rec.autos.simulators

I'm still looking forward as much to GP3 as I did to GP2, but I don't think
it will bump GPL from its first place. It will give a great feeling of an F1
race weekend but it won't be total realism.

I don't think he's trolling, just a little confused maybe :-)

Not sure how much he played the full version, but the demo is not a good
thing to base comments about the handling on as the physics were tweaked
quite a bit for the full version.
The cars back then did slide, drift whatever. I own quite a few '60s
racing videos and GPL seems pretty accurate in that respect IMO.

--
Cheers!
Graeme Nash

% I don't have the cdrom, it's on order though. I did have it when it
% first came out, but moved across country and sold it along with
% everything else. Do have the demo, and am basing my comments on that.
% Just find it very hard to stay on the track at the full realism mode.
% With the trainer it's not a problem. Can't remember from the full
% version, can you race the trainers?

The patch also changed the way the tires react as the slip angle
changes.  I think the patch brought it a little bit closer to what you
are looking for (but still far from 100% accurate).

--
**************************** Michael E. Carver *************************
     Upside out, or inside down...False alarm the only game in town.

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=<[ /./.  [-  < ]>=-=-=-=-=-=-=-=-=-=-=-=-=-=

% I feel I have to disagree with you. GP2 does have a pretty realistic
% physics model. I never saw any of the trigger-event things within it.
% People mostly complain about canned spins. Those spins were indeed not
% canned. They were only very abrupt, but you could catch them with good
% response.

I disagree, but not to mean that GP2 didn't feel right.  Geoff just
programmed the feel so well we were convinced there was a great physic
engine.  If there was a real physic engine we should have been able to
drive a car that reacted like the different marques.  Instead all cars
had the same "canned" physics.  My theory is that was because there was
no real physics engine to plug in the different variables for each
marque.

% If you want to check this theory, and have F1 2000, try lowering the
% inertia values for the cars and engine in the vehicle files by half and
% then increase the tyre stiffness to double the value. In effect you will
% get a feel so much like GP2 that it is almost scary. And F1 2000 does
% model physics particularly well with all 6 DOF, only second perhaps to
% GPL. Apart from large pitch and roll angles I think GP2 modelled pretty
% much everything that is essential. Rest assured that GP3 will be at
% least on the GPL level physics-wise.

GP1/2 tracks never were accurate when it came to camber, crowns, etc.
Basically they were flat tracks with elevation changes.  However, it
does seem from what I have heard the GP3 changes this (I certainly hope
so).

--
**************************** Michael E. Carver *************************
     Upside out, or inside down...False alarm the only game in town.

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=<[ /./.  [-  < ]>=-=-=-=-=-=-=-=-=-=-=-=-=-=

: % I feel I have to disagree with you. GP2 does have a pretty realistic
: % physics model. I never saw any of the trigger-event things within it.
: % People mostly complain about canned spins. Those spins were indeed not
: % canned. They were only very abrupt, but you could catch them with good
: % response.

: I disagree, but not to mean that GP2 didn't feel right.  Geoff just
: programmed the feel so well we were convinced there was a great physic
: engine.  If there was a real physic engine we should have been able to
: drive a car that reacted like the different marques.  Instead all cars
: had the same "canned" physics.  My theory is that was because there was
: no real physics engine to plug in the different variables for each
: marque.

  This sort of thing always annoys me.  There is no such thing as
"canned" physics.  Or alternatively all physics in all physics models
I've ever seen is "canned" physics.

  GP2 makes certain simplifying assumptions, GPL makes different (and
slightly fewer) simplifying assumptions.  If GP2 wanted to make the
different cars handle differently then, for example, the author could
have changed tyre friction, drag, aero downforce and ground effect
down force from each vehicle as well as changing the centre of gravity
and moment of inertia.  The GP2 model was certainly complex enough for
all of these factors to be an effect and this would certainly have had
the effect of changing the handling of the cars.

--
Richard G. Clegg       Only the mind is waving
    Networks and Non-Linear Dynamics Group
      Dept. of Mathematics, Uni. of York
     UPDATED WWW: http://manor.york.ac.uk/

This is what is meant by canned physics, i.e. it is actually not an
approximation or simplification of the real laws but a way around them.
This is also what differentiates simulations from mere games, actually.

-Gregor

I cannot agree with that simply because it is possible to test the car
in GP2 through all the possible regimes and see that it indeed does
behave like a car should. Believe me, I would indeed know canned physics
if I saw it. The sims that indeed do have it canned are rather easy to
sort out.

If your theory was correct then the setup options wouldn't be nearly as
responsive and work as in the real world as they were in GP2 as a canned
physics model would much more easily differentiate between cars as it
would between sometimes really subtle setup changes. Why there is only
one F1 car is quite simple to explain. While the cars of the GPL era had
very differing weight distributions, inertia values, wheelbase and
widths, the different (more) modern F1 cars are much more common in this
respect, in a way due to more strict regulations. It is simply the lack
of knowledge of the subtle differences between the cars of the '94 era
of F1 that made no sense in producing different properties of the cars,
as the differences are not even really well known. Even F1 2000 does not
try to replicate these differences (apart from he engine power) even
though its physics model is superior to GP2.

-Gregor

--
Ian Parker

==

%   This sort of thing always annoys me.  There is no such thing as
% "canned" physics.  Or alternatively all physics in all physics models
% I've ever seen is "canned" physics.

