/consoleText | 08:50 |
Wizzup | so you want me to re-enable u-boot and rebuild the sunxi image? | 08:51 |
freemangordon | yes | 08:51 |
freemangordon | umm...: "Finished: SUCCESS" | 08:51 |
Wizzup | scroll up | 08:51 |
freemangordon | ok | 08:51 |
Wizzup | btw, the sunxi build failed in some python thing for me | 08:51 |
Wizzup | (u-boot) | 08:51 |
Wizzup | Traceback (most recent call last): | 08:52 |
Wizzup | File "scripts/dtc/pylibfdt/setup.py", line 21, in <module> | 08:52 |
Wizzup | from distutils.core import setup, Extension | 08:52 |
Wizzup | ModuleNotFoundError: No module named 'distutils.core' | 08:52 |
Wizzup | make[3]: *** [scripts/dtc/pylibfdt/Makefile:27: scripts/dtc/pylibfdt/_libfdt.so] Error 1 | 08:52 |
Wizzup | that is why I disabled it | 08:52 |
Wizzup | and yes we could install that package | 08:52 |
freemangordon | make: aarch64-linux-gnu-gcc: Command not found | 08:52 |
freemangordon | exactly the same issue | 08:53 |
freemangordon | missing cross-compiler | 08:53 |
Wizzup | that is for kernel, but u-boot also fails | 08:53 |
freemangordon | Wizzup: apt install gcc-aarch64-linux-gnu | 08:53 |
* freemangordon checks | 08:53 | |
Wizzup | https://phoenix.maemo.org/view/Images/job/leste-image-sunxi/3/consoleText | 08:54 |
Wizzup | I can install python3-distutils, but idk if it's using python 3 | 08:55 |
freemangordon | lemme check if I have it here | 08:55 |
Wizzup | both python2 and python3 have distutils, on the host | 08:55 |
Wizzup | ah, python3 does not have distutils.core | 08:56 |
Wizzup | I solved it with python3-distutils | 08:56 |
Wizzup | shall I issue a new build of the sunxi image? | 08:58 |
freemangordon | I have python3-distutils 3.7.3-1 here | 08:58 |
freemangordon | yes, lets check that at least u-boot is ok | 08:58 |
freemangordon | in the meanwhile I am checkoing if kernel bui;ds | 08:58 |
freemangordon | tooks a while, my machine is slow :) | 08:59 |
freemangordon | *takes | 08:59 |
Wizzup | image builds also take 1-2 hours | 08:59 |
freemangordon | ugh | 08:59 |
freemangordon | I tried to run image builder on my laptop with ubuntu | 08:59 |
Wizzup | hence me reverting to uglier fixes ealier | 08:59 |
freemangordon | no dice | 08:59 |
Wizzup | :) | 08:59 |
freemangordon | when I have some spare time I will install leste VM there | 09:00 |
freemangordon | this pal is 4 cores i7 (HT) + 1TB SSD :) | 09:00 |
freemangordon | I guess it will need 20 minutes to build an image | 09:00 |
Wizzup | brb | 09:09 |
freemangordon | the fuck :( "dd: error writing '/home/user/imgbld/image-builder/arm-sdk/tmp/devuan-armhf-build/maemo-leste-1.0-armhf-sunxi-20221014.img': No space left on device" | 09:32 |
Wizzup | where? | 09:33 |
Wizzup | ah | 09:33 |
freemangordon | in my VM :) | 09:33 |
Wizzup | df -h ? | 09:33 |
Wizzup | :) | 09:33 |
freemangordon | it does not lie | 09:33 |
freemangordon | I will have to extend rootfs | 09:33 |
freemangordon | Wizzup: do yuu know if those tablets come with li-po or li-ion batteries and does it make any difference for the replacement? | 10:48 |
freemangordon | *you | 10:48 |
Wizzup | the q8 ones? | 10:50 |
freemangordon | mhm | 10:50 |
Wizzup | no idea sorry | 10:50 |
Wizzup | is your battery dead? | 10:51 |
freemangordon | yep | 10:51 |
Wizzup | too bad | 10:51 |
freemangordon | I will have to find a replacement | 10:51 |
freemangordon | it is some 2800mAh cell | 10:51 |
freemangordon | shouldn't be hard to find | 10:51 |
freemangordon | even mobile phone battery should be ok | 10:51 |
buZz | freemangordon: lipo vs liion should not matter for most chargers | 10:53 |
buZz | both are empty at 2.8 full at 4.2 , and do CCCV charging | 10:53 |
buZz | the biggest difference for a phone, is that lipo slightly expands during its lifespan | 10:53 |
freemangordon | this is a tablet | 10:54 |
freemangordon | and it has lots of space I think | 10:54 |
Wizzup | freemangordon: | 11:10 |
Wizzup | scripts/dtc/pylibfdt/libfdt_wrap.c:149:11: fatal error: Python.h: No such file or directory # include <Python.h> | 11:10 |
