Kitty Keyboard encoding has incorrect modifier bits for modifier keys

[rockorager]

When using kitty keyboard protocol encoding, Ghostty incorrectly sets the modifier bits. Per the specification:

When the key event is related to an actual modifier key, the corresponding modifier’s bit must be set to the modifier state including the effect for the current event.

I tested this on Alacritty, Foot, Kitty, and Ghostty. All except Ghostty have the same encoding for this keypress.

I suspect this is only on a GTK build under Wayland, due to how Wayland delivers modifier presses + modifier events.

Repro Steps

1. kitten show_key -m kitty
2. Press control

Expected Output

CSI 57442 ; 5 u (press)

CSI 57442; 1 : 3 u (release)

Actual Output

CSI 57442 u (press)

CSI 57442 ; 5 : 3 u (release)

Reference: kovidgoyal/kitty#6913 (comment)

[wexman]

Seeing the ;5u here makes me think if my problem is related? It is that some keybindings are not working as expected. For example, ctrl+shift+comma which should reload the config insteat outputs ;5u to the screen.

Is that the same problem?

using zsh on Linux Mint with a german keyboard layout if that matters.

[00-kat]

Unlikely since the Kitty keyboard protocol isn't used for Ghostty's own keybindings. Though it's possible that they have the same underlying cause.

[wexman]

So should I open a separate issue for that?