[ODE] OPCODE trimesh troubles -- collisions with undesired faces

[jon klein]

I just mentioned OPCODE in case the scenario is handled differently  
there -- I haven't tried with GIMPACT (because I'm using ODE with  
double precision).

Below is the vertex and index data I've got for this test shape.  The  
two objects are placed with a 90 degree rotation away from each other  
about the Z-axis, and offset from each other on the Z-axis.  One  
object then slowly drifts into the other along the Z-axis.  Gravity is  
disabled to simplify the demo.

Thanks,

-- jon klein


float v[] = {
   0.500000, 1.000000, 1.000000,
   0.500000, -1.000000, 1.000000,
   0.500000, -1.000000, -1.000000,
   0.500000, 1.000000, -1.000000,
   -0.500000, 1.000000, 1.000000,
   -0.500000, -1.000000, 1.000000,
   -0.500000, 1.000000, -1.000000,
   -0.500000, -1.000000, -1.000000,
};

int i[] = {
   0, 1, 2,
   0, 2, 3,
   4, 5, 1,
   4, 1, 0,
   6, 7, 5,
   6, 5, 4,
   3, 2, 7,
   3, 7, 6,
   0, 3, 6,
   0, 6, 4,
   1, 5, 7,
   1, 7, 2
};






On Nov 20, 2007, at 2:45 AM, Daniel K. O. wrote:

> jon klein escreveu:
>> I've used trimeshes before with some success, but I've now run into a
>> problem using OPCODE with certain meshes that I cannot figure out.
>
> Just wondering, does your mention about OPCODE means that it works on
> GIMPACT?
>
>
> It could be bad data. If possible, give us the trimesh data you are
> feeding to ODE, with the exact positions of the geometries (so we can
> just run dCollide() between the geoms). Your collision callback code
> would help too.
>
>
>
> -- 
> Daniel K. O.
> "The only way to succeed is to build success yourself"
> _______________________________________________
> ODE mailing list
> ODE at ode.org
> http://ode.org/mailman/listinfo/ode