[ODE] Stability of ODE in NovodeX Rocket...

[Achim Moller]

Didn't compile ODEPSCL by myself. Just used the precompiled file from
SourceForge.

>From the ReadMe of ODEPSCL:
[...]
There were a few changes made to Open Dynamics Engine in order to exploit
features in PSCL
[...]
Linear/Angular Damping and Max Angular Velocity. Simply put you can now
apply linear and angular damping on bodies in ODE as well as cap the max
angular velocity of a body. This increases stability and helps makes the
simulation feel better in many cases.
[...]


If I understand you right, you suggest it's better to recompile ODEPSCL
using the latest source tree from CVS? Or from where do these mentioned
features (auto-dampenig, "good" auto-sleep) come?

BTW, I'm also wondering about some of the mentioned limitations:
For sample: "Friction ramp – No material support yet in ODE version, but it
runs". Why isn't there "material support"?

Achim





> 
> Did you make sure there's a little bit of auto-dampening? 
> Was auto-sleep turned on?
> Was it the "good" version of auto-sleep which nukes accumulated
> velocities, forces and torquest to 0?
> 
> Cheers,
> 
> 		/ h+
> 
> 
> -----Original Message-----
> From: ode-bounces at q12.org [mailto:ode-bounces at q12.org]On Behalf Of Achim
> Moller
> Sent: Tuesday, October 19, 2004 9:37 AM
> To: ode at q12.org
> Subject: [ODE] Stability of ODE in NovodeX Rocket...
> 
> 
> Hi,
> 
> I was playing around with ODEPSCL in NovodeX Rocket, but results are often
> not very good meaning that ODE is a lot more unstable in the provided test
> scenes. Basically, boxes jitter a lot more.
> 
> Does anyone have some ideas why ODE seems to be so unstable in contrast to
> NovodeX? Perhaps we can discuss some (missing?) features of the ODE solver
> core which would improve general ODE system stability? 
> 
> Any comments or pointers are welcome.
> 
> Thanks,
> Achim
> 
> 
> _______________________________________________
> ODE mailing list
> ODE at q12.org
> http://q12.org/mailman/listinfo/ode
>