FW: [ODE] Geometry passing through TriList

[Eric Cha]

Hmm, not knowing exactly how Erwin is doing his collision detection, It
seems like he is doing discrete time step sphere to polygon collision
checking.  If that is the case, I would suggest using an LSS (Line Swept
Sphere) vs polygon check using the start and end positions of the sphere
for each time step.  I'm not certain of the most "optimal" way, but if I
were to do it, I would implement it as a gross check (i.e. if the LSS
intersects the poly, then figure out at what time, based on the
intersect, and cut your time delta by that amount to step the simulation
"back" and use the regular sphere to poly collision at that point).

My .02

Eric 

> -----Original Message-----
> From: ode-admin@q12.org [mailto:ode-admin@q12.org] On Behalf
> Of Clay Larabie
> Sent: Thursday, August 01, 2002 9:38 PM
> To: Erwin de Vries; ode@q12.org
> Subject: Re: [ODE] Geometry passing through TriList
> 
> 
> What are your suggestions if I wanted to fix this?
> 
> It seems somewhat vital, to me at least, to have the
> collision detection work despite the step size, especially if 
> it's tied to render framerate, which in my game tends to 
> fluctuate a bit. I'm sure it'll be fairly constant once 
> optimized, but I'm not throttling frame rate to a fixed rate, 
> such as 30 fps.
> 
> Maybe I could go multi-threaded for the physics... though I'd
> like to avoid that.
> 
> By the way, you've done a great job on the TriList collision
> detection. I'm just trying to find the best solution for my situation.
> 
> Clay
> 
> ----- Original Message -----
> From: "Erwin de Vries" <erwin@vo.com>
> To: <ode@q12.org>
> Sent: Thursday, August 01, 2002 10:43 AM
> Subject: Re: [ODE] Geometry passing through TriList
> 
> 
> > This can indeed happen. The maximum stepsize you can take for the
> collision
> > detection to give accurate results is "(Diameter / 2) /
> Velocity" The
> > / 2 part will be fixed in the upcoming release. So a sphere with a
> > diameter of 10 units and a velocity of 1000 units will require a 
> > stepsize of (10 / 2)
> /
> > 1000 = 0.005 seconds
> >
> > The ray collision has nothing to do with this. In step N
> objects might
> > not collide, and in step N + 1 the collision might have
> taken place in
> > the meantime, but that is not simulated. It would be
> possible to fix
> > this of course, but it is not on my list.
> >
> > Erwin
> >
> > ----- Original Message -----
> > From: "Clay Larabie" <c-larabie@rogers.com>
> > To: <ode@q12.org>; <erwin@vo.com>
> > Sent: Thursday, August 01, 2002 18:36
> > Subject: [ODE] Geometry passing through TriList
> >
> >
> > > I've found that fast moving bodies pass through my
> TriList geometry.
> With
> > > the plane collision detection built into ODE, fast moving bodies
> > > will
> not
> > go
> > > right through the geometry.
> > >
> > > Erwin, do you have any thoughts on this? Have you noticed the same
> thing,
> > > and is there a way to fix it?
> > >
> > > You posted this earlier:
> > >
> > > "The next Tri-Collider release using Opcode 1.1 will also
> implement
> > > a
> ray
> > > collision function along with some other not too interesting
> > improvements."
> > >
> > > Will the ray collision detection solve the problem?
> > >
> > > Clay
> > >
> > >
> > >
> > > _______________________________________________
> > > ODE mailing list
> > > ODE@q12.org
> > > http://q12.org/mailman/listinfo/ode
> > >
> > >
> >
> > _______________________________________________
> > ODE mailing list
> > ODE@q12.org
> > http://q12.org/mailman/listinfo/ode
> >
> 
> _______________________________________________
> ODE mailing list
> ODE@q12.org
> http://q12.org/mailman/listinfo/ode
> 
> ---
> Incoming mail is certified Virus Free.
> Checked by AVG anti-virus system (http://www.grisoft.com).
> Version: 6.0.377 / Virus Database: 211 - Release Date: 7/15/2002
>  
> 

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.377 / Virus Database: 211 - Release Date: 7/15/2002
 

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.377 / Virus Database: 211 - Release Date: 7/15/2002