Quote: "That’s not bad at all chap what you was working on why did you give it up?
The handling is not bad although it drives more like a rear wheel drive than a 4 wheel drive."
I was just experimenting, as you do, with a view to making a new version of Victory Road (hence VR2). The thing was, what you see there was weeks of trial and error to eridicate, as much as possible, the infamous "wheel wobble" where the axel joints & suspension just couldn't cope with the forces, so you'd end up
1) Not being able to steer properly
2) The rear wheels not following the car around bends - they'd just try to go straight on!!
Even on that demo right there, you'll notice that the wheels could "Pidgeon Foot" inwards or outward, and once the wheel goes to a right angle, then the car flips over. The speed in that demo was just not sufficient for that to happen though.
What I'm trying to say is that it just didn't really work, I'd find that I needed to set the "MAX_ITERATIONS" of the physics simulation to some way over the top number (like 50 instead of 10) to try to keep things under control. Thats b@ggered the other 40 cars I was gonna have on the track then!
Maybe I was just doing it wrong, maybe there was a bug in the physics engine, maybe the types of joints I had to use were not designed to do what they were doing. It actually occures to me now that perhaps I should have thrown away the "Car demo" I started with and re-done everything from scratch, taking care to simulate wishbone suspension and proper axels etc, basically make the model more realistic in the hope that the physics would follow suit (but not grind to a halt under the processing).
Maybe *good* vehicle physics, even when using a good, but generic physics engine, is still hard to do. Maybe there are just too many parameters for the average bod to figure out what settings to apply for decent results. Maybe either you need a really good example to work from, or the engine needs to have a generic "Vehicle" class built in (as per DP), that can easily be tweaked.
Whatever, the fact was that it started not being very fun anymore and it was clear that I was trying to base the gameplay around something that was outside of my control. That is super fine when its all going well, you stay sane, but when you run into a bug/problems then it all starts unravelling real quick!
I would only hope/assume that this kindof thing would not apply to a commercial physics engine like PhysX, which, as far as I can tell, acts like Direct3D does for HW graphics in that it allows direct manipulation of the PPU, but comes with a "Reference Rasterizer" as well, for when the hardware is not there.
All things being equal, I'm expecting some good stuff and look forward to seeing what you can do with it
