rec.autos.simulators

GPL's excessive Processor demands

Tony Whitle

GPL's excessive Processor demands

by Tony Whitle » Fri, 15 Jan 1999 04:00:00

Excuse me for not updating my PC just to run GPL but I have been wondering
why Papyrus' wonderful programmers could not have added a slider to control
the complexity of the physics engine. It seems from reading this ng and the
interviews with David Kaemmer that:
* it updates the car you are driving 288 times a second (I find that hard to
believe)
* the flow of air through the engine is modelled (why?)
* the AI cars also have a complex physics engine (not quite as complex as
the driver's car, I hope)
* etc., etc., etc.

All of this meaning a major load on the processor in addition to processing
the 3D graphics. Simply reducing the update rate might allow me to get
20fps, while simplifying the AI physics and removing some of the more
esoteric modelling might even get me to 36fps. I realise it is heresy to
suggest anything less than 100% realism (though I don't know what that means
when I am sitting in a swivel chair at a desk, driving the "car" using a
potentiometer) but it just might have improved the sales.

Next week:
How having the option to make the cars easier to drive would have led to
better reviews (followed by my excommunication from r.a.s)

Tony Whitley
Cyrix 200MX/Voodoo2 (which runs Viper/TOCA/CMR/F1RS2 quite happily, thank
you)

Jason Harriso

GPL's excessive Processor demands

by Jason Harriso » Fri, 15 Jan 1999 04:00:00

GPL was written for the crowd who WANT these things
in a simulation and are prepared to run it on a decent
computer in order to get good performance.

BTW - with Pentium3 on the way, your computer really
isn't going to be up to much software being released
in the next 12 months.

Stick to the games that run well on your PC I guess - GPL
isn't about good reviews or simplified physics, it's about
accurate simulation.  Cutting the meat of the sim out because
your PC is too slow would cut the heart out of the game
and make it another arcade game.

Just my 0.02

Jason

[snip comments on reducing realism]

Tim (fusio

GPL's excessive Processor demands

by Tim (fusio » Fri, 15 Jan 1999 04:00:00

[snip]

The more stuff you model this way the less you have to do with
scripted routines, which makes for a very consistent and believable
sim.
I don't truly know if GPL models the cooling dynamically, but if it
does, don't you think that's better than some kind of script that says
"if driver runs at X RPM for X time, the car will overheat"?

I really don't know if turning down the rate at which the sim is
updated would be a viable option.
The computer controlled cars behave very well, and cutting something
like that might cause them to behave crazily. Kind of like the stuff I
read about CART:Precision Racing. :-)

I don't see how anyone could fault you for this. As long as it is a
toggle, nobody should have an objection to it.
I think the fact that you can't adjust the computer cars speed without
screwing with a file was a major mistake on Papyrus' part.

Making the car easy to drive would hurt the spirit of the game in my
opinion, but a slider from 10%-100% driver skill (for computer cars)
would have gone a long way toward making the sim more accessible
without dumbing down the physics model.

I had an older Cyrix 200/Voodoo1, and while I liked it at the time, I
found it to have pretty spotty performance at some games, and be good
with others.
Odd stuff, too. For example:
Mechwarrior2:Mercenaries ran faster on the Cyrix than my P166, but
Interstate 76 (which uses the same engine in modified form) ran at
about 9 FPS on the Cyrix and 25 on the P166.
Basically, I think they're a great Windows processor, but not the
choice for a game processor.

I do think we might see some shouting when NASCAR3 comes out if
Papyrus can't speed up the GPL engine.
Running GPL with an 8-10 car field is fine, but less than a 20 car
field will suck for a NASCAR sim (IMO, of course).
The fact that we'll have 10 times more computer neophyte's buying
NASCAR3 than GPL, then coming here to squawk about it, should make RAS
an interesting place, I'm sure. :-)

--

        http://www.users.fast.net/~fusion1
    (dirt bikes, rat bikes, rental car abuse...)

Swindell

GPL's excessive Processor demands

by Swindell » Fri, 15 Jan 1999 04:00:00

I saw this word and would like to know - if anyone could tell me?

Rob

KPineb

GPL's excessive Processor demands

by KPineb » Fri, 15 Jan 1999 04:00:00

Sorry, Rob, but I find this a little chilling.  You own a computer, but don't
own a dictionary.  With respect ... go buy one and don't embarrass yourself in
front of thousands.  You'll find "heresy" under the "H" section.

I'm bummed.

Kurt

Walk Walke

GPL's excessive Processor demands

by Walk Walke » Fri, 15 Jan 1999 04:00:00

Quote from George Carlin -

"You don't like it? Don't eat it....."

