*** ChanServ sets mode: +v T4 | 05:33 | |
*** phdeswer__ is now known as phdeswer | 07:11 | |
spiiroin | review appreciated: tweak unblanking on autosleep devices | 08:41 |
---|---|---|
spiiroin | https://github.com/mer-hybris/mce-plugin-libhybris/pull/30 | 08:41 |
spiiroin | https://git.merproject.org/mer-core/mce/merge_requests/91 | 08:41 |
*** rubdos_ is now known as rubdos | 09:15 | |
kimmoli | simplest stepcounter qml demo, works direclty with gistpud preview https://gist.github.com/kimmoli/d2910acd355cf9214ea6aa8435a1e616 | 16:54 |
kimmoli | requires stpcntrd daemon and stepcounter sensor support from device | 16:54 |
*** ChanServ sets mode: +v T4 | 17:11 | |
Mister_Magister | ahhh its perfert time for recompiling my kernel :P | 17:28 |
T4 | <DylanVanAssche> @kimmoli [requires stpcntrd daemon and stepcounter senso …], Nice! | 17:37 |
mal | Mister_Magister: did you check all of your devices whether they need the oom fix, should be easy to check from kernel | 17:45 |
Mister_Magister | nah i didnt have time yet | 17:45 |
Mister_Magister | im just fixing one device | 17:45 |
Mister_Magister | mal: any idea for gps? | 17:46 |
mal | I'm not very good at figuring out those API mismatch issues | 17:47 |
Mister_Magister | sure | 17:47 |
mal | I have never needed shim libs on my devices | 17:47 |
Mister_Magister | sure i just thought you may have idea how can i add them to geoclue | 17:49 |
Mister_Magister | as far as i know it doesnt have systemd service | 17:49 |
Mister_Magister | yepp oom error seems gone | 17:49 |
mal | not sure how the shim loading in libhybris should work | 17:50 |
Mister_Magister | im setting evironment variable | 17:51 |
Mister_Magister | well it wasnt merged and krnlyng didnt answer so i dunno | 17:52 |
krnlyng | Mister_Magister, i didn't see any changes to the patch since i last commented | 17:53 |
Mister_Magister | krnlyng: no changes was requested you were talking how it should be implemented | 17:54 |
Mister_Magister | i dont quite understand that tho | 17:54 |
Mister_Magister | mal: except kinzie (the one i just patched), ghost (the one we discovered problem on yesterday) and huawei none seems to have it. | 17:56 |
Mister_Magister | krnlyng: what am i supposeed to do with that patch | 17:58 |
Mister_Magister | mal: lol dunno if its placebo or something but screen seems to work a lot smoother x | 17:59 |
Mister_Magister | xd | 17:59 |
Mister_Magister | mal: ye i started 19 apps and it seems to be a lot lot better now | 18:01 |
Mister_Magister | a lot lot better. i have heavy apps like browser camera and puremaps and other apps 21 actually opened apps and ram usage is only 75% | 18:02 |
* Mister_Magister has urge to open more | 18:03 | |
Thaodan | does sfos uses libinput? | 18:03 |
Mister_Magister | iirc yes | 18:04 |
Thaodan | despite having it not installed? | 18:04 |
* Mister_Magister slaps roof of moto x force this bad boi can fit so many running apps in it | 18:05 | |
krnlyng | Mister_Magister, the patch is kinda broken | 18:06 |
Mister_Magister | krnlyng: wdym | 18:06 |
NeKit | Mister_Magister, also needed it for hermes | 18:06 |
NeKit | (your patch) | 18:06 |
krnlyng | Mister_Magister, as i've said before, the LD_SHIM_LIBS are supposed to be set only once and the same for all processes | 18:07 |
Mister_Magister | thats true | 18:07 |
Mister_Magister | and thats why i actually cant get geoclue to have shims | 18:08 |
krnlyng | i mean your method works but in android they didn't do it this way because of security concerns. | 18:08 |
krnlyng | Mister_Magister, what do you mean? | 18:09 |
Mister_Magister | krnlyng: geoclue is broken on my phone because one lib that it loads needs shim | 18:10 |
krnlyng | Mister_Magister, but with the patch it works? | 18:10 |
Mister_Magister | no it doesnt | 18:10 |
Mister_Magister | thats what im talking about | 18:11 |
