<linkert>
So I'm trying to build rivercarro but having an issue building. On line 8 of ./build.zig it wants to, I guess, import another build.zig under "./deps/zig-wayland/build.zig" - but as the error correctly points out, there is no such file to be found.
<linkert>
Zig version: 0.11.0-dev.1605+abc9530a8
<ifreund>
linkert: pretty sure it requires zig 0.10
<ifreund>
you also might be missing git submodules
<linkert>
O
<linkert>
O
<novakane>
yep sounds like ifreund is right, init the submodules and use zig 0.10
<linkert>
Cool, progress made. Now I get: "Package wayland-scanner was not found in the pkg-config search path.". Sorry this is new ground for me. I tried simply installing the wayland-scanner++ package from the void repos which seems unrelated as its about c++ proxy something and no file called "wayland-scanner.pc" was populated on my system.
<ifreund>
xlocate wayland-scanner
<linkert>
Ah, awesome. wayland-devel, wayland-protocols needed now I get to enjoy rivercarro :) Cheers!
linkert has quit [Remote host closed the connection]
linkert has joined #river
<tleydxdy[m]>
woo, finally got the crash again, again it's the view.output pointer that is messsed up
<tleydxdy[m]>
after my system has been up for over 48h, so definitely trending towards something like use-after-free
<tleydxdy[m]>
I suspect the reason it's always a xwayland window that triggers it is because it have hide to tray, so when the view is mapped again the output was gone already? something like that
<tleydxdy[m]>
will try looking into that direction
<ifreund>
any new new logs?
linkert has quit [Ping timeout: 260 seconds]
<tleydxdy[m]>
nope
<tleydxdy[m]>
even less logs this time
<tleydxdy[m]>
I tried to print view.output at lots of places but they didn't catch it
<ifreund>
the scene graph branch is coming along pretty nicely by the way, and there's a decent chance that it'll replace whatever this bug is with some new ones
<ifreund>
currently working on the big tricky change which is the transaction system, then it'll pretty much just be cleanup till it's merged
<tleydxdy[m]>
I hope 😄
waleee has quit [Ping timeout: 246 seconds]
<ifreund>
I'm rewriting all the core data structures, so it's pretty likely I think :D
<tleydxdy[m]>
but looking at map code right now, it doesn't really check output is valid, and init doesn't seem to listen to output destroy
<tleydxdy[m]>
so might just be it
waleee has joined #river
<ifreund>
> init doesn't seem to listen to output destroy
<tleydxdy[m]>
okay I see where views gets transfered to another output when output destroy
<ifreund>
what do you mean by this?
<tleydxdy[m]>
not that simple I guess
<ifreund>
this is code that's going to be totally different and much better structured on my branch fwiw
<ifreund>
i.e. view.output is gone
<tleydxdy[m]>
na, I'm just checking the current code
jao has joined #river
<ifreund>
yeah, I'm just agreeing that the current code is pretty poor
alexherbo2 has quit [Remote host closed the connection]
zdykstra has quit [Read error: Connection reset by peer]