rec.autos.simulators

Here are the new features:
- Pacejka tire model; better than the original curves.
- Force feedback! Yes, with Pacejka this was a piece of cake. HOWEVER,
the response I feel with my MSFF wheel is too late. Will have to do
some future prediction like GPL, but how...
- Optimized triangle intersection; your FPS may go up a bit.
- Engine torque curve normalized, with a torque factor to get the
maximum torque.
- Released a small track called 'Carrera'; very cute, and it contains
a bridge where you can drive under and on. So it's basically Suzuka
ready. :)
- Physics of the Ferrari once more revised (for Pacejka) so it now
weighs twice as much.
- A shadow is projected onto the ground. Looks much nicer.
- Shifting with buttons.
- Mouse support (select the mouse controller from Config).
- A digital clutch has been included (it's analog, but handled in a
digital way).
- Engine inertia added; the engine must now spin up as well as the
rest of the drivetrain.
- Engine rpm is better simulated.
- A crude open differential is simulated.
- Wheel lock is handled better; you used to be able to steer the car
even with fully locking front wheels.
- Bug fixed; QLOG.txt could get uncontrollably large on Windows
systems. Limit is now 1Mb.
- Source code included (not yet compilable; more coming later)

It's still quite a bit of ice-skating. You *can* however feel the
steering become light when it breaks out. There's too much lateral
sliding, but when I take that out you have no breakouts currently, so
I left the lateral sliding in.

Ruud van Gaal
Ruud van Gaal, GPL Rank +53.25
Pencil art    : http://www.racesimcentral.net/
Free car sim  : http://www.racesimcentral.net/

Like the track, don't so much like the unintentional and seemingly
unavoidable spins when cornering at 60 km/h....

Keep at it Ruud, lots of potential there!

:0)

Jim

Shane

    I have a question about gravity.. How is it set up?  Is it that
everything just falls in a general direction?  or is there a central point
some 300 miles below the track?  Just curious.  Seems like they would be
setup much like the old "Asteroids" game where the space ship would fall
towards an astroid and you had to use the thrusters to maneuver away from
it.  And you could use the thruster and the astroid gravity to swing around
the asroid.

:)

Ah, Friday, 4 pm, WEEKEND!!! And F1 from Brazil, in all digital glory :))))

--
Olav K. Malmin
remove .spam when replying

-Gregor

Hm, well it DOES restore the desktop settings now, Jim. ;-)
But you're right, the main focus should now be the tire model, since
it STILL doesn't feel GPL-ish (making donuts seem impossible,
break-outs are so suddenly). Perhaps a better open differential, where
torque isn't divided 50%-50% but instead more torque flows to the
spinning wheel (like it should) will make the cornering a bit better.
But I want donuts! :)

And splines in my tracks; braking while coming off the bridge is
nearly impossible.

Ruud van Gaal, GPL Rank +53.25
Pencil art    : http://www.marketgraph.nl/gallery/
Free car sim  : http://www.marketgraph.nl/gallery/racer/

...

Weird. What wheel do you have?
If you turn the wheel right, the wheels turn left as well? In that
case, copy msff(_ff).ini to mycontroller.ini and change:

controls
{
  steerleft
  {
    ...
    min=0
    max=-1000
  }
  steerright
  { ...
    min=0
    max=1000
  }

into

controls
{
  steerleft
  {
    ...
    min=0
    max=1000
  }
  steerright
  { ...
    min=0
    max=-1000
  }

In other words, negate the X axis. If it works, mail it to me with a
description of your wheel (split axis? probably) and I'll include it
for a next release.

Gravity pulls at the wheels and the body in the down direction. That
is (0,-1,0) in OpenGL terms. Nothing 'falls', this is the indirect
effect of a gravity *force* that is applied to the 5 affected bodies
(4x wheel plus the body).
For all these bodies, the force is applied at their center of gravity
(which is the center of the model actually for the car, and the center
of the wheel for the wheels).
It seems falling of the bridge fully rolling the car upside down seems
to work anyway. :) (although there's no collision detection, so the
enormous jumps after hitting the ground are because the speeds are too
big to use regular suspension forces; instead, a direct hit (impulse)
model must be used in the future).

Cool, very nice effects can be reached using such a method. Like Lunar
Lander, brilliant in design (costing no processor power, just a single
spring model).

Ruud van Gaal, GPL Rank +53.25
Pencil art    : http://www.marketgraph.nl/gallery/
Free car sim  : http://www.marketgraph.nl/gallery/racer/

Hm, let's try to install DX8 for thrillseeking's sake first, so I can
watch your sim. ;-)

Ruud van Gaal, GPL Rank +53.25
Pencil art    : http://www.marketgraph.nl/gallery/
Free car sim  : http://www.marketgraph.nl/gallery/racer/

...