Wizzup | more pkgs are necessary I guess | 11:10 |
Wizzup | (python3-dev) | 11:10 |
buZz | freemangordon: then you can go crazy :P | 11:10 |
buZz | my fav is just connecting a 18650 batteryholder on the back with wires :D | 11:11 |
buZz | like this tomtom ; https://i.imgur.com/TgLO7Zs.jpeg | 11:12 |
Wizzup | brainparty works on the lime2 | 11:12 |
Wizzup | everything works great except for exiting the application, that hangs | 11:13 |
Wizzup | lol | 11:13 |
buZz | hehe | 11:13 |
Wizzup | I might try to fix that after the presentation is done just to make it more demo-able | 11:13 |
freemangordon | omg | 11:13 |
Wizzup | since it's a great demo | 11:13 |
Wizzup | freemangordon: it's alive again? | 11:14 |
freemangordon | which one? q8? | 11:16 |
freemangordon | Wizzup: do you know where it puts the kernel? | 11:19 |
Wizzup | freemangordon: the sunxi image? probably /boot | 11:21 |
Wizzup | or, what is it? | 11:21 |
uvos__ | hmm brainparty dosen work on the d4 | 11:29 |
freemangordon | is it ok to replace 2800mAh with 2600mAh? | 11:29 |
uvos__ | i mean it kinda works, but touch is messup sometimes and it segfaults often when starting a new challange | 11:29 |
uvos__ | freemangordon: yes | 11:29 |
freemangordon | ok, thanks | 11:30 |
Wizzup | uvos__: yes, it's resolution related | 11:32 |
Wizzup | uvos__: and maybe also screen orientation | 11:32 |
freemangordon | uvos__: shall I keep the original electronics? | 11:32 |
uvos__ | yes - its probubly nessecary | 11:33 |
uvos__ | often the device will read an eeprom or refuse to charge wihout a thermistor it expects etc | 11:33 |
freemangordon | the new one comes with alectronics as well | 11:33 |
freemangordon | this is cheapo allwinner tablet | 11:33 |
uvos__ | right but probubly not what it expects | 11:33 |
freemangordon | ther are only 2 wires | 11:33 |
uvos__ | then it dosent matter | 11:33 |
freemangordon | ok | 11:33 |
uvos__ | you can keep the bms that comes with the battery in this case | 11:33 |
freemangordon | bms? | 11:34 |
freemangordon | s? | 11:34 |
freemangordon | schematic? | 11:34 |
uvos__ | system | 11:34 |
uvos__ | battery managment system | 11:34 |
freemangordon | ok | 11:34 |
uvos__ | the "BMS" i this case is just a chip that disconnects the battery if it goes above 4.25 or below 2.9 usually | 11:36 |
uvos__ | in multi cell batteries its also resposable for balanceing | 11:36 |
uvos__ | but that dosent apply here | 11:36 |
freemangordon | yeah | 11:36 |
freemangordon | thanks | 11:36 |
buZz | uvos__: hey i had dual screen sorta working on d4 :D | 11:41 |
buZz | with something like > xrandr --fb 1920x1080 --output LVDS1 --mode 1600x900 --scale-from 1920x1080 --output HDMI1 --mode 1920x1080 --scale 1x1 --same-as LVDS1 | 11:41 |
buZz | (it can scale outputs that are mirrored to other resolutions!) | 11:41 |
buZz | but, a issue appeared; the DSI-1 output is rotated :P | 11:42 |
uvos__ | sure that applies a transformation matrix | 11:42 |
uvos__ | i mentioned this works | 11:42 |
uvos__ | however its not accelerated | 11:42 |
buZz | yeah, its cool :) | 11:42 |
uvos__ | shuld be very slow | 11:42 |
buZz | it did work pretty fast, not really noticably slow | 11:42 |
uvos__ | hm ok | 11:43 |
buZz | also, -booting- d4 with HDMI attached gets you a dual video output aswell | 11:43 |
buZz | but something gets seriously confused , and ends up just disabling all video outputs :P | 11:43 |
uvos__ | wierd that its fast enough | 11:43 |
uvos__ | freemangordon: do you know if omapddx accelerates transformation matrixices? | 11:43 |
buZz | uvos__: the matrix would just be x2/4 btw, 960 * 2 = 1920 , 540 * 2 = 1080 | 11:44 |
uvos__ | freemangordon: ie xrandr --outputie --transform | 11:44 |
Wizzup | is pkgmaster.devuan.org slow for anyone else? | 11:44 |
buZz | that might help the speed | 11:45 |
uvos__ | buZz: maybe it optimizes that yeah | 11:45 |
buZz | 12ms ping Wizzup | 11:45 |