-/- Walk Walker


>Excuse me for not updating my PC just to run GPL but I have been wondering
>why Papyrus' wonderful programmers could not have added a slider to control
>the complexity of the physics engine. It seems from reading this ng and the
>interviews with David Kaemmer that:
>* it updates the car you are driving 288 times a second (I find that hard
to
>believe)
>* the flow of air through the engine is modelled (why?)
>* the AI cars also have a complex physics engine (not quite as complex as
>the driver's car, I hope)
>* etc., etc., etc.

>All of this meaning a major load on the processor in addition to processing
>the 3D graphics. Simply reducing the update rate might allow me to get
>20fps, while simplifying the AI physics and removing some of the more
>esoteric modelling might even get me to 36fps. I realise it is heresy to
>suggest anything less than 100% realism (though I don't know what that
means
>when I am sitting in a swivel chair at a desk, driving the "car" using a
>potentiometer) but it just might have improved the sales.

>Next week:
>How having the option to make the cars easier to drive would have led to
>better reviews (followed by my excommunication from r.a.s)

>Tony Whitley
>Cyrix 200MX/Voodoo2 (which runs Viper/TOCA/CMR/F1RS2 quite happily, thank
>you)

Daxe Rexfor

GPL's excessive Processor demands

by Daxe Rexfor » Fri, 15 Jan 1999 04:00:00


>Excuse me for not updating my PC just to run GPL but I have been wondering
>why Papyrus' wonderful programmers could not have added a slider to control
>the complexity of the physics engine.

I think I recall reading in an online review somewhere that GPL was unique
as an offering from Papyrus in that it is able to run with all of its
options turned on/up on hardware available at its release.  In other words,
it was designed LESS in anticipation of future hardware than any of their
other releases.

On the other hand, look at the N2/ICR2 family of games.  How old is this
engine by now? 4 years?  By doing little more than adding native support for
3Dfx, an API that wasn't terribly popular when they came out, they have
extended to useful life of at least N2 right up through the beginning of
1999.

I don't think anyone is going to argue with the success of N2, and that has
in part to do with it having staying power on hardware that has grown in
capabilities by leaps and bounds.  At this time in 1994 I had a 486DX2/66
with 1 MEG of on-board VL bus graphics, 24 MEG of RAM, an SB16 ASP, a 2X
CD-ROM and a cruddy joystick.  ICR2 was fun then, ran pretty darn good on
that hardware (considering), and N2 ran even better with a few tweaks from
Papyrus for the framerate and so forth.  While both of these games are
looking a little dated (Though 3Dfx  N1999E looks great) they are still
enjoyable and playable on my P2/450 with 2 graphics card totalling 20 MEG,
256 MEG of RAM, SB Live! Value, a 32X CD-ROM and a MSFF wheel.

Something tells me that if they went out of their way to accomodate people
running 386sx25s when ICR2 came out, the physics modelling would be so poor
by comparison these days that no-one would be buying N1999.  Designing games
to run best on very high end machines at the time of the games release
simply ensures that the game will continue to be enjoyable farther into the
future.  People in this NG are running GPL sucessfully on well-sorted P200s
and enjoying it, and people are running it on much faster hardware, too, and
enjoying it just even more.

Your P200MMX is dated.  That is the reality of the computer hardware
treadmill.  I get a new computer (or at least MB/CPU) every 18 months so I
can always have something capable of running the latest.  I don't HAVE to, I
do it by choice.  IF I don't do that, pretty soon there will be a whole slew
of games that won't run for sh*t on my computer.  Believe it or not, it
isn't the routine upgraders like the members of this group that drive the
software guys to write bigger and better games, it is John and Jane Q.
Public and their snotty-nosed kids who just got a new family computer and
want some 'kewl' games.  It is the fact that they bought the biggest and
fastest computer to run their games even though they can barely find the
START button, that compels the software companies to meet the escalating
hardware capabilities.

The hardware bar is raised by mainstream PC buyers/users and the game
companies are simply responding.

Whatever,

~daxe
"religion is remedial spirituality"

-----------== Posted via Newsfeeds.Com, Uncensored Usenet News ==----------
 http://www.newsfeeds.com/       The Largest Usenet Servers in the World!
-----------== Over 66,000 Groups, Plus  a  Dedicated  Binaries Server ==----------

Zonk

GPL's excessive Processor demands

by Zonk » Fri, 15 Jan 1999 04:00:00


>Sorry, Rob, but I find this a little chilling.  You own a computer, but don't
>own a dictionary.  With respect ... go buy one and don't embarrass yourself in
>front of thousands.  You'll find "heresy" under the "H" section.

>I'm bummed.

>Kurt

http://www.m-w.com/netdict.htm

(and i wonder if he was about to order from amazon....)

