[ODE] OPCODE

[Flavien Brebion]

The solution i'm using now is to cast a ray from the previous
body's position to its next one (which is already a first problem,
since to know the "next position" in advance, you have to run a
timestep first, then step back to the previous position), and find
if there is an intersection with the world.

If there is an intersection, i then recursively split the timestep
at the intersection's time.

It gets complicated with more than one object, and is not perfect,
but seemed to fix 99.99% of my problems.

F. Brebion


-----Original Message-----
From: ode-bounces at q12.org [mailto:ode-bounces at q12.org]On Behalf Of Aras
Pranckevicius
Sent: Friday, February 13, 2004 3:39 PM
To: ode at q12.org
Subject: Re: [ODE] OPCODE


>       Anyone, explain please, as OPCODE takes into account the
> velocity of the object at finding of the collisions?      I use
> TriMesh and bodies moving quickly often fly through him.      Can
> anyone will prompt output their given to situations?

No, OPCODE doesn't do that. ODE also doesn't do that. Continuous collision
is
a hard problem to solve...

Easy solutions to try:
. decrease timestep.
. make larger bodies :)
. make slower bodies :)
. you could explicitly use some swept-geoms for some things. Eg. a fast
travelling bullet might use a ray segment, a cylinder or capsule (ccylinder)
for collision so that it covers the path travelled during timestep.



Aras Pranckevicius aka NeARAZ
http://www.gim.ktu.lt/nesnausk/nearaz/


_______________________________________________
ODE mailing list
ODE at q12.org
http://q12.org/mailman/listinfo/ode