Setting Initial Rotation

[optimisticmonkey]

I am trying to set set the initial rotation of a body before beginning a simulation.

It seems after the first step, the rotation is reset to 1,0. The attached shape is a ball, and is not in contact with anything at start.

Is there any reason that the rotation should be being reset?

Once running, the rotation changes caused by collisions work fine.

Thanks for any feedback

[ker]

This should not happen, are you sure you aren't doing anything after the first step? 1.0 is a very weird value since chipmunk uses radians (so 1.0 is like 70 odd degrees or something)

have you checked if setting the rotation after the first step will cause everything to work? (increases probability you are doing something to the ball)

[slembcke]

Are you talking about the rotation vector and mean you are setting it to the vector (1,0)?

There are several linked values in cpBody for optimization purposes. i/i_inv, m/m_inv, a/rot. You must use the setter function to set these values or you will get weird behavior.

[optimisticmonkey]

Thanks for the quick reply -

I was setting the body->rot directly....and leaving body->a unmodifed.

I will try using cpBodySetAngle instead.