XCR6

GPL's excessive Processor demands

by XCR6 » Fri, 15 Jan 1999 04:00:00

<<I do think we might see some shouting when NASCAR3 comes out if
Papyrus can't speed up the GPL engine.
Running GPL with an 8-10 car field is fine, but less than a 20 car
field will suck for a NASCAR sim (IMO, of course).
The fact that we'll have 10 times more computer neophyte's buying
NASCAR3 than GPL, then coming here to squawk about it, should make RAS
an interesting place, I'm sure. :-)>>>>

 Keep in mind that in NASCAR3 they wont have to model the suspension movements
that you see on the GPL AI cars, or the suspension movements that you see from
inside the GPL***pit. let alone haviing to model tires to look realisticly,
and act the same way. I think that N3 will be much easier on CPU demands
comparitively  speaking. N3 you'll only have to look at a body flexing with
weight transfer, and maybe see a bit of the tire where it meets the road, or
see it spinning from the side view. keeping these things in mind, I believe we
have a good thing on the horizon to look forward to, and by then most hardware
will  run it adequatly<forgive the spelling <G>>.

T_K

'John' Joao Sil

GPL's excessive Processor demands

by 'John' Joao Sil » Fri, 15 Jan 1999 04:00:00

GPL does indeed model dynamic cooling, my Eagle setup for Spa runs
very high 5th gear RPM's as my Eagle peaks at 201mph into the Masta
kink and Stavelot, where I can see the engine temperature get dangerously
hot, I am forced to shortshift the rest of the lap on the way up from
Stavelo all the way until after Eau Rouge to reduce the engine temperature,
so that the engine has cooled and is ready to peak out at high RPM's into the
long downhill straights again on the next lap.

You can visibly see the effects of running extended high RPM's on the
engine heat, and I've found that my Eagle's engine will blow at Spa
on PRO damage settings if I do not keep an eye on the temp and manage
it as I describe above. *This* is probably very close to what real
drivers have to do to manage engine performance -vs- engine wear...

Tony, it's detailed realism touches like this that make me a Racing
Simulation nut, I love GPL for things like this, when I want more dumbed
down physics and realism I load up one of my favorite arcade racers like
the NFS series or Screamer series, which I also love for racing when I'm
in a less serious mood, but when I don't feel like worrying whether I
am redlining the RPM's on my Ferrari Testarossa, or why the AI magically
knows which side of the road I am trying to pass on and automatically moves
to block.

While I agree that a difficulty slider might have been a good option
for novices in GPL and I think it is an important feature for Nascar3,
I *don't* think that every sim necessarily needs to cater to the
lowest common denominator out there both in skills or level of computing
power, maybe I'm biased by also being a flight-sim fan, but there are
plenty of great arcade racing games out there that not every serious racing
sim has to be muddied up to also please everyone.

If features can be added that are toggleable on and off then yes by
all means I am for them, but fudging the physics and realism numbers
just to try and please everyone leads to mediocre sims/arcade-racer
products like Cart Precision Racing which neither satisfy the diehard
auto sim enthusiast nor the arcade sim racer... IMHO (not a flame)

And periodic computer upgrades are just a reality of being a simulation
enthusiast, Falcon3.0 urged my upgrade from a 386 to a 486, ICR2 GP2 and NFS
were partly responsible for my upgrade from a 486/66 to a P166 a few
years ago, and now GPL and Falcon4.0 encouraged my upgrade to a
Celeron P2/300a system a few months back, that is just a price of getting
closer and closer to *real* simulations, more computing power needed to
calculate all the extra variables, I'm just glad at the amount of computing
power that you can buy nowadays for much less money than just one or two years
ago. My latest system cost around one third the price of my two previous
system upgrades, and I've done system upgrades usually every 2-3 years or so.

With the right hardware GPL is amazing, and IMHO GPL does justify a hardware
upgrade. Yes that is the fanatic part of being an auto sim fan.

Seeyas on the track.

--John (Joao) Silva




>[snip]
>>* the flow of air through the engine is modelled (why?)
>>* the AI cars also have a complex physics engine (not quite as complex as
>>the driver's car, I hope)
>* etc., etc., etc.
>I don't truly know if GPL models the cooling dynamically, but if it
>does, don't you think that's better than some kind of script that says
>"if driver runs at X RPM for X time, the car will overheat"?

SNIP!
Alan Chandl

GPL's excessive Processor demands

by Alan Chandl » Fri, 15 Jan 1999 04:00:00

On Thu, 14 Jan 1999 09:03:58 -0000 "Tony Whitley"


>Next week:
>How having the option to make the cars easier to drive would have led to
>better reviews (followed by my excommunication from r.a.s)

I downloaded a utility that I found on one of the standard GPL
websites that lets you downgrade the opposition (actually you can do
it manually - tweaking driver.ini but its much easier with the
program).  It doesn't take long to control the car well enough to at
least get round the track - and you can pull back the opposition to be
as bad as you are :-)

Alan


http://www.chandler.u-net.com

ymenar

GPL's excessive Processor demands

by ymenar » Fri, 15 Jan 1999 04:00:00

XCR650 wrote

Actually even if it's not shown on the screen it has to be modeled. The same
goes for your rear suspension. Even if you don't "see" it it's modeled of
course.. the same will happen with N3. That's why with AI in GPL it almost
exponentially takes your CPU. Because it has to model it even if you don't
see it "somehow" (I won't go in the prediction code and so...).