uvos__ | buZz: can you try some other matrix? | 11:45 |
Wizzup | buZz: I mean I got 4kB/s after a few seconds | 11:45 |
Wizzup | s/got/get/ | 11:45 |
buZz | isnt pkgmaster. a round robin? | 11:45 |
Wizzup | well I get the same robin :) | 11:46 |
uvos__ | something like xrandr --output DSI-0 --transfrom <whatever non integer scaleing> | 11:46 |
buZz | lets see | 11:46 |
freemangordon | Wizzup: it is extremely slow | 11:46 |
freemangordon | uvos__: no idea | 11:46 |
buZz | uvos__: 'xrandr --output DSI-1 --scale-from 800x600' -> black screen | 11:47 |
freemangordon | uvos__: but copy with resising is accelerated through gpu | 11:47 |
freemangordon | *resizing | 11:48 |
freemangordon | or rather all copies | 11:48 |
uvos__ | buZz: ^^ ok that explains it for trivial matricies | 11:48 |
buZz | ah, cool, that might help to make some 'maemo friendly' videooutput :) | 11:48 |
uvos__ | (--scale-from just applies a simple matrix) | 11:48 |
buZz | yeah | 11:48 |
uvos__ | that also explains why rotation dosent work | 11:49 |
buZz | i think it would be nice to just have a 'enable screen mirror' option somewhere | 11:49 |
uvos__ | if you have a rotated primary and a non rotated secondary | 11:49 |
buZz | i didnt try to rotate myself yet | 11:49 |
uvos__ | x has to rotate the buffer for one output | 11:49 |
buZz | it might just work through a transformation matrix aswell? | 11:49 |
uvos__ | but it cant unless you accelerate non trivial matricies too | 11:49 |
buZz | hmhm | 11:49 |
uvos__ | yes thats how modesetting works | 11:49 |
freemangordon | and is slow as snail | 11:50 |
uvos__ | no | 11:50 |
buZz | rotating is slow? | 11:50 |
freemangordon | yes, see on PP | 11:50 |
uvos__ | the sw fallback is slow | 11:50 |
uvos__ | that works totaly differantly | 11:50 |
buZz | freemangordon: but we're rotating 24/7 on d4 | 11:50 |
uvos__ | buZz: yes | 11:50 |
uvos__ | thats not modesetting | 11:50 |
freemangordon | buZz: but, we use TILER to rotate | 11:50 |
* buZz confused :) | 11:50 | |
buZz | freemangordon: ah | 11:50 |
uvos__ | the sw fallback of modeseting to rotate the display | 11:50 |
freemangordon | buZz: uvos__ was just giving an example | 11:51 |
uvos__ | dosent use the transformation matrixies | 11:51 |
buZz | and i guess TILER doesnt want to do differnent stuff on multiple outputs? | 11:51 |
uvos__ | it uses a custom function to rotate the frame | 11:51 |
freemangordon | it does whatever you tell him to do | 11:51 |
buZz | hmm :) | 11:51 |
freemangordon | the issue is not with tiler but with rendering | 11:51 |
Wizzup | buZz: what about hdmi capture device with the same res as the d4 screen? | 11:51 |
buZz | Wizzup: seems my capture unit doesnt like 960x540 :/ | 11:52 |
buZz | maybe i can get xorg to see 1920x1080 as 2/4x smaller? | 11:52 |
buZz | but , the rotation is still a slight issue then | 11:52 |
buZz | oh yeah, connecting a USB mouse, enabling HDMI as primary, restarting hildon, gets you mousecursor on the LCD (rotated) , and all maemo output to HDMI :P so weird | 11:53 |
Wizzup | that's unfortunate @ capture unix | 11:54 |
uvos__ | 960x540 is a unloved resoultion for some reason | 11:55 |
uvos__ | really theres no good reason any 1080p pannel should not accept it as an input | 11:55 |
uvos__ | but they dont | 11:55 |
uvos__ | at least in edid | 11:55 |
uvos__ | often they do all kinds of resolutions if you force it via modeline | 11:56 |
Wizzup | yeah | 11:58 |
buZz | yeah i'll need to try that, but will have to be later, otherjob today | 11:58 |
Wizzup | :) | 11:58 |
Wizzup | I have brainparty working I think now, without hang upon exit | 11:58 |
Wizzup | so I'm happy | 11:58 |
buZz | w00t | 11:58 |
Wizzup | these was a hang in a pthread join in the audio closing | 11:58 |
Wizzup | I justcommented it since it was just before sdl_quit anyway | 11:58 |
Wizzup | will figure out a proper fix later :p | 11:59 |
buZz | hmm last i tried brainparty did work on d4 btw | 12:00 |
Wizzup | yeah but it's buggy | 12:00 |
