rec.autos.simulators

by **Jeff Johnso** » Fri, 23 Oct 1998 04:00:00

Don't get me wrong. I've have every Papyrus sim and I really like GPL but
as far as I'm concerned they need to work on their 3D programming.
When Papyrus released the original ICR2 I liked it alot. The problem
was that even the fastest PC available at the time couldn't run it at 30fps
with all details on. Then along comes Rendition and ICR2 flies on even
modest hardware.
Now we have GPL and we are in the same boat as we were with the original
ICR2 and NASCAR Racing. Even with new 3D cards that have 3 to 4 times the
amount of memory, speed and sophistication, you once again need one of the
fastest PC's available to run GPL well. It's almost as it they have sucked
all the life out of the 3D cards (not to mention the fact that they are only
supporting 2). Maybe I would not have noticed it as much if others had not
been able to do it. Say what you want about F1RS and MGPRS2, you don't need
a Pentium II 400 to run them well. And I don't believe that if they had
GPL like physics in their sim engines you would need as much processing
power to run them well either.
I applaud Papyrus for getting the car physics as close to what I feel
they should be (since I never really raced that's just a guess). But I
don't applaud for their 3D. From a hardware standpoint GPL is like GP2 even
with a 3D card.

by **Chris Schlette** » Fri, 23 Oct 1998 04:00:00

Read all the posts on this newsgroup again concerning this topic.
Then perhaps you'll be a bit more educated. But to give you a quick hint:

ITS NOT THE 3D GRAPHICS THAT REQUIRE ALL THE CPU HORSEPOWER ITS THE PHYSICS.

Can we grasp that yet?

F1RS and MGPRS2 do NOT have the physics that GPL does, period...so of course
you wont need as much machine to run the game. Start adding in the multiple
AI cars which acts just as adding another real driver to a game, so you have
all the physics (the exact same physics engine that acts on the player's
car) acting on more and more AI cars and it takes a fast computer to handle
it. But even if those two sims did YES THEY WOULD NEED the same hardware
requirements.

And as for "sucking the life out of the 3D cards"...um, WHAT? Lets see, yes
they supported 2....Rendition they had experience with and previous titles
had support for it so their customer base had Rendition cards
installed.....and when they started writing the game 3dfx cards were
becoming very big. Not to mention that very few games actually use
OpenGL...and D3D was really quite a chore to write in back again when they
started over a year ago.

If you want eyecandy..go drive F1RS or MGPRS2..if you want physics, dont
*** about the steep requirements. Take the physics engine out of GPL and
your system requirements plunge quite drastically.

by **David Mocn** » Sat, 24 Oct 1998 04:00:00

Yes, you're right... GPL requires some damn fine hardware (PII 266+) and 3D
card. Though, from what I remember, there was no hardware to run GP2 when it
came out (some would say there still isn't anything that can run GP2 with
everything, without slowdowns at no time). On the other hand, there are a lot
of people who can run GLP with evrything all the time and get framerate
between 30-36 fps.
GPL requires top of the range hardware but unlike GP2, this harware at least
exists.

Regards,
David Mocnay

>Don't get me wrong. I've have every Papyrus sim and I really like GPL but
>as far as I'm concerned they need to work on their 3D programming.
> When Papyrus released the original ICR2 I liked it alot. The problem
>was that even the fastest PC available at the time couldn't run it at 30fps
>with all details on. Then along comes Rendition and ICR2 flies on even
>modest hardware.
> Now we have GPL and we are in the same boat as we were with the original
>ICR2 and NASCAR Racing. Even with new 3D cards that have 3 to 4 times the
>amount of memory, speed and sophistication, you once again need one of the
>fastest PC's available to run GPL well. It's almost as it they have sucked
>all the life out of the 3D cards (not to mention the fact that they are only
>supporting 2). Maybe I would not have noticed it as much if others had not
>been able to do it. Say what you want about F1RS and MGPRS2, you don't need
>a Pentium II 400 to run them well. And I don't believe that if they had
>GPL like physics in their sim engines you would need as much processing
>power to run them well either.
> I applaud Papyrus for getting the car physics as close to what I feel
>they should be (since I never really raced that's just a guess). But I
>don't applaud for their 3D. From a hardware standpoint GPL is like GP2 even
>with a 3D card.

by **Zonk** » Sat, 24 Oct 1998 04:00:00

>Read all the posts on this newsgroup again concerning this topic.
>Then perhaps you'll be a bit more educated. But to give you a quick hint:

>ITS NOT THE 3D GRAPHICS THAT REQUIRE ALL THE CPU HORSEPOWER ITS THE PHYSICS.

