Re: [simspark-devel] [Sserver-three-d] Fw: [sserver-commits] CVS:rcsoccersim/rcssserver3D/lib/oxygen/physicsserverspace.cpp, 1

[Feng X. <hen...@ma...>]

Hi Joschka,

    Great idea! I was somehow blind with this great feature of ODE!
    I will change the code ASAP.

Thanks a lot!



Feng Xue
2008-04-14



Sent By： Joschka Boedecker
Sent At： 2008-04-14 12:58:03
Sent To： Feng Xue
CC To： simspark-devel
Topic： Re: [Sserver-three-d] Fw: [sserver-commits] CVS:rcsoccersim/rcssserver3D/lib/oxygen/physicsserverspace.cpp, 1

Hi Feng,
On Apr 12, 2008, at 4:38 PM, Markus Rollmann wrote:
> Hi,
>
> Feng Xue wrote:
>>    Yes, the nao model has been modeled like that(some bodyparts are  
>> in
>> some bodyparts totally). And if I use universal joint instead, it
>> is probably okay. But the universal joint is not equal to two hinge
>> joints and is hard to control for the joint limits according to  
>> current
>> work. Besides, refering to Webots' implemention, they use hinge  
>> joint on
>> all the joints too.
>
> I still don't see why it is necessary that two geoms are constantly
> colliding by design in the nao robot model. But if that's a  
> requirement,
> so be it.
Just another question on this: if some parts of Nao are completely  
enclosed in others, wouldn't it be a solution just not to add  
colliders for these bodies? Or does ODE not allow this (don't remember  
now)?
Cheers,
Joschka