[ODE] When rockets hit objects

Graham Fyffe gfyffe at gmail.com
Mon Feb 21 10:24:20 MST 2005


You might want to use a few rays instead of a sphere to test the
rocket collisions... basically use a ray that extends from where the
rocket was in the previous frame to where the rocket is in the current
frame.  You can use several rays to fill out the shape a little
better, or if you're clever you could even use a capped cylinder as a
swept sphere.  I forget if OPCODE will do capped cylinders correctly. 
This is still not perfect, but it should reduce the number of false
negatives you get.  For example, rockets will not be able to pass
through walls.  I've used the ray technique in other engines before
and it seemed reliable enough.

- Graham Fyffe


On Mon, 21 Feb 2005 08:35:57 -0300 (ART), Jose Marin
<jose_marin2 at yahoo.com.br> wrote:
> Hello.
> 
> I'm using ODE on my game, and would like to use OPCODE
> (inside ODE) to collision detection.
> 
> The game has a terrain, some moveable objects (tanks),
> some static objects (trees, buildings, etc) and lots
> of rockets flying!
> 
> I use spaces and trimeshes geoms for all objects,
> including the terrain.
> 
> Is possible to use OPCODE to detect the collision
> between the  rockets and the other objects?
> 
> The problems are:
> 1) the rockets are very fast (may lose some
> collisions?)
> 2) there are lots of rockets (say, 20 at one time!)
> 3) how to detect when a rocket hit some object?
> 4) how to identify this object?
> 5) the rockets may use boxes or spheres to detect
> collisions; which one is faster (may be a matter of I
> make some tests)?
> 
> It's possible to use only OPCODE to collision
> detection, in this case, or I must use a custom one?
> 
> Thanks for any tips!
> 
> Jose
> 
> 
> _______________________________________________________
> Yahoo! Acesso Grátis - Instale o discador do Yahoo! agora. http://br.acesso.yahoo.com/ - Internet rápida e grátis
> _______________________________________________
> ODE mailing list
> ODE at q12.org
> http://q12.org/mailman/listinfo/ode
>



More information about the ODE mailing list