[ODE] Question about variable step size and collision detection

Per Bull Holmen pbholmen at yahoo.com
Mon Dec 18 15:42:39 MST 2006

--- Filip Joelsson <filip at blueturtle.nu> skrev:

> If you want pixel perfect collision detection, I'd
> suggest that you
> implement your own sprite class - instead of using
> cylinders. For
> collision detection with sprites, all you need is to
> store beginning and
> end for each row (which you'll do anyway if they are
> run length encoded)
> - and check for overlaps. ODE allows you to plug
> your own CD in - so
> this can be _very_ efficient.

Yeah, this is a good idea. But for my first attempt
I'll try out Bullet instead. Bullet handles continuous
collision detection even when the objects are
rotating, out of the box. Modelling the geoms for cd
should be a breeze - the balls will have varying
sizes, so it's easier to render them realtime anyway.
I see someone saying the dynamics of ODE is better,
but for my game super-realistic dynamics is not of
primary concern.

> I have seen that ODE is used in a couple of 2D
> applications. Would there
> be an interest in a sprite class for the main tree?

This could be a great idea for other users, especially
if it can derive the outer shape of the objects by
looking at the alpha layer of the bitmap, so that
normal vectors are computed correctly for the
collisions. I probably won't use that solution anyway.
A class like that should probably be designed in such
a way that a separation between graphics rendering and
physics modeling is still expressed.


Alt i én. Få Yahoo! Mail med adressekartotek, kalender og
notisblokk. http://no.mail.yahoo.com

More information about the ODE mailing list