[ODE] Problem with Trimesh

Stuart O Anderson soa at andrew.cmu.edu
Mon Nov 17 12:34:25 MST 2003


The collision tests seem to work fine, so I don't think there's a problem
with getting the data into ODE/OPCODE.  Just something weird when I try
and get it back out.

Incidentally, has anyone implemented Plane-Mesh collisions for ODE?  I'm
working on it now, but if it already exists that would save me some time.

stuart

On Mon, 17 Nov 2003, Flavien Brebion wrote:

> Two things come to my mind (by the way, i haven't looked at that
> part of the code for some time, so i'm not 100% sure of what i'm
> going to say:)
>
> 1. When you specify the vertex array for the tri-mesh, make sure
> you aren't deleting the memory afterwards. I don't think ODE keeps
> a safe copy of the vertex array, so the pointer must still be valid
> at run-time.
>
> 2. Maybe more probable: invalid stride parameter ? If your vertex
> array is made of 3 floats per vertex, the stride should be 12 bytes.
>
> F. Brebion
>
>
>
> -----Original Message-----
> From: ode-bounces at q12.org [mailto:ode-bounces at q12.org]On Behalf Of
> Stuart O Anderson
> Sent: Monday, November 17, 2003 3:47 PM
> To: ode at q12.org
> Subject: [ODE] Problem with Trimesh
>
>
> Hi -
>
> I've got a problem with dGeomTriMeshGetTriangle, which appears both in my
> own code and with the test_trimesh code.  Essentially I'm getting the
> wrong values back from the function.  For example, I retreive vtx 1 of
> triangle 1 in test_trimesh right before calling dsSimulationLoop for the
> first time.  It get <1,-5,-5,0>, which doesn't make any sense since no
> vertex should have '1' as a component.  (the correct result would have
> been <-5,-5,2.5>).  I checked out the relevant code but can't see anything
> obviously wrong.  Maybe something inside of OPCODE?  has anyone else run
> into this?
>
> Stuart
> _______________________________________________
> ODE mailing list
> ODE at q12.org
> http://q12.org/mailman/listinfo/ode
>
>


More information about the ODE mailing list