Mister_Magister | geoclue has no process i could give variable for loading shims | 18:11 |
Mister_Magister | i need to set shims for all processes | 18:11 |
Mister_Magister | mal: dunno what happened maybe some process wasnt functioning properly or smth but after reverting that patch its seriously smoother (or its placebo) | 18:12 |
mal | it shouldn't directly affect that | 18:13 |
Mister_Magister | mal: yee im aware of that | 18:13 |
Mister_Magister | maybe its just placebo as always ¯\_(ツ)_/¯ | 18:14 |
Mister_Magister | krnlyng: so i have no idea how can i fix this patch so it will just be there forever not merged? | 18:22 |
krnlyng | Mister_Magister, if it doesn't work properly it doesn't make sense to merge it, if you change it how i proposed, i'll consider it. it's not perfect but better than nothing | 18:24 |
Mister_Magister | krnlyng: i dont know what should i change or how | 18:25 |
Mister_Magister | i didnt get any of what you were saying there | 18:25 |
Mister_Magister | im too stupid sorry | 18:26 |
krnlyng | Mister_Magister, all i am asking is to replicate this with the shims: https://github.com/libhybris/libhybris/pull/377/commits/c214b8f59c57d147eda23082928122c8d7cfc2b9#diff-cf1210eb0ad9879f7d812c63baf11fb4R4643 | 18:27 |
krnlyng | then it would also work with geoclue... | 18:27 |
Mister_Magister | krnlyng: oh that looks like doable | 18:54 |
Mister_Magister | krnlyng: but where do you keep these default paths? | 18:54 |
Mister_Magister | krnlyng: oh its on compile time so it has to be set in .spec file or something or some build config | 18:55 |
mal | Mister_Magister: https://github.com/libhybris/libhybris/blob/07b547e90db625685050bdfd00c92ccafc64aa09/hybris/configure.ac#L140 | 18:56 |
Mister_Magister | mal: i see serious problem here as shims library path is higly depend on device | 18:57 |
Mister_Magister | it cant be like that | 18:57 |
Mister_Magister | not possible | 18:57 |
Mister_Magister | krnlyng: thats not possible | 18:57 |
Thaodan | should mce handle the camera button? | 18:58 |
Mister_Magister | ye why not | 18:58 |
mal | Thaodan: yes, or something like that, why? | 18:59 |
Thaodan | because of KEY_CAMERA_* | 18:59 |
mal | it should be used automatically | 19:00 |
mal | assuming you have correct events | 19:00 |
mal | install mce-tools and check evdev_trace -i to see what you have and then you can also see the signals using evdev_trace -t | 19:01 |
Thaodan | thats what I did but I found nothing about in mce | 19:01 |
mal | what is the problem? | 19:02 |
mal | do you see the events when you press the camera button? which exact events? | 19:02 |
krnlyng | Mister_Magister, why wouldn't it be possible? that variable must be generated *somewhere* during build | 19:02 |
Mister_Magister | krnlyng: but if we put that in configure that wont work it needs to be in some kind of config or something as its device depend you cant set default path for all devices | 19:03 |
krnlyng | that's why you would detect it during build | 19:04 |
Thaodan | mal: KEY_CAMERA_FOCUS KEY_CAMERA_SNAPSHOT | 19:04 |
Thaodan | when pressing the camera button | 19:04 |
mal | Thaodan: so what is the problem? | 19:04 |
Thaodan | it should open the camera app | 19:05 |
mal | no it shouldn't | 19:05 |
Mister_Magister | krnlyng: what do you mean | 19:05 |
mal | Thaodan: actually it does if you long press the button | 19:05 |
mal | didn't remember that | 19:05 |
Thaodan | mal: ah thanks. I don't remember the android behaviour. I thought it was a simple pres | 19:06 |
Thaodan | s | 19:06 |
kimmoli | this is no Android. .. | 19:07 |
kimmoli | you're looking for | 19:07 |
* Mister_Magister is sharing sfos religion on zenfone telegram group | 19:08 | |
Thaodan | doing that Qt also helps | 19:09 |
Thaodan | *that with | 19:09 |
T4 | xdasu1 was added by: xdasu1 | 20:39 |
T4 | <xdasu1> hi | 20:39 |
T4 | <xdasu1> bot | 20:39 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!