[03:10] <RAOF> Woot! Now with fewer memory leaks and with fewer crashes on display hotplug!
[03:30] <RAOF> Wait, what? Now X11 captures the wrong pixel format?
[07:33] <Saviq> Mornang o/
[08:08] <RAOF> Today:... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/69bfba9432c340eee28e634f79c3b97706e7b910>)
[08:10] <RAOF> Additionally: pondered over the weekend about making cross-GPU scanout less of a "construct an EGL context, import as EGLImage, render to a fullscreen rectangle texturing from that image" fandango. Proposed a thing in !BdPaVRYnmXhCMtKNua:matrix.org that would make it much nicer.
[09:53] <Saviq> Meh. Would be cool to have ARM packages from CI...
[10:07] <alan_g[m]> Occasionally, yes. Was tedious building on RPi400, but we don't do that often...
[10:08] <Saviq> I have the multiarch build-snap action that works on some snaps... but not others, maybe remote-build is The Way(TM).
[15:45] <Saviq> Is there an easy way to reset VT state after kmscube do you know? I can toggle VTs and it comes back, but if I can kick something else, that would be preferable :)
[15:45] <alan_g[m]> Probably, but not a way that I know of.
[15:46] <Saviq> Will leave this for Chris when he wakes up, then :)
[15:47] <alan_g[m]> Is it as simple as reset? ?
[15:48] <Saviq> No :)
[15:57] <alan_g[m]> Then you're stuck with something like: vt=$(fgconsole);chvt $((vt+1));chvt $vt
[15:58] <Saviq> That's my last resort, yeah
[16:05] <Saviq> Today:... (full message at <https://libera.ems.host/_matrix/media/v3/download/libera.chat/d53c89104dff438aa7c9a36610c3fd28a0cc3b25>)
[17:28] <alan_g[m]> Today:
[17:28] <alan_g[m]> * Re-reviewing mariogrip's PR
[17:28] <alan_g[m]> * Reviewing a couple of Saviq's PRs
[17:28] <alan_g[m]> * Pulse review/planning