Re: [PATCH v2 0/5] MR4092: user32: Actually implement RealGetWindowClass{A/W}.

[Jinoh Kang (＠iamahuman)]

> Depending on whether any of the builtin winprocs tolerate being called on another process's window,

AFAIK the real-window-class changing behavior can be observed via mere `WM_NULL`, which is unlikely to rely on process-specific state.
> it may not be possible to test.

You mean it may not be possible to test NtUserSetRealClassId *indirectly* via builtin winprocs? If so, what prevents us from calling it directly? `win32u:win32u` already has a few tests for class names.