>Can we grasp that yet?

>F1RS and MGPRS2 do NOT have the physics that GPL does, period...so of course
>you wont need as much machine to run the game. Start adding in the multiple
>AI cars which acts just as adding another real driver to a game, so you have
>all the physics (the exact same physics engine that acts on the player's
>car) acting on more and more AI cars and it takes a fast computer to handle
>it. But even if those two sims did YES THEY WOULD NEED the same hardware
>requirements.

Correct me if i'm worng, but didn't we have some posts 6 months or so ago
which detailed physics modelling as account for around 10% of actual CPU
usage?

I'm not inclinded to believe this is changed, either.

Z.

by **Zonk** » Sat, 24 Oct 1998 04:00:00

>On Fri, 23 Oct 1998 08:11:07 GMT,

>>Correct me if i'm worng, but didn't we have some posts 6 months or so ago
>>which detailed physics modelling as account for around 10% of actual CPU
>>usage?

>I guess you're VERY wrong as it relates to GPL.
>I can also tell you that you are VERY wrong when it relates to ***
>combat flight sims. Physics occupy a large part of processing in these types
>of sims.

Then what kind of CPU time it taking? Anyone care to speculate?

Z.

by **Daniel Dru** » Sat, 24 Oct 1998 04:00:00

: Then what kind of CPU time it taking? Anyone care to speculate?
: Z.

Have Papyrus send you the source, and put a step-through de*** on it.

by **Daniel Dru** » Sat, 24 Oct 1998 04:00:00

: Yes, you're right... GPL requires some damn fine hardware (PII 266+) and 3D
: card. Though, from what I remember, there was no hardware to run GP2 when it
: came out (some would say there still isn't anything that can run GP2 with

Just for sake of argument, I run GPL in 640x480 with some details turned
off (set to "most") and the sky texture turned off. these aren't big
losses, the sun glare is still there, and the stands and stuff just
distract.

I run a P200MMX with a Voodoo2. I get 33fps, down to about 29 with 5 CPU
opponents. This isn't bad, IMO.

by **Daniel Dru** » Sat, 24 Oct 1998 04:00:00

: I would guess that the 'physics' take greater than 50% of, say, a
: Pentium 133. That would strictly be speculation. The upshot is that in
: order to run GPL on a low-end system, the graphics must be scaled back
: dramatically from what is possible on a more powerful system. If not,
: its a slide show.

Why don't you just ask WB's HT or somebody who knows just how much time
the CPU spends doing the floating point math functions for aerodynamics?

Everyone here is jumping around making wild speculations, and nobody's
produced an ounce of code.

by **Paul Jone** » Sat, 24 Oct 1998 04:00:00

Ok, so take the Voodoo2 card out of your box and replace with a Voodoo1 card
and you should see no loss of fps if it's all physics, provided the Voodoo1
can handle 36fps.
I agree - so much speculation and no evidence. MGP has substantially better
physics than F1RS but seems to run at about the same graphics speed. Then
again, than proves nothing, they may have just corrected errors in the F1RS
physics model without increasing the time spent calculating them.
I doubt that either physics calculations per second or graphics calculations
per second are linear plots against time. There is almost always memory
structure building and loading involved in programming so that only when
these structures are build can the plot be linear. Therefore I would expect
a doubling of instructions per second to give you more than a doubling in
fps from both the physics and graphics standpoint.
But MGPRS2 is a much better game on a P400/Voodoo2 than a P200/Voodoo1 - I
know I've got both machines. Yeah it runs fine on the 200 but you can't get
as good lap times as on the P400 because of the fps.
Paul

> : Then what kind of CPU time it taking? Anyone care to speculate?
> : Z.

> Have Papyrus send you the source, and put a step-through de*** on it.

by **Ben & Kare** » Sat, 24 Oct 1998 04:00:00

> Again, there is *no* option in GPL (or any *** simulations that I am
> aware of) to scale back the physics. Just the graphics,etc.

Perhaps not in racing sims, but scaleable physics is a relatively common feature in
*** flight sims. Longbow 2, for instance, allows you to turn on or off a number
of different physics options.