uvos__ | it "wroks" | 12:00 |
Wizzup | > 4,553 B/s 24min 46s | 12:00 |
Wizzup | :( | 12:00 |
buZz | yeah just works here, but touchscreen is confused because of smaller window? | 12:00 |
Wizzup | brainparty was never made for anything other than 800x480 | 12:00 |
buZz | ah no, its accurate, just a bit smaller | 12:00 |
buZz | right | 12:00 |
Wizzup | there is a newer version for different screens | 12:00 |
uvos__ | somhow the coordinates get confused | 12:00 |
uvos__ | also it sefaults alot | 12:00 |
Wizzup | but I never tried to reconcile them | 12:00 |
buZz | Wizzup: we'd have to rework all the graphics? | 12:01 |
uvos__ | *segfaults | 12:01 |
Wizzup | the author made newer versions | 12:01 |
buZz | or just stretch perhaps | 12:01 |
buZz | ah ok | 12:01 |
uvos__ | buZz: the version we have is an ainchent maemo one | 12:01 |
uvos__ | theres a desktop linux one that supports more | 12:01 |
uvos__ | thats newer | 12:01 |
buZz | k k | 12:01 |
Wizzup | but the maemo one is nice :D | 12:01 |
uvos__ | on n900 maybe | 12:01 |
Wizzup | on this lime2 tablet as well :) | 12:02 |
uvos__ | it also segfaults on n900 for me btw | 12:02 |
uvos__ | when loading the next game | 12:02 |
Wizzup | just tried it | 12:02 |
uvos__ | ok | 12:02 |
Wizzup | ? | 12:02 |
uvos__ | yeah | 12:02 |
Wizzup | I suspect this might be (more) libsdl stuff | 12:02 |
uvos__ | more like a bug in brainparty that old libs wher hideing | 12:02 |
Wizzup | maybe | 12:04 |
Wizzup | did you attach gdb? | 12:04 |
uvos__ | no, i have to say, brainparty was never a priority before :P | 12:05 |
Wizzup | :D | 12:05 |
Wizzup | I just want to attracts folks to the stand ;) | 12:05 |
Wizzup | attract* | 12:05 |
buZz | also, closing brainparty just worked | 12:06 |
Wizzup | yeah, maybe it's related to no audio on the lime | 12:06 |
Wizzup | not sure | 12:06 |
buZz | ah hmhm | 12:06 |
Wizzup | (the lime2 has no audio port, you have to use the pins unfortunately) | 12:06 |
buZz | maybe hookup a usb soundcard? | 12:07 |
Wizzup | then we'd need headphones or speakers as well | 12:09 |
* buZz hands Wizzup 10kg of amplified speakers | 12:11 | |
buZz | :P | 12:11 |
buZz | which gps application is nicest on d4 to grab maps over gprs and show (and maybe record) gps location? | 12:12 |
buZz | maep? | 12:12 |
Wizzup | I like maep, but others work too | 12:12 |
Wizzup | cloudgps can be 'smooth' | 12:12 |
Wizzup | but maep is my preferred one usually | 12:12 |
Wizzup | there's also modrana which is more featured | 12:13 |
Wizzup | and one uvos uses but it is not hildonized iir | 12:13 |
Wizzup | c | 12:13 |
buZz | navit | 12:13 |
Wizzup | could be yes | 12:13 |
uvos__ | yes navit | 12:13 |
buZz | i got em all installed now, i'll try to record some while in the car today | 12:13 |
uvos__ | navit is the only one of these that works offline | 12:13 |
Wizzup | looks like there is a maemo navit, I remember I looked at this before | 12:13 |
Wizzup | uvos__: does it also do routing? | 12:13 |
buZz | navit doesnt grab maps dynamically at all | 12:13 |
uvos__ | Wizzup: yes | 12:13 |
Wizzup | if you could make a guide some times :) | 12:13 |
Wizzup | I would love that | 12:13 |
buZz | uvos__: really? doesnt navit require some external daemon for routing? | 12:14 |
uvos__ | no | 12:14 |
buZz | oh | 12:14 |
uvos__ | or maybe its included in the devuan package | 12:14 |
uvos__ | it just works | 12:14 |
buZz | nice :) | 12:15 |
buZz | maybe i should quickly put a navit map dump on my d4 | 12:15 |
Wizzup | if it works semi well I'm happy to work on porting the hildon port | 12:15 |
buZz | navit is quite amazing, they also have a whole distro they maintain for those ancient tomtoms | 12:16 |
uvos__ | it works but: its really really slow (even on a very fast pc), the ui is "touch optimized" and intended for in dash navigation but ultra klunky and counter intuative, configuring it via its xml file is a pain | 12:17 |
buZz | https://www.youtube.com/watch?v=bqftmI6es_w <-- | 12:17 |