The visual effects you see on the screen are just the answer to the
mathematical calculations of the game engine.

I don't see any difference between modeling a stock-car than a 67 F1 "in
general".  They are both a race car with real life physics.. they act with
the same laws of gravity and forces.  They have 4wheels that have weight
shifting depending on the actions of the car.  The only differences are in
the various differences in chassis, aerodynamics, tyre grip, shock/spring
settings, etc..

-= Fran?ois Mnard <ymenard/Nas-Frank>
-= NROS Nascar sanctioned Guide http://www.racesimcentral.net/
-= SimRacing Online http://www.racesimcentral.net/
-= Official mentally retarded guy of r.a.s.
-= May the Downforce be with you...

"People think it must be fun to be a super genius, but they don't realise
how hard it is to put up with all the idiots in the world."

Nathan Won

GPL's excessive Processor demands

by Nathan Won » Fri, 15 Jan 1999 04:00:00


> screwing with a file was a major mistake on Papyrus' part.
> Making the car easy to drive would hurt the spirit of the game in my
> opinion, but a slider from 10%-100% driver skill (for computer cars)
> would have gone a long way toward making the sim more accessible
> without dumbing down the physics model.

I agree. Yes, there is a feeling of accomplishment when you finally do
beat car #19 (out of 20 cars), but it's also fun to actually win a race
every now and then. I used the file that someone created that actually
lets you modify the strengths of the opponents and it truly does make the
sim much better, in my opinion. I could see why Papyrus left it out, but
as a marketing stand point they shouldn't have.
Motorcar Journa

GPL's excessive Processor demands

by Motorcar Journa » Fri, 15 Jan 1999 04:00:00

GPL is a simulator you know. Why cut out all the good stuff just for some
people who think their PC has eternal lives?
Buy a PII 300 and put in some extra 96 Mb and you will be ***ed to
realism in a few hours (minutes, seconds)

It think GPL could be even more realistic than it allready is (no offence
btw. I love it even more than Falcon 4.0)

Natan

--
Natan Tazelaar
Founder & Editor-in-Chief
Motorcar Journal
Free Dutch Carmagazine
www.motorcar-journal.com


>Excuse me for not updating my PC just to run GPL but I have been wondering
>why Papyrus' wonderful programmers could not have added a slider to control
>the complexity of the physics engine. It seems from reading this ng and the
>interviews with David Kaemmer that:
>* it updates the car you are driving 288 times a second (I find that hard
to
>believe)
>* the flow of air through the engine is modelled (why?)
>* the AI cars also have a complex physics engine (not quite as complex as
>the driver's car, I hope)
>* etc., etc., etc.

>All of this meaning a major load on the processor in addition to processing
>the 3D graphics. Simply reducing the update rate might allow me to get
>20fps, while simplifying the AI physics and removing some of the more
>esoteric modelling might even get me to 36fps. I realise it is heresy to
>suggest anything less than 100% realism (though I don't know what that
means
>when I am sitting in a swivel chair at a desk, driving the "car" using a
>potentiometer) but it just might have improved the sales.

>Next week:
>How having the option to make the cars easier to drive would have led to
>better reviews (followed by my excommunication from r.a.s)

>Tony Whitley
>Cyrix 200MX/Voodoo2 (which runs Viper/TOCA/CMR/F1RS2 quite happily, thank
>you)

Dave Hawn

GPL's excessive Processor demands

by Dave Hawn » Sat, 16 Jan 1999 04:00:00


> GPL was written for the crowd who WANT these things
> in a simulation and are prepared to run it on a decent
> computer in order to get good performance.

> BTW - with Pentium3 on the way, your computer really
> isn't going to be up to much software being released
> in the next 12 months.

> Stick to the games that run well on your PC I guess - GPL
> isn't about good reviews or simplified physics, it's about
> accurate simulation.  Cutting the meat of the sim out because
> your PC is too slow would cut the heart out of the game
> and make it another arcade game.

> Just my 0.02

> Jason

You took the words right out of my mouth Jason.... 100% correct!
Dave Hawnt (UK)

rec.autos.simulators is a usenet newsgroup formed in December, 1993. As this group was always unmoderated there may be some spam or off topic articles included. Some links do point back to racesimcentral.net as we could not validate the original address. Please report any pages that you believe warrant deletion from this archive (include the link in your email). RaceSimCentral.net is in no way responsible and does not endorse any of the content herein.