[ODE] Jitters

[Tim Rightnour]

On 08-Nov-2004 Jon Watte wrote:
> Note that the value of the rotation is "90" in all the steps, 
> yet your code ought to try to make that rotate towards 0. I 
> think the source of your problem is elsewhere.

I guess I should be more specific here, the NEW/OLD value is the current
position of the object.  Ie, the object is at 1200,400,1300 with a euler facing
of 0,90,0.  Thats not a rotation torque, thats the direction the box has been
rotated.

> Are you running collisions? Is this sword held by a character? 
> Does the sword collide with the character? Does that generate 
> a contact?

There is a dTerrainY beneath the sword.  I am running collisions.  What I find
most puzzling about this is, there is a rock, just to the NE of the sword. 
That rock stabilizes quickly (far faster with worldstep than quickstep, but it
stabilizes either way), and is not affected by the damping routine at all.

> Perhaps you could put a data breakpoint where the Y force and 
> velocity of that object is updated? Perhaps you could log all 
> the joints/contacts that affect the thing?

When the damping is applied, the sword never creates any contacts, nor has any
joints.  It immediately speeds up and shoots straight down through the ground.

first step:  lvel = {0, -1.96000004, 0, 0}, avel = {0, 0, 0, 0}
             facc = {0, 0, 0, 0}, tacc = {0, 0, 0, 0}
after damp: lvel = {0, -1.96000004, 0, 0}, avel = {0, 0, 0, 0}
            facc = {0, 0.0196000002, 0, 0}, tacc = {0, 0, 0, 0}
second step:  lvel = {0, 94.3329697, 0, 0}, avel = {0, 0, 0, 0}, 
              facc = {0, 0, 0, 0}, tacc = {0, 0, 0, 0}
after damp: lvel = {0, 94.3329697, 0, 0}, avel = {0, 0, 0, 0}, 
            facc = {0, -0.943329692, 0, 0}, tacc = {0, 0, 0, 0}

I'm rather confused by this, I don't see a connection between those linear
velocity numbers at all.

---
Tim Rightnour <root at garbled.net>
NetBSD: Free multi-architecture OS http://www.netbsd.org/
NetBSD supported hardware database: http://mail-index.netbsd.org/cgi-bin/hw.cgi