Now themes are loaded once per scale and stored centrally not per
window. They are not unloaded till application shutdown. Since there
is unlikely to be more than two or three scales this is not a big waste
of resources. Since cursor lifetime is tied to theme lifetime and cursors
are stored per window, destroying a theme requires destroying all
cursors win all windows referring to that theme, which is too much work.
Should hopefully fix#2810
Now pressing a modifier key in one keyboard and a normal key in another works. Fixes#2362
Don't rebuild keymaps on new keyboard events that only change geometry. Fixes#2787
Better handling of multiple keyboards with incompatible layouts (this is for free from the above fixes). Fixes#2726
On my system reduces the number of getpid() calls from 67982 to 17327
I dont know why get framebuffer config attributes requires getpid() and
why it is so slow, probably a driver bug. Fixes#2754
Previously, the mouse back and forward buttons sent the same codes as
scroll up and down. Now they instead send the same codes as xterm. Mouse
button 10 (in X11 numbering) also now sends the same as xterm, instead
of not sending anything.
This also changes the `send_mouse_event` function which can be called
from kittens to use X11 numbering for mouse buttons instead of what it
previously used, which turns out to be a hybrid of X11 and GLFW. It was
documented to use GLFW numbering, but GLFW doesn't have numbers for
scroll events (that's separate events with x/y offsets) and 4 and 5 in
GLFW is actually back and forward, while `send_mouse_event` interpreted
it as scroll up and down.
That means that this is a breaking change for `send_mouse_event` because
it swaps the number for the middle and right button to be consistent
with X11. I did this because I think it's better to use one consistent
numbering scheme for the function, and because people probably know X11
numbering better than GLFW numbering and GLFW doesn't have numbers for
the scroll buttons.
This required the kitty package to be moved from Frameworks to
Resources, because Apple cant codesign .pyc files and its notarization
requires everything in Frameworks to be signed. Hopefully that does not
break anything. Also removed the kitty-deref-symlink since it is not
needed anyway and was only present for backwards compatibility.
Fixes#2040
