04:03 <NaviTheFairy> Hey, I seem to have gotten locked out of ssh access by fat fingering my password a time or two. Is there a way to disable this fail2ban like behavior in the gui?

04:46 <mack> NaviTheFairy: I've never heard of such a feature in LE. Something else has likely changed. There are some good troubleshooting steps in this forum thread, though -- like how to see a good log and how to login textmode. https://forum.libreelec.tv/thread/25004-ssh-not-allowing-login-nightly-build-on-rpi4-over-pxe/

09:03 <newuser566> kodi works finally, thx for your work

09:48 <adema> jernej, I asked around where the most recent patches were and if there were some upstreaming in progress, I was here that if wasn't on the roadmap, so I am sorry if I overstep any boundaries. I thought this work was abandoned and that people kept rebasing the old patches.

09:51 <adema> I can definitely not pursue this though, I haven't wrote the code, you guys did, I just mostly refactored it. I am just not sure why not try to merge experimental support that can be improved by the community? I feel like it would gain in quality if more people would use it and review it?

16:33 <LEBot> [slack] <jernej> adema: did you asked here? If so, I missed your post

16:34 <adema> I asked some bits here and some on ffmpeg-devel channel

16:34 <LEBot> [slack] <jernej> anyway, we certainly need someone like you to upstream the patches, but before that, some work needs to be done

16:35 <adema> but no worries, I won't be pushing forward my attempt for the time being

16:35 <LEBot> [slack] <jernej> if you're still interesting in this work, please take a look at RPi version

16:35 <adema> I really just thought it would be genuily better to have *something* in FFmpeg that we can iterate over

16:36 <LEBot> [slack] <jernej> RPi uses improved design, which should lower latency

16:37 <LEBot> [slack] <jernej> so that should be combined with stable codecs, tested and then sent to ffmpeg ML

16:39 <adema> I honestly don't much knowledge about the inner working for v4l2 request APIs, I just read a lot of code and decided to clean / refactor some bits and send a v2 of the original patches. You can probably do a better job than me upstreaming it. I just thought no one was willing to do it so I volountered :p

16:39 <LEBot> [slack] <jernej> yeah, but the thing is that I don't have much time for such work

16:41 <LEBot> [slack] <jernej> this is also the reason why it wasn't pushed to ffmpeg yet

16:42 <adema> No worries, I fully understand :)

16:43 <adema> If I were to go forward how would I test all of this? I just run ffmpeg with -drm and -drm_device option without a rawvideo codec and it seems to decode fine but do you have a test harness somewhere for this? Or just instruction on how to verify it works fine

16:43 <LEBot> [slack] <jernej> in other words, it would be great if you learn all the details and the benefit would be twofold - upstream ffmpeg support and another dev with request api experience :)

16:44 <LEBot> [slack] <jernej> don't you have any RK or AW board?

16:44 <adema> I have a RK3328 board

16:44 <LEBot> [slack] <jernej> well, that's good for dev purposes

16:44 <adema> but you cannot use hwaccels with ffplay right?

16:44 <LEBot> [slack] <jernej> no, but it can be tested via ffmpeg

16:45 <LEBot> [slack] <jernej> you can then even run conformancy test via fluster

16:45 <adema> right, that's the bit I need to know more about

16:46 <adema> This fluster I presume https://github.com/fluendo/fluster ?

16:46 <LEBot> [slack] <jernej> yes

16:46 <LEBot> [slack] <jernej> you can join #linux-media on oftc

16:47 <LEBot> [slack] <jernej> ndufresne is another dev, which cares about request api

16:47 <LEBot> [slack] <jernej> but he is gstreamer developer

16:47 <LEBot> [slack] <jernej> we work in tandem to test different userspace and kernel drivers

16:47 <adema> Oh great

16:48 <LEBot> [slack] <jernej> so in general, you want both, ffmpeg and gstreamer packages with request api support

16:48 <LEBot> [slack] <jernej> and run fluster with both and compare results

16:48 <adema> gstreamer has a pretty good support for the request API already I believe

16:49 <LEBot> [slack] <jernej> yes, in 1.20 is pretty solid

16:50 <LEBot> [slack] <jernej> well, it still has rough edges in some cases, but it's already usable

16:52 <shibboleth> chewitt: ping

16:53 <shibboleth> <shibboleth> i've noticed that 10/12 (30/36)-bit playback is now possible on rpi4, is it possible to instruct the pi to also output the gui in 10/12 to avoid the seconds of delay when switching for playback?

16:53 <shibboleth> <LEBot> [slack] <chewitt> shibboleth not possible to preset higher but depths. Also not desirable for GUI speed reasons

16:53 <shibboleth> yes, that was my question. maybe i misunderstood your answer, but the depth is all i'm trying to adjust

16:54 <shibboleth> i remember back when using a debian vm with a passthrough amd gpu for kodi, ever since the radeon 6950 i had in what... 2011? the UI was output as 10/30-bit

16:55 <shibboleth> now, this wasn't possible until very recently on the pi, but outputting the UI as 10/30 or 12/36 by default will save me a few seconds when switching to 10/12 iptv sources

16:56 <shibboleth> surprised my plasma from 2008 can do 12-bit though. but that's what you get for shelling out 500=USD in 2008-money for a Pioneer :)

16:56 <shibboleth> 5k usd even

16:57 <LEBot> [slack] <jernej> adema: example ffmpeg command line for request api: ffmpeg -loglevel debug -hwaccel drm -hwaccel_output_format drm_prime -i video.mkv -f null -

17:24 <ndufresne> LEBot: hi :-D

17:25 <LEBot> [slack] <Nicolas Dufresne> Nice, didn't know we could also chat here but from slack

18:57 <Successus> Hi, could you please tell me, can I run LibreELEC using VirtualBox or Qemu? Should I use the ova or img file? Thank you in advance

18:59 <colo> is gpu_mem=76 the recommended setting for a RPi4? (2GB model, so plenty of RAM to spare)

19:03 <LEBot> [slack] <CvH> last time (a year ago) vbox was lacking a gpu driver

19:04 <Successus> I found this, I will see if I can follow the guides: https://forum.libreelec.tv/thread/20970-virtual-libreelec/ https://libreelec.tv/virtual/

19:04 <LEBot> [slack] <CvH> qemu should work, ova->qcow2 is also possible

19:04 <Successus> I see

19:04 <Successus> I need to extract something from ova, or ova directly?

19:05 <Successus> It said 2. Extract the LibreELEC-Generic.x86_64-9.2.0.vmdk 3. Convert the vmdk to qcow2

19:05 <LEBot> [slack] <CvH> qemu-img convert -O qcow2 some.ova output.qcow2

19:05 <Successus> I see, thank you

19:05 <LEBot> [slack] <CvH> not tried it with LE but this works in general

19:07 <Successus> what is LE?