Cumulative error when moving mouse long distances

[Qarj]

In my testing there appears to be some kind of possible cumulative error that creeps in when moving the mouse long distances.

In the function move_to_area I removed the random component for the target location.

In a test I found moving the mouse from 229,199 to 6312,1589 the final calculated point by the function bezier(ts) is in fact [6414.064814991564, 1628.489157870313] which is in fact way off target (using deviation of 30 and speed of 4).

The workaround solution for me appears to be call the function twice.

Thanks very much for this by the way, it looks like a very useful project!

[dav-ell]

Running into this same issue.