buZz | > OpenTom fork NavitTom on a 2007 Tomtom One XL new OSM maps | 12:17 |
buZz | slow gets a new definition there ;) takes ages to boot, start up , etc | 12:18 |
uvos__ | the wierd thing with navit | 12:18 |
buZz | but, does work | 12:18 |
uvos__ | is that the map rendering takes constant time | 12:18 |
uvos__ | dosent matter if its on a droid4 or a ryzen9 | 12:18 |
uvos__ | and its allways slow | 12:18 |
uvos__ | its like it dose sleep(1) before rendering a map tile xD | 12:19 |
buZz | :P | 12:19 |
uvos__ | also idk what you would hildonize even | 12:23 |
uvos__ | it uses sdl to render its own ui | 12:23 |
buZz | yeah, like cloudgps (?) | 12:23 |
buZz | which is somehow hildonizing itself, it seems | 12:24 |
uvos__ | theres not mutch here to do except maybe use liblocation instead of gpsd | 12:24 |
uvos__ | buZz: i renders a ui that sorta looks like hildon | 12:24 |
uvos__ | thats all (for cloudgps) | 12:24 |
buZz | ah yes, that was some issue, gpsd doesnt start normally | 12:24 |
buZz | uvos__: oo thats right, navit has 'ui layouts' in xml files | 12:24 |
buZz | maybe we can make a 'hildon-esque' xml layout | 12:24 |
uvos__ | buZz: yes | 12:24 |
uvos__ | buZz: but have a fun time changeing anything | 12:24 |
buZz | :D | 12:24 |
buZz | yay! funtimes are fun | 12:25 |
Wizzup | there is/was gtk | 12:25 |
Wizzup | whic iirc the hildon one uses | 12:25 |
uvos__ | there are also different frontends | 12:25 |
uvos__ | theres a qt one the default sdl one and there used to be a gtk one untill it was removed | 12:25 |
uvos__ | they are universaly terrible | 12:25 |
buZz | :) | 12:26 |
uvos__ | its like some alien decided to do a humanz ui :P | 12:26 |
buZz | navit is a weird bunch :P but they are very helpful on github | 12:27 |
uvos__ | they are clearly aliens that are preparing the invasion by implementing mapping systems of earth to use | 12:28 |
buZz | osm_bbox_2.9,50.7,7.3,53.8.bin 61% 979MB 277.3KB/s 38:24 ETA | 12:47 |
buZz | pfff | 12:47 |
buZz | keeps hanging during transfer a lot | 12:47 |
buZz | really start to think something is wrong on my d4 which 'disk io' | 12:47 |
freemangordon | Wizzup: 5.18 does not boot :( | 12:55 |
freemangordon | what's worse is that there is nothing on the screen | 12:55 |
uvos__ | buZz: iw phy phy0 set txpower fixed 1000000000 | 12:56 |
uvos__ | try that | 12:56 |
uvos__ | for some reason sometimes the wifi fw dosent ramp up tx power on auto | 12:56 |
uvos__ | when its clearly nessecary to not get terrible packet loss | 12:56 |
uvos__ | maybe is also kerel bug instead of fw | 12:57 |
uvos__ | *kernel | 12:57 |
Wizzup | freemangordon: mhm | 13:12 |
freemangordon | I found an old kernel here, trying it now | 13:14 |
freemangordon | no dice :( | 13:14 |
freemangordon | well, that kernel seems to boot at least | 13:16 |
freemangordon | hmm, maybe I was not patient enough | 13:18 |
freemangordon | or mce licked the screen :) | 13:18 |
freemangordon | *locked | 13:18 |
freemangordon | "no space left on the device" :) | 13:30 |
freemangordon | seems 2GB are not enough | 13:30 |
freemangordon | "welcome to leste" :) | 13:30 |
freemangordon | colors are weird and TS does not work | 13:31 |
Wizzup | freemangordon: /etc/expandcard.sh ? | 13:32 |
freemangordon | I did that with gparted | 13:32 |
buZz | does gparted also resize the fs? | 13:32 |
freemangordon | yes | 13:33 |
buZz | nice | 13:33 |
buZz | oh, the SD card is literally just 2GB? :) | 13:33 |
freemangordon | Wizzup: we have the same issue 32bit vs64bit, do you rmember? | 13:33 |
freemangordon | *remember | 13:33 |
freemangordon | when we build 32 bit image on 64bit | 13:33 |
Wizzup | sorry, what issue? | 13:34 |
Wizzup | yes I remember, something with themes | 13:34 |
freemangordon | missing icons/pixmaps etc | 13:34 |
freemangordon | yes | 13:34 |
Wizzup | it was some qemu / glib thing iirc | 13:34 |
freemangordon | yes | 13:34 |
freemangordon | oh, sorry | 13:34 |
freemangordon | it is my fault | 13:34 |
