02:22 <Wizzup> yes, backend should not be in settings

02:45 <Wizzup> with tp we'll have many backends

02:45 <Wizzup> uvos__: do you think it's worth adding the mz617 hacks to our experimental or devel kernel?

06:08 <uvos__> Wizzup: thatswhz i

06:08 <uvos__> upps

06:08 <uvos__> Wizzup: thats why i rebased the kernel on 6.6

06:09 <uvos__> but its currently no use untill the omap-dai is fixed

06:18 <Wizzup> ok

07:05 <nela> someone was tracking down the IR TX issue on N900 and found the timing issues, did we ever get to the end of it?

07:06 <nela> also I now acquired a second worse condition device for debugging purposes and I will solder a UART header to it shortly

07:26 <freemangordon> nela: see https://patchwork.kernel.org/project/linux-media/patch/1696501739-29861-1-git-send-email-ivo.g.dimitrov.75@gmail.com/ and https://patchwork.kernel.org/project/linux-media/patch/1560b474f7d426bc77100665c14c3a29c3af3e75.1697193646.git.sean@mess.org/

08:30 <Guest224> dear developers Leste wiki has error at PP: Keyboard, Yes, When a PP keyboard is connected, it works out-of-the-box

08:32 <Guest224> it should be this way: Keyboard, WIP, When a PP keyboard is connected, all exept third level of first row

08:33 <Guest224> It had not be working right way in any stable version.

08:35 <Guest224> you find problem quite fast when writing with PP keyboard: setxkbmap -query | grep model

08:40 <Guest224> solution is in that link I put yesterday, but it should merge to next version of Leste.

08:42 <Guest224> or is it coming in next version already?

14:40 <uvos> freemangordon: sure that fixes the issue with ir-tx

14:40 <uvos> but is the underlying issue of the delay taking longer than expected still there

14:41 <uvos> if so i suspect it might be the cause for bluetooth audio being semi-broken on d4

14:42 <uvos> (it faces fairly constant ~100ms clicks where something is unable to keep the stream going, even when pm states are disabled)

14:42 <uvos> us

17:13 <sicelo> arno11 ... i did test the transitions.ini. all is well indeed. just one nitpick - when in the application list, not blanking/dimming the wallpaper makes things seem confusing/cluttered

17:13 <arno11> sicelo: yes, need a bit of work for this

17:15 <arno11> sicelo: i have another transitions.ini file wich is faster (but still the confusing app list)

17:15 <arno11> *btw

17:16 <arno11> uvos: sicelo: https://ibb.co/MpBhkRR

17:16 <arno11> this is the best result i have ATM

17:17 <arno11> still using hbox (vbox is tricky to use with the keypad)

17:18 <arno11> but user friendly even with big fingers IMO

17:18 <sicelo> this looks good! i think you can bring Recent to the bottom as well, and have three buttons on that bottom row

17:18 <sicelo> but i don't see 'Cancel' ;p

17:19 <arno11> i removed it lol

17:19 <arno11> but i can add it again

17:20 <arno11> (if uvos is hangry)

17:20 <sicelo> i don't think it'll be approved :p

17:20 <arno11> *angry

17:20 <arno11> yes indeed :p

17:22 <arno11> maybe i can add 'recent' at the bottom row and add 'cancel' on the top one ?

17:22 <arno11> *instead of 'recent'

17:31 <sicelo> i'm quite sure they want Cancel to be at the bottom ... presumably to close the application easily, even in portrait mode

17:33 <arno11> ok

17:34 <uvos> pretty mutch exactly correct

17:34 <arno11> so call, contacts, cancel

17:34 <uvos> i also really dont like this layout for another reason: its too different to the portrait one

17:34 <uvos> the relative possitioning of buttons should stay the as mutch the same as possible

17:34 <uvos> to avoid too mutch user suprise

17:35 <arno11> agreed but dealing with vbox columns is tricky

17:35 <arno11> *with the keypad

17:36 <uvos> hmm

17:36 <uvos> the keypad is one widget anyhow

17:36 <uvos> so just have a everything in a vbox inside a hbox that contains just the vbox in portait

17:37 <uvos> and remove the keypad from the vbox and add it to the hbox where the vbox is also in when switching to landscape

17:37 <uvos> or rather when switching to a window that too small for the portrait layout

17:39 <arno11> vbox inside a hbox is what i'm trying to use ATM with no success

17:39 <arno11> question of time :p

17:39 <uvos> ok, not sure what issues would crop up there

17:39 <uvos> but its been a whlile since i wrote a gtk2 ui layout

17:44 <uvos> oh btw

17:44 <uvos> i would recommend "scrot" to take screenshots over ssh

17:44 <uvos> mutch less work than makeing a photo

17:51 <arno11> thx for the tip

18:15 <arno11> question of POV but i think the last example is not so different from the portrait mode: in every cases we have to add buttons on the right/left/bottom/top anyway

19:20 <freemangordon> uvos: yes, and unfortunately I have no idea what the reason could be. However, seems to be process priority/scheduler related, like, on omap4 with proper priority latency is low. Keep in mind the shortest measurable period is ~30us, because system timer is running @ 32kHz, so maybe scheduler has no option when we have 3-4 high-priority tasks but to delay each task re-activation with couple of 30us periods

20:05 <uvos> freemangordon: if you think its related to scheduler behavior maybe we should try a different one than cfs, like pds or bmq or the like, those are mostly more optimized for interactive systems anyhow

20:32 <freemangordon> uvos: if you give me some hints I can try

20:33 <freemangordon> but yeah, seems scheduler related