In the test patch forum scawen mentioned he was going to try and up the racer and spector limits on servers for the Next Multiplayer incompatable. False starts are also on the may-be list if I remember correctly.

OT:
The things I am looking for before S2 Final are:

Propper Diffs
Propper Transmission/Clutch
More Fully Developed Damage
More Optimised & Brand Difference Tires
Aero Fixes (Undertray aftected by closeness to ground and relitive attitude)
Final LX and UFR Models
Final GTR Interiors
Class Ballance (I still maintain weight/power ratio and tire size and shape need to be equal)
AI
Flags & Penalties
Turning off Road Surface Smoothing (kill the splines and let us drive on a not glass smooth surface)

Quote from Becky Rose :

WR's and laptimes hold no rellevance to me in the slightest. All the respect I give to follow racers is earned when my wheel arches are gently brushing theirs.

I could not have said it better then Becky. Speed and records have mo meaning for me. I respect the drivers that I can race door to door with lap after lap and battle for position with constantly in a ever changing dance for position.

The drivers I respect most are the ones that show me level the respect that they want to recieve from me and that can race on any line no matter how far from ideal it may be.

Right now we are at version S2 0.6v8 Scawen generaly will change the last number with every little change(Compatable). He generaly will change the preceeding leter when all of the bugs are worked out with the latest added features so when all the bugs are worked out with the version S20.6vXX we will get a new base version of S2 0.6w.

Now when he releases the first version of the multiplayer patch it will likely be either 0.6w1 or 0.6x. I do not remember of if incriments the leter on a feature add. When the version goes final we will probaly be at verison 0.7.

Having been here for over 4 years I can agree with what everyone has said. Patience is a virtue you have to have in spades if you are a fan of this sim. The funny part is after 4 years of new things my favorite car is still the XRT.

Of all the cars in LFS the one I am having the most fun with right now and the only one I am interested in driving is the XRT. Thats one of the original 3. As for additional cars before S3 there are several possiblities. There is the DP1 if the devs decide to include it. We could get some other yet unknown real world or independant concept car, there is the seldom mentioned Rally pack and what ever cars it would bring if it gets done. There is also out in the design limbo the LX8 which was pulled in favor of the UF1 and UFR. Prior to the BF1 it was rumored to be the fastest car in LFS and was in its own class.

As for what version LFS is in as the mod said we are still in S2 Alpha. There are still features of this version that are not yet included in our version. Once all planded features are in LFS S2 will become a Beta and once all bugs are crushed and the version is stabe it will become S2 Final and we will have a long time of little contact with the devs as they work on what ever they plan to do for the next release version.

If you are bored with the cars and tracks now then you will be really bored in 1-2 years when we are waiting for the first release of S3.

I like it. If the profiler can do it I see no reason why it can not be done ingame and it would make it far easier.

I do see some issues with this Becky. Drivers in the pit lane always throw up a yellow flag for other drivers bases on their speed difference. Also at least 50% of the time the one that spins and throws the yellow flag after contact is the victom not the driver that caused the incident. A system like this also will flag drivers that take evasive action to avoid contact if they slow too much or get out of shape even if it was done to minimise disruption to the race. I do not think that throwing yellow flags is a good indication of the quality of a racer based on the current way yellow flags are generated.

It a shame this project did not have anough acitivity to stay in opperation. I would not mine lending some time to a project like this but I am just a driver, not an orginizer or web site wizard.

My guess is that the 22PSi remark is for boost pressure.

I dought it is related but who knows and it was damed odd. after installing V5 my DFP software would not see all the buttons on my contoler. It saw my DFP as a Wingman Pro. I had to get new controler software and after that is was like it never happened.

I have no idea how an update to LFS could cause that or if its even possible but I just wanted to see if it was just me. Personaly I have never had a problem with any other test patch.

Lest see... S1 was version 0.3. I started when it was version 0.1k I think back in late 2002. Found the forum some time later in Feb 2003, and have been racing ever since. I have no idea how many miles I had before S1. I know have 42k miles since they started keeping track. If that is since the begining then I am averaging 10k a year in virtual reality and 20k a year in reality. I think I am driving too much.

This could be very doable via a cuminity Insim application and private database. I woudl post the question in the programing section and see if anyone there thinks its doable.

Its still a rather simplistic AI, Whats makes them interesting is that unlike most AI in racing sims/games they only have the same controls as a human does. They also only know where the racing surface is not an optimal line. So they have to learn the tracks just like a human does.

They are no match for a fast racer. But they can be a great way to practice some race craft skills such as overtaking, close racing and learning less then ideal lines while trying to pass on the outside or searching for a new line.

