[ODE] Ball joints

[Brad]

That's what I don't understand. I'm adding the torque in global 
coordinates, so why would it matter where the body is in its cycle 
around the Z axis? It seems more like I'm adding torque around the 
body's local axis instead of the global axis.

Jon Watte (ODE) wrote:

>
> Second, dBodyAddTorque() adds torque in global coordinates, so the 
> result will be different depending on where in its cycle around the Z 
> axis the rotating body is (assuming it's not just attached at the 
> center).
>