There is such a thing as "canned" physics.  Instead of modeling the
individual physics of a car (ala GPL), an algorhythm is built that says
if A & B are applied at the same time C will happen.  If 1/2 A & 3/4 B are
applied then D will happen.  These are then built up to a further
degree.  Actual physics are not modeled, just the behavior.  Now the
other version of "canned" physics is not what I would really call canned
(but that is just semantics as far as you are concerned).  This version
models actual physics of the car and track (tire patches, engine torque,
spring loads, tire camber, track camber).  To me there is a difference
between modeling physics and modeling the effects of physics without
modeling the affects.  While both maybe contained within a program
(i.e., wrapped in a container -- thus canned in a sense), the latter is
truly "canned" in the sense of a "canned" laugh track.

Main Entry: canned
Pronunciation: 'kand
Function: adjective
Date: 1904
1 a : prepared or recorded in advance; especially : prepared in
standardized
form for nonspecific use or wide distribution <canned laughter> <canned
music> <canned speeches> b : lacking originality or individuality as if
mass-produced <canned sales pitch>

The author could have, IF the actual physics were modeled.  However, if
they weren't, the process of coding the effects (without having modeled
the affects) could be extremely complicated and time consuming.
Especially if one wanted the effects to be equal to the standards that
Geoff has set himself in his programs.  That is why I theorize that GPx
is basically a "canned" (notice quotes) physics simulation.  In GPL the
individual cars react differently not because the programs wrote new
canned routines for each car, but because they plugged in the variables
in each car's physics.  This was possible because GPL was based around a
physics model.  I doubt that this is/was possible with Geoff's approach
to coding GPx.

--
**************************** Michael E. Carver *************************
     Upside out, or inside down...False alarm the only game in town.

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=<[ /./.  [-  < ]>=-=-=-=-=-=-=-=-=-=-=-=-=-=

% I cannot agree with that simply because it is possible to test the car
% in GP2 through all the possible regimes and see that it indeed does
% behave like a car should. Believe me, I would indeed know canned physics
% if I saw it. The sims that indeed do have it canned are rather easy to
% sort out.

This theory in part is based on past posts (a long long time ago) in
r.a.s. about Geoff's approach to programming Grand Prix 1/2.  Some were
based on interviews with Geoff.  Unfortunately, deja.com is
re-organizing their system and I have no way to go back in time to track
down these discussions.  

% If your theory was correct then the setup options wouldn't be nearly as
% responsive and work as in the real world as they were in GP2 as a canned
% physics model would much more easily differentiate between cars as it
% would between sometimes really subtle setup changes. Why there is only
% one F1 car is quite simple to explain. While the cars of the GPL era had
% very differing weight distributions, inertia values, wheelbase and
% widths, the different (more) modern F1 cars are much more common in this
% respect, in a way due to more strict regulations. It is simply the lack
% of knowledge of the subtle differences between the cars of the '94 era
% of F1 that made no sense in producing different properties of the cars,
% as the differences are not even really well known. Even F1 2000 does not
% try to replicate these differences (apart from he engine power) even
% though its physics model is superior to GP2.

Basically I think Geoff is an extremely talented and artistic programmer
and that is why he was able to deliver the nuances in setup vs.
performance you discribe.  Not necessarily because GPx was loaded with a
complex physics engine.  That's the beauty of his creations and also why
his creations differ so much from the ones from Papyrus Group.  Once
again, I am not "knocking" Geoff or his creations, just the opposite.

The physics in both GPx and f12k do exist to some degree, but not close
to the level that are coded into GPL (and appears to be in WSC and
MotorSims programs).  Yes, they may model a varying amount of tire
contact, and possibily a gross approximation of drag (i.e., wings), grip,
etc. f12k ups the ante by modeling more axis points than GP did, but I
highly doubt that aerodymanics are modeled, nor indiviual suspension
parts, engine torque, weight distribution, etc.  

So if I undertand you correctly we could put a Ferrari engine in a
Jag chassis and we would have an even match between the two cars?  Or if
all of the marques ran Mercedez engines, all teams would be even?  Some
one should tell the engineers and owners of the F1 teams that they are
wasting millions of dollars in chassis and suspension development costs.
<G>  If what you say about strict regulations preventing much difference
between the performance of one chassis over the other were true, then
running an F1 team wouldn't be so expensive.  

--
**************************** Michael E. Carver *************************
     Upside out, or inside down...False alarm the only game in town.

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=<[ /./.  [-  < ]>=-=-=-=-=-=-=-=-=-=-=-=-=-=

:-)

Thx, that's all I wanted to know. Is sliding around a bend/corner
realistic or not. Now I can look forward to my copy of GPL arriving
knowing that when I slide off the track into a fence that the same
thing would have happened in real life. :-)

--
Nos

The full version is easier to drive, but you have to have 36 fps all the
time or any version will be undriveable.
Yes, you can race the trainers, even online - have a look at VROC.net for
details.

Bye,
Leo

Well, if I can race the trainers then I'll have no problem. :-)
I have a P3 500 128mb ram (soon to be 256mb) and V3 3000.
Should get an OK frame rate in GPL with that if I thin out the field.

--
Nos

Stephen

--
Regards,
Bruce Kennewell,
Canberra, Australia.
---------------------------