Well for this to happen we first need real mirrors. Right now LFS just generates a Mirror view, the side mirrors are each half of the mirror view. Real reflective surfaces which are needed fro mirrors as far as I can tell are a real Power hog in DX8 and older engines. LFS does not use a DX9 engine and this would be a real performance hit.

I like the idea, but the base graphics engine in LFS is not yet up to the challange to do it well.

From what I have seen and I have played this alot they will drive the cars in to the ground ignoring tire temps, wear, damage to a degree and fuel.

Your best bet for the AI and their ability to race is to give them a setup that is forgiving and easy on tires. Also having them use different sets, and running in different cars will help them improve. And yes the AI can be great at drifting as well.

They also have to be taught to pass, overtake and draft but they will learn. They also learn much faster when driving with other people or AI. They learn from thier own experance and the other cars on the track. If you run with them and constanty drive a different line then they do they will eventualy alter their line to be closer to yours, if its faster. Thier ability to adapt and learn is rather suprising.

This is why I buy older cars with simplistic computers. No OBDII for me.

I personaly do not feel that with todays technology we can simulate a morotcycle in such a way for it to feel as natural in the simulation as a car currently can feel in a simulation. So much of the bike experance is the physical sensation and the way the g-forces play on you and the bike. Sure its posible to simulate but as a biker I woudl never use them in LFS.

Hyperactive,

I looked at the design you came up with and that is a nice slick way to do it. It also adds some other nice and useful graphic displays for things like the HP/Torque curves, grearing and a notes section. I really like the concept. It have a more polished and usfull look then we have now.

Why should the devs have to cover the extra cost of storing and distibuting larger high res skins to the general public. Until LFS has a built in P2P file sharing system which each driver can configure I do not see a reason to force the devs to pay the cost for an option like this. If you really want High Resolution skins. Build an insim app and a skins repository to check for high res skins and download the them if they are in the repositiry. You want high res skins then you front the cost.

I think there are a lot of topics getting confused in this thread. I have now seen references to engine braking, scavenging (Only happens in N/A engines) and some other oddly related topics.

As others have said until we get a more complex engine physics model some of these thiong are just not possible to add in to the simulation yet and things like scavenging, electromicly controled engne braking reduction, and the like are just not likely to happen as the physics they need to work are not in the physics model yet or are just placeholder pieces.

Some how I do not see LFS becoming so compled as to accualy model the physics of how an engine works, how the transmission, shifting and clutch work and some of the other rather complex componantes of the cars. Even though we would love to see such complecity.

Personaly I would love to see LFS become a truely complex vehicle simulation which works as closely to how reality does as can be done on a modern PC. But then it may detract for what makes LFS as great as it is, which is being a very fun, accesable and fairly realistic racing simulation.

There are other projects that are striving for more extreme accutacy of vehicle physics. They are targeting a different crowd then LFS is though. If LFS can get the vehicles behaving a little more like real cars, get the tire feel closer and build the damage system in to a nice complex componant then I think LFS will have achieved its intended and correct level of realisim and retain it fun factor.

nKPro, i-Racing and Drivers Republic are all trying to more accurate vehicle simulations, but none will be as accessable or easy to use and the racing will not be as fun for the casual racers like LFS is. LFS will always be the sim of choice for me when I just want to have fun and a good race. I will look elseware when I want a real experance.

I think the point Cue-Bal is trying to make is that as it is right now the level of adjustability is so high that the driver is not forced to make as many choices in the setup. This leads to a situation where everyone runs bases their sets of the WR sets. And setups can be so fine tuned that they drive exactly like the driver wants.

This has a net result of allowing those people with very good setup skills to create a car that is perfict for them while those that don't have such finely honed skills are left with having to addapt to a not ideal setup. The end result is races that are not very close and competitive which is the point of a sim like this. To have close and fun races.

By limiting setup options to more realistic values and number of choices drivers now will all have to make compromises and adapt to the limitatios of the setup options, this also brings the racing closer togeather and makes the racing more about your racing abilities and less about your setup. This is a good thing.

I do not see a problem with very granular setup option for the suspension, wheels, brakes and tires. I do think that the transmissions and diffs need to be modeled and their setup options adjucted to be in line with what is really posible in a real car. I just spent some looking and most of the gear boxes under 10k come with a choice of a few ratio sets and rear diffs come in with 5-10 ratio choices from most of the racing gearbox vendors I could find.

The only car in LFS that woudl use a box where indifidual gears can be altered indipendantly is the BF1 and even then you might have no more then 10 choices for ratios of any 1 gear. We do not need transmission and final drive ratio adjustability down to the 1/10000th. Model a real transmission with real teeth and let us adjust within real limitations.