freemangordon | :) | 13:34 |
Wizzup | so you have some life in the tablet? | 13:34 |
freemangordon | my VM is 64 bits | 13:34 |
freemangordon | yes | 13:35 |
Wizzup | cool | 13:35 |
freemangordon | unfortunately with some old kernel | 13:35 |
Wizzup | mhm | 13:35 |
freemangordon | 4.11 | 13:35 |
Wizzup | might be mali even (not lima) | 13:35 |
freemangordon | FB even :) | 13:35 |
Wizzup | right | 13:35 |
freemangordon | will ask on #linux-sunxi later on | 13:35 |
Wizzup | they're on oftc now I think | 13:36 |
freemangordon | sees so | 13:37 |
Wizzup | https://phoenix.maemo.org/job/leste-image-sunxi/6/console kernel is building | 13:39 |
freemangordon | yes. and does not boot :) | 13:39 |
Wizzup | mhm, at least on your tablet | 13:40 |
Wizzup | do you have serial? | 13:40 |
freemangordon | ah, which kernel is that? | 13:40 |
freemangordon | 5.11? | 13:40 |
Wizzup | uh | 13:40 |
Wizzup | yes | 13:40 |
Wizzup | 5.11.y | 13:40 |
freemangordon | do not change it, lets see if it will boot here | 13:40 |
Wizzup | it's still on uboot btw, I was wrong | 13:40 |
Wizzup | let's see if kernel builds though | 13:41 |
freemangordon | it will | 13:41 |
Wizzup | the image-builder's are on hdd now, not on ssd | 13:41 |
Wizzup | maybe I should move them back once I have a bit of time | 13:41 |
buZz | Wizzup: with bcache you could use (a part of) your SSD to cache the HDD | 13:43 |
buZz | at nurdspace i made a bcache on a raid0 over two nvme drives for a raid1 on hdds , gets ~1.5GB/s reads now | 13:43 |
Wizzup | this is a qemu qcow image | 13:44 |
buZz | if its on a blockdevice, you could add bcache | 13:44 |
Wizzup | it's not | 13:44 |
buZz | but i think you'd have to reinitialize it | 13:44 |
buZz | the image file isnt on a blockdevice? | 13:44 |
Wizzup | it's on a btrfs fs | 13:44 |
Wizzup | which consists of 6 block devices | 13:44 |
Wizzup | so please let's not go down the path | 13:45 |
buZz | yeah i think thats still supported | 13:45 |
Wizzup | I'm not interested in bcache | 13:45 |
buZz | ok :) | 13:45 |
Wizzup | buZz or uvos__: how does one go about connecting a bt keyboard | 14:01 |
Wizzup | assuming bt drivers work etc | 14:01 |
buZz | you could apt install blueman | 14:02 |
buZz | and connect it with that | 14:02 |
buZz | but i havent tried keyboards yet, headphones and speakers work though | 14:02 |
buZz | its a gtk gui, i think | 14:02 |
Wizzup | ok | 14:07 |
Wizzup | buZz: it's not obvious how to use it, but maybe I am missing something | 14:09 |
Wizzup | it looks like it just backgrounded | 14:09 |
buZz | oh yeah, its very unobvious indeed | 14:09 |
Wizzup | ah hcitool sees no device | 14:10 |
Wizzup | ok then | 14:10 |
Wizzup | won't bother | 14:10 |
sicelo | iirc hcitool is deprecated:-( | 14:11 |
Wizzup | in any case there's no bt device seen | 14:13 |
buZz | did you remove the blacklisted bt? | 14:14 |
buZz | at least on d4 afaik its still blacklisted | 14:14 |
buZz | not sure | 14:14 |
Wizzup | this is rpi | 14:14 |
buZz | ah | 14:14 |
buZz | not all rpi have bt iirc | 14:15 |
Wizzup | rpi4 does | 14:15 |
Wizzup | p sure | 14:15 |
buZz | but might not be in the image then | 14:15 |
Wizzup | I made the image | 14:15 |
buZz | yeah pi4 should | 14:15 |
buZz | random tutorial installs 'pi-bluetooth' on it | 14:15 |
buZz | https://github.com/RPi-Distro/pi-bluetooth | 14:16 |
uvos__ | buZz: modprobe hci_uart, start bluez, have blueman installed | 14:19 |
Wizzup | on d4/bionic I guess? | 14:19 |
uvos__ | run blueman-manger | 14:19 |
buZz | seems pi4 has firmware needed for bt that isnt in other package? | 14:19 |
uvos__ | search for a device, click on it pair | 14:19 |
uvos__ | etc | 14:19 |
uvos__ | everything works pretty much as expected from that point | 14:20 |
uvos__ | i use it all the time | 14:20 |
uvos__ | Wizzup: yes | 14:20 |
uvos__ | but on other devices it should be the same | 14:20 |
uvos__ | besides the modprobe of the mapphone driver ofc | 14:20 |
buZz | seems pi4 also uses hciuart | 14:20 |
