[ODE] Patch to add a new joint in ODE

[Thomas Paviot]

Hi Remi,

Remi Ricard a écrit :
> Hi Thomas,
>
>>
>>> Hi,
>>>
>>> I'm only posting this here to let the developers with svn access know
>>> that there is a new patch.I submitted the patch on the sourceforge 
>>> website.
>>>
>>> The patch add a new joint to ODE.
>>>
>>> The new joint is like a slider joint but the bodies are free to 
>>> rotate in one direction. The axis of rotation is the same as the 
>>> slider axis.
>>>
>>> The patch include all the modification to ..h and .cpp files.
>>> A simple demo demo_piston.cpp and finally patch for the premake.
>>>
>>> This is a patch against the latest trunk code (revision 1173).
>>>
>>> The patch was produced with:
>>> cd openode; svn diff joint_piston.patch
>>>
>>>   
>> Hi,
>>
>> I'm surprised there's no dJointSetPistonAnchor method in the patch 
>> you submitted (there's only a dJointSetAxis one).
>>
>> Around wich axis ( a point and 3 vects) is computed the rotation? How 
>> is it possible to modify this point location?
>>
>> Thomas
>
> It was kind of busy here so it took time before finding the good answer.
>
> Before submitting the joint I only tested when body1 and body2 have 
> their center of mass as follow
> cm1 = cm2 + A x axis_of_prismatic
>
> i.e cm aligned with the axis of the prismatic.
>
> yesterday someone here tested the new joint when there is an offset 
> and was also asking about the anchor point.
>
> So I think you are right to have a more general joint an anchor point 
> will be needed.
> I might have time to had that after you rush for a demo.
>
> For now you can consider the anchor point as the center of body2 if I 
> remember correctly.
>
>
> Remi
>
>
>
That's the answer i expected. Your PistonJoint, enhanced with an 
explicit anchor, could become a new joint named CylindricalJoint. The 
SetPistonForce method should be useless.

I'm really impatient to test your work.

Regards,

Thomas