-Ben

by **Jeff Johnso** » Sat, 24 Oct 1998 04:00:00

First, no one was ***ing. I stated in the very first sentence that I liked
the sim. Second, from a hardware standpoint I'm not ***in either since
I'm running it on a Pentium II 400, ABIT BH6 motherboard, 64MB SDRAM w/ECC,
Hercules Thriller 8MB AGP, Obsidian2 X-24, Yamaha Waveforce 192XG PCI
Soundcard , Creative PC-DVD Dxr2 and TSW Steering Wheel.
I still say that I shouldn't need all of this to run it at nice
framerate (as per "Four Wheel Drift") and that Papyrus could have done a
better job in that respect. And as far as the AI cars goes, the fps aren't
that great when you're driving alone. And I definitely need some proof on
your statement about the AI cars using the same physics engine.
As far as I can see, the only one ***ing is you because someone isn't
as big a GPL and rec.autos.simulators zealot as you appear to be. I've got a
life.
As far as your eyecandy crack, I do drive all three :)

by **Skun** » Sun, 25 Oct 1998 04:00:00

not going to bother reading the rest of these msg's
this post is RIGHT
IF YOU DON'T AGREE.. YOU ARE WRONG

>Read all the posts on this newsgroup again concerning this topic.
>Then perhaps you'll be a bit more educated. But to give you a quick hint:

>ITS NOT THE 3D GRAPHICS THAT REQUIRE ALL THE CPU HORSEPOWER ITS THE
PHYSICS.

>Can we grasp that yet?

>F1RS and MGPRS2 do NOT have the physics that GPL does, period...so of
course
>you wont need as much machine to run the game. Start adding in the
multiple
>AI cars which acts just as adding another real driver to a game, so you
have
>all the physics (the exact same physics engine that acts on the player's
>car) acting on more and more AI cars and it takes a fast computer to handle
>it. But even if those two sims did YES THEY WOULD NEED the same hardware
>requirements.

>And as for "sucking the life out of the 3D cards"...um, WHAT? Lets see,
yes
>they supported 2....Rendition they had experience with and previous titles
>had support for it so their customer base had Rendition cards
>installed.....and when they started writing the game 3dfx cards were
>becoming very big. Not to mention that very few games actually use
>OpenGL...and D3D was really quite a chore to write in back again when they
>started over a year ago.

>If you want eyecandy..go drive F1RS or MGPRS2..if you want physics, dont
>*** about the steep requirements. Take the physics engine out of GPL and
>your system requirements plunge quite drastically.

by **Don Burnett** » Sun, 25 Oct 1998 04:00:00

I was wrong once, but found out later I was mistaken.

--
Don Burnette
Palmetto Racing
Dburn on Ten
AOLL Iroc Administrator

>not going to bother reading the rest of these msg's
>this post is RIGHT
>IF YOU DON'T AGREE.. YOU ARE WRONG

by **Jeff Johnso** » Sun, 25 Oct 1998 04:00:00

>On 23 Oct 1998 19:53:50 PDT,

>Are you an AI and vehicle dynamics simulation programmer?
>Oh, and who does vehicle physics better than Papyrus??

>Huh?

No I'm not and I'm positive you aren't either. If bothered to try to see
someone elses point of view you would have seen that's not what I feel the
problem is. I'm not saying the physics are bad. What I am saying is that
their coding of the entire sim leaves something to be desired in terms of
hardware requirements. As quite a few software vendors are.

I have no problems with my system in fact I'm getting 30+ fps with a full
field running at 640 x 480. But you're making my point, you shouldn't need
even a P2-333 to run this sim decently. I bet no one's getting 36fps with
Papyrus's recommended machine. And you really shouldn't have to have such a
machine to drive alone :)
To tell the truth, after about 6 sims, some of you guys seem to feel Papyrus
can do no wrong. Even though I have all of their sims and enjoy GPL, I
might add. I don't look at Papyrus as some infallible gods.

by **Jeff Johnso** » Sun, 25 Oct 1998 04:00:00

>On 24 Oct 1998 14:26:08 PDT,

>>To tell the truth, after about 6 sims, some of you guys seem to feel
Papyrus
>>can do no wrong.

>Who does it better?

So, I take that to mean that you think GPL and Papyrus are perfect and
cannot be criticised in any way shape or form.

rec.autos.simulators

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D

Papyrus knows physics but not 3D