I'm happy to see that the tire physic is going stronger in LFS, but I'm worring on something

if I'm right, you are fall in a hole of "update > more problem > update the tire > more problem > update the tire > more problem > update the tire > more problem".........

but... could you sure that the problem is on the tire?

I'm not very sure what you did, but as what I see in S2, there are no unsprung weight effect, the wheel and suspension are always absort all sprung on road, but it is not the true, the wheel & suspension have their own weight & mass, their inertia during they moving are also need to calcute. I hope that you can at leass sure that the problem now you face are not from that, or you will just find more problem after you make the tire more perfect

Prepare to be flamed for your insolence , LFS supports unsprung mass.
See here
I prefer to wait and see for the tire physics update.

Not quite correct, Mr Bogey. The unsprung mass is calculated but LFS uses a single rigid body model, so there mass for the wheels does not move separately in the physics engine. Likewise the inertia for the body always includes the wheels, rather than being separate. Scawen has then done some clever trickery in order to compensate for this (adjusting forces and such) so that spring rates work out, and unsprung mass does have largely the correct effect. I don't know the technical difference that going to a multi-body physics engine would provide.

Shouldn't it be a rule, that the more parts, the more to go wrong with? I mean that would be alot of work and testing to be done. I agree, it could be more realistic, but then again much could go wrong

I am now disillusioned. Can we deduce that supporting multi-bodies calculation would take too much CPU resources?
Do you think that this may take part of the tire physics update? (no sarcasm here)

The reason is unknown but it was rumored before that Scawen internally complained that the update already would be available if there was more CPU power.

I understand the wish to keep CPU power consumption as low as possible but my i7 with 4(x2HT) cores is looking pretty bored

Quote from cargame.nl :

I understand the wish to keep CPU power consumption as low as possible but my i7 with 4(x2HT) cores is looking pretty bored

That wish makes no sense whatsoever when you want cutting edge physics Scawen, please

Of course it makes sense. You want it to be playable for an average amount of people not only for the computer geek with the latest hardcore computer shit.

Quote from Bogey Jammer :

Can we deduce that supporting multi-bodies calculation would take too much CPU resources?

Not much more, I would have thought. Although more realistic, it does introduce it's own problems. I've not played with a multi-body engine first hand, however.

All this technical talk has lost me completley, but if im correct the tyres are nearly complete, but with a few problems. I know im gonna get flamed for this but maybe come off perfection a little and get the issues sorted and release an update, then for the following update could be perfection of the tyre model. This way you could keep the community happy and not have as much pressure on you to force out an update. Because in theory pressure produces problems.

I doubt Scawen feels particularly pressured. Tyres are by far the least understood part of a vehicle, and also the most important to get right. This is why it is worth spending the time on.

Aerodynamics are probably next most complicated but those effects are, for the majority of cars, not so significant. LFS could definately benefit from improvement in that area though.

Pretty much any other area of vehicle simulation is fairly straightforward, however, so it's usually just a case of making it right using known formulae. The same cannot be said of tyres and aerodynamics, although there are many known simple approximations of the two, all approximations have their own weaknesses. If Scawen can pull this off, I for one would be interested in licencing his tyre model, if it can be written in a way it can plug in to other vehicle dynamics engines (I doubt it's being written in exactly that way know, but I'm pretty sure Scawen has componentised his vehicle engine to some degree).

Quote from cargame.nl :

Of course it makes sense. You want it to be playable for an average amount of people not only for the computer geek with the latest hardcore computer shit.

Still makes no sense. The reason it's such a problem to develop is because we need to do an approximation if the simulation needs to run in realtime on old hardware. The more power we have available, the less simplified the approximation needs to be, and the more accurate it will be at the same time.

To me, it seems that if you want cutting-edge physics, you should abandon old hardware. I'm not saying up the minimum requirements to the best quad core processors etc.
I believe the calculations are quite memory and CPU intensive and would benefit majorly from parallelisation, so a processor with 2 cores or just 1 core with HT could help majorly if the code was optimized for it (multithreaded).