uvos__ | haveing bluez running costs a tone of power tho | 14:21 |
uvos__ | on d4 | 14:21 |
uvos__ | so better exit it right after | 14:21 |
buZz | uvos__: this is probably for the demos at booth, so likely has power available | 14:23 |
Wizzup | yes | 14:23 |
uvos__ | sure im just saying i use it daliy on d4 | 14:24 |
uvos__ | but have to stop bluez | 14:24 |
buZz | alright, maybe some day we can reduce that powerdraw | 14:26 |
Wizzup | starting bluez, that is /etc/init.d/bluetooth start ? | 14:27 |
buZz | yeah | 14:28 |
buZz | should be .. | 14:28 |
buZz | don't think i ever started it differently :P | 14:28 |
Wizzup | freemangordon: oh right I forgot that for sunxi I had also disabled the kernel build by skipping it | 14:33 |
Wizzup | freemangordon: for some reason I had to set roamingallowed to true for my a1 bg sim to work here | 16:32 |
Wizzup | on leste | 16:32 |
Wizzup | the ui would blink forever | 16:32 |
Wizzup | the mdbus command failed immediately with 'gprs not attached' | 16:32 |
Wizzup | setting roamingallowed to true made it work | 16:32 |
freemangordon | weird | 16:34 |
Wizzup | yup | 16:38 |
Wizzup | got stickers :p | 16:55 |
freemangordon | Wizzup: Oct 14 15:51:56 localhost kernel: [ 0.126953] [drm] Initialized lima 1.1.0 20191231 for 1c40000.gpu on minor 0 | 17:55 |
freemangordon | this is with 5.15 | 17:56 |
Wizzup | freemangordon: that is good, right? | 18:01 |
freemangordon | mhm | 18:01 |
freemangordon | I have to play a bit because of the panel, but that should be trivial | 18:01 |
Wizzup | ok | 18:02 |
Wizzup | great :) | 18:02 |
freemangordon | TS is tricky too | 18:02 |
freemangordon | anyway, going on a ride, ttyl | 18:02 |
Wizzup | ttyl | 18:04 |
buZz | meh :( | 20:06 |
buZz | recorded a 6hr track on Maep , went to 'export' , selected location , (no save button???) tapped outside of panel, Maep closed/crashed and zero recorded track | 20:06 |
buZz | :( | 20:06 |
rafael2k | what is this version "+2m7"? | 20:10 |
rafael2k | should I use in new packages too? | 20:10 |
buZz | m7 = maemo 7? | 20:11 |
buZz | (aka leste) | 20:11 |
rafael2k | ah | 20:12 |
rafael2k | right | 20:12 |
rafael2k | : ) | 20:12 |
rafael2k | Unpacking connui-conndlgs-internet (2.79.7+2m7) over (2.79.6+2m7) ... | 20:15 |
rafael2k | dpkg: dependency problems prevent processing triggers for gconf2: | 20:15 |
rafael2k | gconf2 depends on default-dbus-session-bus | dbus-session-bus; however: | 20:15 |
rafael2k | Package default-dbus-session-bus is not installed. | 20:15 |
rafael2k | Package dbus-session-bus is not installed. | 20:15 |
rafael2k | Package dbus-x11 which provides dbus-session-bus is not configured yet. | 20:15 |
rafael2k | apt-get upgrade just broke here | 20:15 |
rafael2k | it gets in a look | 20:16 |
rafael2k | *loop | 20:17 |
rafael2k | :( | 20:17 |
buZz | tried apt-get -f install ? | 20:17 |
buZz | just without a pkg behind it | 20:17 |
rafael2k | yes | 20:17 |
buZz | :( | 20:17 |
rafael2k | crazy loop | 20:17 |
buZz | dpkg-reconfigure dbus-x11 ? | 20:17 |
rafael2k | never saw this before in 12 yrs using debian | 20:17 |
rafael2k | 12 no, much more | 20:18 |
rafael2k | lemme try to fix | 20:18 |
rafael2k | dpkg: dependency problems prevent processing triggers for gconf2: | 20:18 |
rafael2k | gconf2 depends on default-dbus-session-bus | dbus-session-bus; however: | 20:18 |
rafael2k | Package default-dbus-session-bus is not installed. | 20:18 |
rafael2k | Package dbus-session-bus is not installed. | 20:18 |
rafael2k | Package dbus-x11 which provides dbus-session-bus is not configured yet. | 20:18 |
rafael2k | dpkg-reconfigure dbus-x11 | 20:19 |
rafael2k | dbus-x11 is broken or not fully installed | 20:19 |
buZz | :( | 20:20 |
buZz | was this just a update on leste today? | 20:20 |
rafael2k | I dunno | 20:22 |
rafael2k | I did not do a apt-get upgrade in some days... | 20:23 |
rafael2k | I did not manage to unbreak yet | 20:23 |
rafael2k | uff | 20:26 |
rafael2k | unbroke | 20:26 |
