04:13 <toxic_fume> I have returned, and I checked the channel logs for the answer. Is there a preferred place to upload logs like that? I haven't done stuff like this a whole lot

04:53 <bwbuhse> toxic_fume a pastebin service like 0x0 is usually easiest

07:40 <toxic_fume> Sorry for the long gaps, and that I keep disappearing, but know that I'm determined to figure out this issue and will check channel logs for answers, so those answers will not go to waste. Also, thank you for pointing me to 0x0, it seems like a gem. Here is a full log: https://0x0.st/XFWr.txt

07:40 <toxic_fume> I'm on Arch Linux, running an AMD Ryzen 7 5800U with integrated graphics on the Mesa drivers, on a laptop and using only a single monitor. To repeat my problem so all of the info is accessible in one place, when I boot River, it comes up with a kinda teal background, and the mouse cursor displays and moves properly. Keybinds like brightness,

07:40 <toxic_fume> volume, and exiting River work, but if I try to launch any applications, they seem to simply not launch. Examining from a different tty shows wofi not running when I try to launch it, so it seems like the applications actually aren't launching, rather than just having trouble displaying. I generally try running River in a second tty while my first

07:40 <toxic_fume> is running an X11 session, but I've tried doing it without that, and saw no noticeable changes to the logs. Sway, which is also based on wlroots, seems to run fine, as does Hyprland, which isn't.

07:57 <ifreund> toxic_fume: this sounds like a configuration issue, your key bindings don't seem to be working as you intended

07:58 <ifreund> I suggest running a terminal emulator directly from your init script (e.g. `foot &`)

07:59 <ifreund> you can also start programs or run riverctl commands from outside of river by setting WAYLAND_DISPLAY=wayland-1 or whatever number river uses

08:48 <tiosgz> toxic_fume: there is "error(command): Return: unknown command" in the logs, so you most likely have a typo/error in your init, related to how you specify the key binding

08:54 <TheAnachron> typo seems to be: 255. > Warning: Could not resolve keysym XF86KbdInputAssistPrevgrou

08:55 <TheAnachron> missing a p here? ;-)

08:55 <tiosgz> but also, "Return" isn't a valid riverctl command, so there must be a superfluous argument or something :)

08:58 <tiosgz> (and river would complain differently: "error: invalid keysym 'nonsense-key'" immediately upon trying to map the key)

08:58 <TheAnachron> yeah maybe posting the config would be a good idea.

09:53 <ifreund> TheAnachron: that message is just normal Xwayland noise iirc

15:30 <gbrlsnchs> Hey folks, when I'm playing games (more specifically Steam games) and I "minimize" (or in other words unfocus) a game's window there's a big FPS drop (the giveaway here is the game's sound, it starts to sound weird and slow). I have been working around this by wrapping games with Gamescope, but that breaks the Steam overlay (a Gamescope problem).

15:31 <gbrlsnchs> Question is, is there a way to stop this? Sometimes it causes online games to glitch and soft lock (not River's fault of course, but if I could avoid this altogether would be great)

15:33 <gbrlsnchs> Gamescope has an option described as "game refresh rate when unfocused", so I use this for games like, say, Elden Ring, where the FPS drop when unfocused would kick you out of the online experience

15:33 <gbrlsnchs> But Gamescope breaks the Steam overlay, not to mention everything with it is either windowed or fullscreen

16:28 <ifreund> gbrlsnchs: this is caused by an impedence mismatch between the wayland frame callback mechanism and the event loop of the (usually X11) games in question

16:28 <ifreund> iirc the fifo-v1 protocol is supposed to improve the situation but has been stuck in review for quite a while

16:32 <ifreund> hmm, maybe the newish suspended xdg-toplevel state is also meant as a workaround for this

16:33 <ifreund> I dont recall if wlr_scene sends that automatically for us or not

16:39 <vyivel> it doesn't

16:46 <ifreund> hmm, thats not great since wlr_scene will stop sending frame callbacks due to totally occlusion of a surface

16:47 <ifreund> and iirc doesn't provide a good way for the compositor to query whether or not a surface is fully occluded

18:28 <gbrlsnchs> how does Gamescope works around that?

22:07 <toxic_fume> Turns out I double-mapped the keybind for my terminal and forgot to pass --show to wofi. I thought since I'd only made small adjustments to the example config, it was unlikely an issue, especially since neither of my binds for running applications were working, but it turns out I was wrong :/

22:07 <toxic_fume> Thank you everyone for the help, and sorry for bothering you with this, the EDL errors tripped me up into assuming it was a harder problem than it was.