[ODE] interactive and continuous physics collisions?

[Jon Watte (ODE)]

You need a fixed frame rate for physics (fixed step size).
http://www.mindcontrol.org/~hplus/graphics/game_loop.html

You can also do manual detection, where you sweep a sphere across the 
movement arc (creating a capsule), and do the appropriate detection on 
that, instead of on a static sphere.

Cheers,

             / h+

Onur Tugcu wrote:
> Hello,
>
> I'm working on a networked 2d table game. My scene is simple,
>
> a ball sphere
> a player1 sphere
> a player2 sphere
> constrained onto the plane, sphere centers are right on the plane.
>
> 4 thin boxes on the sides constraining the ball within the table's surface.
>
> The players control massless geometries (spheres) to hit the ball and score.
>
> When the geometries stand still, the ball collides and bounces fine,
> same works for the side boxes.
>
> When the geometries hit the ball slowly, things are still well, and
> the game is somehow playable.
>
> But when the players move too quickly, they will pass through the
> ball, or do a strange bouncing (since the next simulation step has the
> ball inside the player's sphere)
>
> Also when the ball hits the boxes too fast, it will pass through them
> and continue flying off.
>
> So is there any way of constraining the ball on a limited surface area
> or constraining the ball to never pass through certain objects like
> spheres?
>
> Am I doing something wrong? is there a better way of doing all this?
>
> Thanks in advance.
> _______________________________________________
> ODE mailing list
> ODE at ode.org
> http://ode.org/mailman/listinfo/ode
>
>
>