buZz | :( | 20:26 |
rafael2k | force-removed dbus-x11 and installed it again | 20:26 |
rafael2k | : ) | 20:27 |
buZz | fixed? | 20:27 |
rafael2k | yeap | 20:27 |
rafael2k | good ol' debian... it seems it is not unix... but it is | 20:27 |
buZz | ooo new xorg-video-omap again? does it implement Vulkan now? :D | 20:29 |
rafael2k | wow | 20:30 |
rafael2k | I get some crazy shit | 20:30 |
rafael2k | iphb? | 20:30 |
buZz | (highly doubt it'll be vulkan) | 20:30 |
buZz | what kind of crazy shit? | 20:30 |
rafael2k | Your kernel headers for kernel 5.15.7-3-MANJARO-ARM cannot be found at | 20:30 |
buZz | ... MANJARO? :D | 20:30 |
buZz | what device is this? | 20:31 |
rafael2k | Your kernel headers for kernel 5.9.0-rc7-00376-g2e49b58e9b10 cannot be found at | 20:31 |
rafael2k | Maemo-West | 20:31 |
rafael2k | Wild Wild West | 20:31 |
buZz | maemo -west- ? :D | 20:31 |
buZz | lol | 20:31 |
buZz | weird image you're running rafael2k :D | 20:32 |
rafael2k | Your kernel headers for kernel 5.10-sunxi64 cannot be found at | 20:32 |
rafael2k | just apt-get'd dist-upgrade !! | 20:32 |
buZz | what device are you using rafael2k ? | 20:32 |
rafael2k | the same image from 2 years ago I never changed | 20:32 |
rafael2k | PP | 20:32 |
buZz | ok .. well i did see Wizzup do some sunxi stuff today, maybe some packages got updated incompatible? | 20:33 |
buZz | i'll try updating the PP here | 20:33 |
buZz | btw i never do dist-upgrade | 20:34 |
buZz | only on upgrading distributions :) | 20:34 |
buZz | i'll try dist-upgrade now, yeah , nothing | 20:35 |
buZz | 0 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. | 20:35 |
rafael2k | hum | 20:37 |
rafael2k | do you use -devel? | 20:37 |
rafael2k | iphb-dkms | 20:38 |
rafael2k | try this ^ | 20:38 |
buZz | how could you do -dkms if our kernels dont have source? | 20:39 |
buZz | but i do have that installed | 20:39 |
buZz | iphb-dkms/testing,now 1.2+2m7.2 all [installed,automatic] | 20:39 |
rafael2k | : ) | 20:42 |
rafael2k | so all good | 20:42 |
buZz | oh i guess just -headers is enough | 20:44 |
rafael2k | true | 20:44 |
rafael2k | interesting I have not asked to install this package | 20:45 |
rafael2k | it was a dependency of something else | 20:45 |
buZz | yeah might be pulled in by a metapackage i guess | 20:46 |
rafael2k | uhum | 20:47 |
rafael2k | me no like | 20:47 |
rafael2k | :P | 20:47 |
rafael2k | hildon-connectivity-something | 20:47 |
buZz | afaiui iphb can help us reduce powerusage a lot on some applications | 20:48 |
rafael2k | we could at least put the linux-headers as depency | 20:48 |
rafael2k | dependency* | 20:48 |
rafael2k | right, first time I hear about it | 20:49 |
buZz | me too (when i saw it suddenly in this channel) | 20:49 |
buZz | i think its from maemo too | 20:49 |
rafael2k | cool | 20:53 |
norayr | they say pocophone and plus6 (with same soc) work with mainline very well. does anyone have one of those to support with maemo? | 21:43 |
Wizzup | I think maybe parazyd tried for the pocophone | 21:45 |
norayr | eh. | 21:47 |
norayr | it is the fastest device which supports mainline as i understand. | 21:47 |
sicelo | yes, lots of devs working on that SoC, so progress is quite good | 21:49 |
Wizzup | what I mean is that arm-sdk supports pocophone | 21:51 |
uvos | pocophone wored | 21:57 |
uvos | *worked | 21:57 |
uvos | and had an image even | 21:57 |
uvos | but it hasend been tested in quite some time | 21:57 |
uvos | but yeah its an interesting non-mapphone to support i have it on a short list next to xt1602 and friends to try | 21:58 |
uvos | oh look related article on the bbc news main page: https://www.bbc.com/news/science-environment-63245150 | 22:01 |
freemangordon | buZz: rafael2k: https://github.com/maemo-leste/iphb-dkms/blob/master/debian/control#L10 | 23:01 |
buZz | yeah nice, i didnt have any trouble with iphb-dkms though | 23:03 |
buZz | rafael2k had issues around dbus-x11 i hink? | 23:03 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!