> > Poly-based physics = garbage.
> Well, poly-based game engines with physics incorporated into that is the
> future of ***. I said in the past, that when this will happen
> completely, games will become "less" realistic but it's the risk they have
> to take. The computing power just isn't there, heck still in F1C there's
> way too much stuff that's not poly-based and still running in it's own
> virtual world.
> Is poly-based physics garbage presently? Yes. But I don't want sims in 10
> years to still have completely different entities, physics-wise and
> game-engine wise. Like Papy has. I don't want my car, graphically on
> screen, to be unrelated in it's world to what really happens behind the
> scene with the physics. That's not the way to go.
Papy's tracks are surface based, which means the surface is described by a
formula. Any position on each surface can be found by using U/V barycentric
(sp?) co-ordinates (generated by collision detection). This accurately
determines the height of the ground and therefore how much the tyre is
"squashed".
This can be done relatively easily (and quickly) with todays processing
power (and yesterdays!). Also this "physics" representation is "tesselated"
(turned into polys) depending on how close the viewer is to each surface
(more polys for closer surfaces). Therefore the "gfx" representation is
exactly the same as the physics representation.
I'm pretty sure thats how its done for Papy, it can certainly done on
today's hardware. Native support for surfaces will be (is?) supported in
hardware soon (if not already).
Thats how I intend to do things at least :)
All the best,
Ash
http://www.racesimcentral.net/
