Page 1 of 1
How many iteration should be used for collision detection
Posted: Wed Aug 27, 2014 6:33 pm
by apapaxionga
When handling collision contact, collision constraint should be iteratively solved. However, I have a question that what kind of collision constraint should be iteratively solved ? Obviously, there are some kinds of collision constraint do not need be iteratively solved. Can someone explain to me what kind of role iteration play in collision constraint solving?
Re: How many iteration should be used for collision detectio
Posted: Wed Aug 27, 2014 8:47 pm
by bone
Iterative solving is one way to handle multiple constraints that are not independent of each other.
It's not clear why it is "obvious" to you that some kinds of collision constraints do not need to be iteratively solved. If it is the only constraint that needs to be solved, or it is completely independent of any other constraint, that may be true. Perhaps to back up for a moment and get some context, what kind of physics system are you using?
Re: How many iteration should be used for collision detectio
Posted: Wed Aug 27, 2014 11:00 pm
by apapaxionga
bone wrote:Iterative solving is one way to handle multiple constraints that are not independent of each other.
It's not clear why it is "obvious" to you that some kinds of collision constraints do not need to be iteratively solved. If it is the only constraint that needs to be solved, or it is completely independent of any other constraint, that may be true. Perhaps to back up for a moment and get some context, what kind of physics system are you using?
That's exactly I want, thx!