Bad rotation algorithm

[eyqs]

The rotation algorithm doesn't work when viewing things in 4D. Especially worrying is how rotating stuff the opposite way doesn't actually unrotate it when points start to fall apart. At least everything works in 3D...

Ideas:

• come up with a new algorithm
• steal an old algorithm from somewhere
• use like, actual matrix rotations or quaternions or whatever
• keep the old algorithm because it's easier to extend to higher dimensions? is it?