Yes, I model the tracks curved, then apply it to an Earth-sized curve,
then process each vertex to have the gravity direct to the center of
the planet (the center of gravity), then normalize these vectors
(align to the universe Y-axis) to get a virtual uncurved version of
the track (where all gravity vectors point along the universe Y-axis).
Strangely, as it turned out, the resulting processed track was flat!
So I took the code out.

Sja! ;-)

*writes down* A turret, yes, that's missing too.

Ruud van Gaal, GPL Rank +53.25
Pencil art    : http://www.marketgraph.nl/gallery/
Free car sim  : http://www.marketgraph.nl/gallery/racer/

Ash

|
| ...
| >Excellent! But isn't a atronomical gravity model a bit too over the
| >top in racing sims ? I don't even think the curvature of the earth is
| >modelled in the tracks, and if it was, gravity would be a force
| >straight down.
|
| Yes, I model the tracks curved, then apply it to an Earth-sized curve,
| then process each vertex to have the gravity direct to the center of
| the planet (the center of gravity), then normalize these vectors
| (align to the universe Y-axis) to get a virtual uncurved version of
| the track (where all gravity vectors point along the universe Y-axis).
| Strangely, as it turned out, the resulting processed track was flat!
| So I took the code out.
|
| Sja! ;-)
|
| *writes down* A turret, yes, that's missing too.
|
|
| Ruud van Gaal, GPL Rank +53.25
| Pencil art    : http://www.marketgraph.nl/gallery/
| Free car sim  : http://www.marketgraph.nl/gallery/racer/

    I was just curious as to how gravity worked in a Sim.  I used Asteroids
as an example...  There's apparently a point in the center of each Asteroid
where the ship is pulled to (center of the Sims Earth) and the outer crust
of the Asteroid is a surface that the Ship can't enter to continue it's
jurny to the center of that asteroid (surface of the Sims Earth).    The way
I understand your statement about Gravity in the Sim, If I created a track
that started at the base of a 20 mile high mountain and ran to the top, I
would still be pulled towards the track at the same rate at the top of that
mountain as I would at it's base. (an exaggerated concept to help me further
understand gravity in a Sim).

Crossing fingers .... :)

-Gregor

Now, back to some reading... And F3000 and F1... Geez. Gotta get some
food to keep the body running.

Ruud van Gaal, GPL Rank +53.25
Pencil art    : http://www.marketgraph.nl/gallery/
Free car sim  : http://www.marketgraph.nl/gallery/racer/

...

Ah, the homemade 4-axis one. Download the JoystImm.exe util from my
download page and try that to see which axes are present. Print out
one of the MSFF controller files and see if you can work it out.
Reverse the 1000/-1000 in steerleft{} and steerright{} for a quicker
try.
You can try to attach the clutch to an axis btw, but it's not really
analog just yet (quite a few nifty things that need work when you've
got a real analog clutch).

...

Yes, you've got it. The reason for this is like any other aspect of
racing sims; if it's too insignificant, leave it out first, then if it
all works brilliantly, you might want to get it in for detail's sake.
As most tracks don't have a lot of height difference, you can assume
the gravity as constant. The only effect that *might* make enough
difference is whether you were driving on one of the Poles. IIRC the
gravity is larger there, since the Earth is squashed, like a
fast-rotating waterfilled balloon.
What you *will* encounter in some of the sims (GPL has some remarks on
this in its files) is the air density. Altitude, or rather, air
pressure is quite important for aerodynamics, so some sims adjust that
for a given track (though it remains constant while driving).

The sim works (in my case and most others) with a Newtonian model;
forces, torques (which are forces but rotational ones instead of
translational) and MKS (meter, kilogram, second). You apply forces and
torques to your bodies, then calculate accelerations using F=m*a (one
of Newton's laws). As all is dependent on one another, you take small
steps, say h=0.01s, then figure out all the forces again. In GPL this
all is done 288 times per second.
Anyway, one of the forces is a gravity force, which always pulls
directly down (in world coordinates that is). Close enough for all
racing done on this planet. Although ofcourse a SuperGalactic Stunt
Car Racer might break this rule and need orbital calculations. ;-)

Ruud van Gaal, GPL Rank +53.25
Pencil art    : http://www.marketgraph.nl/gallery/
Free car sim  : http://www.marketgraph.nl/gallery/racer/