forked from bartvdbraak/blender
added notes on collisions between dynas and "static moving" geometry (kinematically controlled)
This commit is contained in:
parent
1ca9bded73
commit
ae459bfc12
@ -97,6 +97,17 @@
|
|||||||
/// \todo moving infinite mass objects should impart extra impulse to objects they collide with
|
/// \todo moving infinite mass objects should impart extra impulse to objects they collide with
|
||||||
///
|
///
|
||||||
/// currently ODE's ERP pushes them apart but doesn't account for their motion.
|
/// currently ODE's ERP pushes them apart but doesn't account for their motion.
|
||||||
|
/// you have to detect if one body in a collision is a non-dyna. This
|
||||||
|
/// requires adding a new accessor method to
|
||||||
|
/// KX_IPhysicsInterfaceController to access the hidden m_isDyna variable,
|
||||||
|
/// currently it can only be written, not read). If one of the bodies in a
|
||||||
|
/// collision is a non-dyna, then impart an extra impulse based on the
|
||||||
|
/// motion of the static object (using its last 2 frames as an approximation
|
||||||
|
/// of its linear and angular velocity). Linear velocity is easy to
|
||||||
|
/// approximate, but angular? you have orientation at this frame and
|
||||||
|
/// orientation at previous frame. The question is what is the angular
|
||||||
|
/// velocity which would have taken you from the previous frame's orientation
|
||||||
|
/// to this frame's orientation?
|
||||||
///
|
///
|
||||||
/// \todo allow tweaking bounding volume size
|
/// \todo allow tweaking bounding volume size
|
||||||
///
|
///
|
||||||
|
Loading…
Reference in New Issue
Block a user