I have on many occasions slid the rear tires of the XRT on approach to T1 on Blackwood due to the engine braking being more then the tires have grip to handle. Its there, its noticable but with out the seat of your pants g-forces you just can not sence it well.

I don't even want to get in to the mistakes and assumptions you make reguarding how engine braking works and how you think its adjustable.

Last edited by Gimpster, Fri 23 Mar 2007, 19:46.

Ok what I wrote before did not make any useful addition to this thread so I am re-writing my thoughs.

As far as the level aof adjustability and the costs infolved for a street car here is what I know. My car, a miata, from the factory has the ability to adjust the camber and toe. on the wheels. There are 2-3 final drive ratios that were avalable and depending on the year of transmission used there maybe three differnt ratios that were used on the 5 speed boxes. The only differences were in the first two gears and the difference were very minor.

For few thousand dollars I can get a fully adjustble suspension system which 20+ levels of damping adjustment + preload, several ARB choices and a few sets of springs. There are also a few choiced for aftermarket gearsets for the transmission that are more optimised for the power curve of the engine but each is over 3k.

Thats a lot of adjustment but is no where near what is abalable in LFS, Caster, Ackeman, Brake Bias & Pressure are not adjustable. Final drive is limited to 3 choices, an gear box ratios might as well be limited to 2 predefigned sets. The number of spring rates is also very limited as are the ARB choices. And thats all I have to work with. Everything become a ballancing act and the driver becomes more of the determining factor.

LFS allows adjustment in gear ratios down to 0.0001, and do do that you need like 10,000 teeth on the gear, you might be lucky if you have 100 in a real transmission. The number of teeth & there pitch determin the number of choices that exist. In LFS it is as if instead of gears we have friction wheels which we can very in very very small imcrements to optimise ratios and this is not in any way close to reality.

Patch W will be a multiplayer update as Scawen mentioned, avter that will be the incompatable. So the earliest I can see a fix to these issues is Patch X. As S2 has been said to be about the Cars this is the right time to take this step. Its time to recode the vehicle setup process and options, it time to move from infinate adjustbility to realistic values and number of choices. Its time to make it realistic.

Last edited by Gimpster, Fri 23 Mar 2007, 17:05.

But it would still require additional logic to be programed in to the existing replay function to allow for the posibility of the "Ballast" paramater not being there and to then assign a value of zero for that missing paramter. Seems like a lot of extra work to me just to slip in a new feature that could just as easily wait until the next physics/feature update, so as to not have to add extra code that would be made un-nessasary once the next non-compatable version is released.

Thats my point, why do it at all if the work is only going to be needed until the following patch?

Quote from AndroidXP :

Where do you have this knowledge from? Did Scawen tell you? Did you have a look at the code? Or is it just an assumption?

If deciding in a replay that "value not there" = "value is 0" is so hard (it is not), how come that replays were still compatible after Scawen strengthened the suspension of the BF1, which was actually really a physics change? I can even recall him saying, that upon loading the replay he was checking if he should let the physics engine run with strong or weak suspension, to stay compatible with replays.

Sorry, but I think you're pulling this out of your butt.

I dare you to find me two versions where there is a feature difference that affects the cars where an older client can connect to the more updated host which has the new feature.

You will not find one.

Regurading the change to the BF1's suspension. That was not the case of a missing paramiter. The value of a paramiter changed. The replay only tracks user input and then then recreats the replay from those inputs and the current game physics and vehicle paramiters. Altering the threshold of when damage occures and how much damage occures if the threshold is passed would not invalidate the replay but whould display slightly different as the physics applied to the repay would differ. It was not deamed in-compatable as it affected a single car, was done right after the car was introduced and was a relitivly small change to two paramiters.

Thats a far cry from adding a new paramiter.

Thats the point AndroidXP, LFS doen not have the code to assume a missing paramiter means that the parameter should return 0. The base engne is not written that way. That would be even more new code needed just to allow the newer client to reconize this state and not error out when it does not find an expected paramiter.

LFS has not been desined to allow for clients of differing versions to interact. Part of that is cheat pertection, part is code efficiency. For the engine to allow for clients of differing versions to play togeather it would need to be have a check and logic function for each possible feature/paramiter differece and then automaticly default to the feature set of the lowest common dinominator(only allow features/paramitors which all clients have in comon).

While very doable the cost in terms of code complexity is extreamly high and its not very often done as the need for such adaptability is often not worth the cost and its far easier to control by forcing client version compatability.