Consider Wayland mode

* https://github.com/owl-compositor/ow
* https://github.com/ravynsoft/ravynos/tree/main/CoreServices/WindowServer

If we are ever going to use Wayland, it would be absolutely non-negotiable to support all features which WindowManager in X11 mode already supports. For this, we would likely need to own the compositor in order not having to debate with fd.o (which is the upstream for wayland-protocols) which features it may and may not have. I could envision something where we:
* Use the most capable library abstracting away most of the Wayland standard protocol
* Introduce our own protocol to simulate arbitrary X11 atoms
* Implement all features we need that WindowManager supports in X11 mode without any discussion or debate whatsoever
* Document what we have implemented so that others (e.g., applications wanting to position windows on the screen) can take or leave it

Essentially out-engineer the debating circle that is fd.o that has been holding back proper desktop features in Wayland for decades.