Have you even checked how little it takes to run LFS currently? It's possible to attain 30 fps with a 1.2 GHz Celeron, 512 mb slow RAM and an onboard video card with 32 mb memory. That was slow 10 years ago..

hi fellars and dolls

pls try out GPL, an old game, nearly 2 old, so i had probs to make it run on my new pc
but it runs on evry historic pc. the grafics are awful (4 2day), the sounds an epic poem, the physics, especially the feeling on the wheel are quiet insane (my o.)

for ex.:
-the front of this old massive RWD-formel cars fells lighter by kick the gas and then the sterring does also. .
(real behave?!- don't know, nevr driven one, but it feels good)
-or the mass effect in changing corners (not bad in LFS also)
-or the slippy reactions and feelings when driving in limits range

and on the other hands there are games like n4s shift, f1 2011, which stressed the most, even newer pc's- but- don't feel very real or plastic or ductile . . .
think about it

i think these autonomous and brave guys try and will make a good job. in the past they did. forum lfs world etc. speaks 4 itself we will see . .

out of my buero racing seat
nille

The simple answer for that is, NFS and similar games don't TRY to simulate tyre physics. They use computer power (mostly GPU) for graphics.
GPL is quite awesome. LFS is already better IMO though

Quote from RasmusL :

Have you even checked how little it takes to run LFS currently? It's possible to attain 30 fps with a 1.2 GHz Celeron, 512 mb slow RAM and an onboard video card with 32 mb memory. That was slow 10 years ago..

No..

Thats not possible, you are exaggerating .. Celeron yes but you need at least some GPU otherwise it's not playable. Yeah on 640x480 maybe.

Quote from nille[ger] :

-the front of this old massive RWD-formel cars fells lighter by kick the gas and then the sterring does also. .

same happens in lfs when you drive a car like the lx4 or lx6

Quote from cargame.nl :

No..

Thats not possible, you are exaggerating .. Celeron yes but you need at least some GPU otherwise it's not playable. Yeah on 640x480 maybe.

Alright, I had no actual numbers but estimated, it wasn't too far from true. Lets get some proof from posts on this forum (idk how to quote from another topic properly :shrug.

Quote :

Okay, so I installed windows 98, but I also added a couple of upgrades to the slower computer. I swapped the ATi Rage 128 16mb video card for an ASUS V7700Deluxe 32mb video card. I also added a 256mb stick of ram for a total of 512mb. I apologize for adding two more variables if anyone is using this for a reference. The FPS now ranges from 14-20. This still may not sound great but, LFS looks pretty nice, and is more than what I would call "playable".

Quote :

Intel Celeron CPU 1000 mhz
256 mb RAM
Trident Blade 3D/ProMedia 8mb (onboard)

fps - 18-28

Quote :

I have a P III 660 Mhz, with 196 MB RAM and some shitty on board graphics (must be 4mb) i guess.Well for me my sytem delivers about 10-20 fps

Certainly these are on the lowest settings possible, but discounting graphics this is what's required to run LFS, albeit a bit slowly. It's time for an upgrade.

Quote from RasmusL :

The simple answer for that is, NFS and similar games don't TRY to simulate tyre physics. They use computer power (mostly GPU) for graphics.
GPL is quite awesome. LFS is already better IMO though

mmh, okay u re right . .
i was in hurry so i missd to point it exactly.
i meand that GPL can do that without x-tra large CPU
+ these "eyecatchingracinggames" does use a lot of cpu
(without special physics)

by the way- is dualcore, multicore etc useful for lfs ??

Quote from Shotglass :

same happens in lfs when you drive a car like the lx4 or lx6

sure- i nevr felt that. maybe i need a better wheel than ^^

Quote from cargame.nl :

The reason is unknown but it was rumored before that Scawen internally complained that the update already would be available if there was more CPU power.

I understand the wish to keep CPU power consumption as low as possible but my i7 with 4(x2HT) cores is looking pretty bored

hey throw me some dough please.


Anyways, my PC cant handle LFS as it is that well, so why make more PC consumption

Quote from Bmxtwins :

hey throw me some dough please.


Anyways, my PC cant handle LFS as it is that well, so why make more PC consumption

Try upgrading? Let's stay in the dark ages because some spotty kid in America has a shit PC.

Steam platform uses polls often to know the global customer's pc performances status.

Why not do the same here? it would give the devs a "go ahead" for new physics features, or not... :twocents:

I have an idea too: Let Scawen do whatever he wants. It's his creation. He knows better how to proceed with LFS development than any of us do

Quote from pearcy_2k7 :

Try upgrading? Let's stay in the dark ages because some spotty kid in America has a shit PC.

ooh that's a good idea

the pc performance doesn't cares, i think. some guys give 100's of bucks
for cockpit, wheel etc. - then some new pc parts shoudn't mess it. .

i only want
-even more physics as now
.
.

- more grafics would be nice

Quote from Bob Smith :

I've not played with a multi-body engine first hand, however.

So you never played NetKar Pro or rFactor?

I suspect his meaning of "played with" meant something a bit more than that