Consumed modifier keys required in keybindings

[OJFord]

System

OS: Linux
Version: alacritty 0.10.1 (2844606d)
Linux/BSD: Wayland, sway

Logs

[12.571405548s] [INFO ] [alacritty] glutin event: WindowEvent { window_id: WindowId(Wayland(WindowId(94144256298704))), event: KeyboardInput { device_id: DeviceId(Wayland(DeviceId)), input: KeyboardInput { scancode: 47, state: Pressed, virtual_keycode: Some(Paste), modifiers: LOGO }, is_synthetic: false } }

---

When binding to At on a US layout, for example, it is necessary in Alacritty to supply mods: Shift, when really this was already consumed in producing the At.

I came across this in adding a Super+v binding for XF86Paste, and initially thought I'd done something wrong in XKB, since Alacritty would see Super+XF86Paste rather than just XF86Paste. I've confirmed though with wev, wshowkeys, and by trying something similar with Shift+2 (for At) that the binding is correct.

In case it's of any use, it seems rofi had and fixed a similar issue: davatorium/rofi#485

[chrisduerr]

It's a known issue that the current keyboard handling isn't ideal. This requires some upstream work and is tracked there and some other Alacritty issues. At the moment there's no way around this in Alacritty itself.

[OJFord]

Ok, thanks @chrisduerr. If you have a chance, could you link me one of the others or upstream issues please?

I had another look but I assume they're more generic/nuanced, so difficult to find without prior knowledge. Just so I have something to subscribe to to keep track of. Cheers!

[chrisduerr]

Upstream is rust-windowing/winit#753. There should be a couple open Alacritty issues linked too already.