[ODE] Demo of upcoming kart game

[Jeroen Schmitz]

Erwin de Vries wrote:
> This looks pretty nice. Can you share more details about how you
> implemented the physics? I believe i did notice some AI cars going to
> the air very shortly when bumping into a wall, but most of the time
> the dynamics seem to be solid and predictable.

I will post some more details about the way I implemented it tomorrow,
when I'm back in the office. The physics are indeed ok most of the time,
but it should be solid all the time :)
After all, I don't think people will appreciate it when their kart goes
flying around every now and then, unless we would release this game as a
goofy stunt racer (which it isn't supposed to be ;) ).

> Some questions. Did you model the kart as 1 box, or did you use rays
> or spheres for the wheels? There seems to be no suspension in the
> karts, so i assume its just 1 box?

The kart is implemented using a box (for the car body) and 4 spheres
(for the wheels), each connected to the box with a 2-hinge joint. So
it's done in the most common way that vehicles are implemente in ODE
applications, I think. The karts do indeed have almost no suspension,
but that's also the case with real karts (actually, the only
'suspension' you have in a kart comes from the tires).

> When a kart collides with a wall
> or another kart its position is influenced only a little, and speed
> is preserved. Did you do anything special to get this behavor? Did
> you use the tricollider to do your collisions with the walls?

I used the tricollider for both the walls and the road. Only the wheel
spheres have collision with the road triangle mesh and only the car body
box has collision with the wall triangle mesh. I think part of the
problem with the karts going up the wall may be caused by the
not-so-perfect working box-tricollider collision (at least, I read that
this wasn't perfect yet). Kart-kart collision is only done by box-box
collision.


> My fastest time was 25:67. Is that good? :-)

I think it is :)

Jeroen