Re: Collision problems

On Jun 22, 2005, at 2:38 PM, Andrei Androsoff wrote:

> Hi, Jon and all,
>
> I'm having problems with collision detection - I think it's the  
> same problem James Grimes pointed out in Vol 17, Issue 3. It's with  
> the -X axis again, as well as the -Z axis, balls colliding with  
> Stationary cubes. I never see embedding, only passing right  
> through, and sphere/wall size seems to make no difference.
>
> Version 2.1b was to have fixed this but looks like it's reared its  
> head again.
>
> I wrote a sample program that tests +/- X/Z axes - the source is  
> linked here:
>
> http://www.magma.ca/~xI2aL0tK2/breve/collision_problems.txt

Hi Andrei,
thanks for the report.  After James' sent his demo I discovered some  
ugly problems with the collision detection code.  After spending some  
time working through these problems, I'm happy to say that collision  
response is far more stable that it's ever been before -- it makes  
James' demo run flawlessly (no penetrations -- the "bounce" issue is  
still somewhat unresolved).

If you're using Mac OS X, I have a test build which includes these  
fixes:

http://gibson.hampshire.edu/~jklein/breve_2.3b.dmg

I'd appreciate feedback on the physics changes in this build.  If  
there is interest, I can make some builds for Linux and Windows as  
well, though I'm not ready to make an official 2.3 beta yet.

- jon klein
_______________________________________________
breve mailing list
[email protected]
http://lists.spiderland.org/mailman/listinfo/breve