Quote from Scawen :

Yes! I can reproduce it that way too!

I can simplify the reproduction method slightly :

1) Start LFS (windowed or full screen, doesn't matter)
2) If you started windowed, go to full screen
3) Window while connecting : set to ON
4) Click for a list of games (LFS goes to window - now you can stop the list)
5) Press back (LFS goes back to fullscreen)
6) Now, SHIFT+F4 to window (LFS still OK in window)
7) SHIFT+F4 back to full screen (Now you have a black screen)

Strangely enough, automatic windowing must be switched on. If I just start LFS and keep pressing SHIFT+F4, it doesn't go wrong. Apparently, LFS must have at one point, automatically gone full screen (i.e. coming out of an auto-windowed state). After that, you will never successfully go to full screen from windowed again.

[ EDIT : I'd like to know if this doesn't cause the bug for anyone, or if you can get it if LFS hasn't auto-returned-to-full-screen at some point ]

No black screen for me. XP, U23, 6600 GT AGP, 84.43 drivers.

Quote from tristancliffe :

Of course, if your clutch has gone and you come to a stop there's not much you'll do about it...

Just keep it in gear and use the starter motor. This combined with clutchless shifting has got me home twice already when my clutch cable broke. The shifting part does work better on some cars than on others. On my Renault 5 from third gear upwards was so easy that pressing the clutch was just a waste of time and effort.

Set "Port 80 0" to "Port 80 1" in cfg.txt.

Maybe Ferrari can borrow Daniel Pedrosa for a race.

Has anyone noticed that situation is exactly the same as Suzuka '96? If Villeneuve would have won Hill would have needed exactly one point to take the title. In the race Hill disappered into the distance and Villeneuve retired. Good times.

Lets hope Alonso does the same.

I had the same problem a couple of years ago. The only port we had open was 8080 TCP. After fiddling around with lots of tools I managed to unlock and play LFS online. When LFS started using UDP packets for online play it was all over though. But maybe the tools have improved since then so you might have a chance.

Anyway these are the names of the tools that are still in my "Tunnelling" folder on my HD:

HTTP-Tunnel
Socks2Http
SocksCap
SSHProxy

I think the Socks thingy was for filesharing not LFS though. Good luck!

Quote from Hyperactive :

I have always thought that LFS is the purest example of how some things are more important than others. Thing like tire physics will always be the number 1 of LFS. Every incompatible patch we get has some tire physics updates, and I can see most of us agreeing with that. Car handling characteristics are and always will be the number 1 point of LFS. Online part is the 2nd. I guess that's why we the other stuff so unfinished at this stage. With tires we can overheat them, flatspot them, blow them (oops) etc. But some other stuff like engine damage or proper race procedures are very basic.

LFS is defenately a wip, but it does its core stuff 1000x better than anything else. And that is physics.

As said earlier, realism is a subjective thing. On the other side we have people who want ghost cars and on the other side we have people who could only take ghost car if it was used in real life racing. There is a lot of people in the middle too, some closer to other side than others. The extreme ends tend to be the loudest ones.

The first group thinks that "it is a computer program, making a ghost car is not something very hard and there are (or can be) some benefits gained of using using it. For this group LFS is better with ghost car. The second group sees LFS as a computer program that aims to be as close to real life as possible as it can be on a computer screen. With this group every suggestion and feature request goes through the "is it used in real life? / how is it used? /where is it used etc." and if the suggestion has valid it gets the OK.

I see myself as 80% part of the second group and 20% part of the 1st group. I want LFS to simulate real life racing as closely as possible, but I don't want to wait hours to get my cars fixed. With this belief I am, of course, pursuing my own interests when I see a ghost car topic on the forum. I don't see ghost car as part of racing sim, because ghost car mods aren't used in real life. I understand that ghost cars are not probably used in real life because of technical limitations but imho that doesn't mean that we could have it in a sim, just because it is possible and could be useful.

That sums it up very nicely. :up:

Things can always be made more realistic.

Some people would like realistic times for setup changes. How about first having to drive from the hotel to the paddock, wait around there, go to the pre-grid, wait around again, go to the grid, drive the warm-up lap, drive the race and then not drive on the same track again for a whole year?

It's not unrealistic but it doesn't make sense for a simulator. But it would be pointless to hope that all simracers could in agreement about which features cross the line and which don't.

Quote from TagForce :

Dangerous statement... If he gets enough people to state the same, then he'd just say 'no', get your money, and pick up racing again.

LOL. I had to read that a few times to get what you were saying.

Quote from Leifde :

@Niki - why did you stop racing?

My money is on no money.

Quote from Niels Heusinkveld :

I looked at the 'nap mod' physics.. "New Advanced Physics"..

Well.. I looked at the lateral tyre curve.. It isn't advanced there.. The grip peak is the same as 90% of all ISI curves, peaking at 6.2 degrees. If you look at the graph there is less dropoff which seems more right than most ISI curves, but calling that advanced is not right..

However, the curve has two peaks, which is strange..

However, the creator doesn't seem to understand how the lateral 'step' size works in ISI (it takes the SIN of the value) so the SIN can't be bigger than 1 (mathematic limit) ASIN 1 = 90 degrees and surely a lateral grip curve can't exceed 90 degrees.. you can't have 'more lateral angle' than 90 degrees, as that is as sideways as the tyre can work!

So what happens, if you translate the lateral curve to DEGREES, you see the peak at 6.2degs (as all isi sims, just about) then dropping to somewhere between 80 and 90% at 90 degrees slip... Now this second peak in the curve, happens way beyond 90 degrees slip and is thus disregarded by the sim as it can't take the SIN of >1!!

Now I ask you.. Someone makes 'new advanced' physics, somehow introducing a second peak in the lateral tyre curve, driving, and believing it is the best thing.... not knowing that this whole thing is disregarded as its beyond 90 degrees?? the A in NAP sounds more like Amateur..

The result is not that good either, its still ISI.. you can make it somewhat better with the grip curves that have less drop off but things are still weird. Snap back is VERY violent and the drive isn't convincing imo..

In the simracing / modding scene the quality of tracks and cars is graphically often substandard.. but the physics stuff is often far far worse! :/

Interesting stuff. I especially like the violent ISI snap-back. If you ask the RSC crowd why you never see cars spinning off in the opposite direction of the corner in real life you're told it's because all real racing drivers are the equivalent to gods in terms of driving skills. And because they never go to the limit because they are in constant fear of getting killed to death. lol.

I don't see the LOL in serious kart crashes.

Whoa! 1920x1080p@60fps. Now that's what I call HD. Are there any screens around that can do that? Shame there is no mention of any physics changes as the racecar physics were quite poor in GT4, road cars were fun. And the idea of racing against AI can't really get me exited either although they did mention something about online beta testing going on for future GT games.

First of all: Good Luck Victor in whatever you may be doing in future. Keep us informed.

About LFS development: I think one day the LFS guys will move on to something else when they feel they have spent enough time of their lives with LFS. Or they may even be caught by new sims that have more manpower and more resources but ATM I think iRacing are the only ones that may be able to do this.

Call me a fanboi but for me LFS has had better physics than any other game from 0.04K to S2U. That's over 4 years. Talented people have tried but they couldn't keep up the pace. So who knows how long this will go on?

The ASS article made me download the current rFactor demo and have a look for myself.

First of all the default settings of rFactor are really a disgrace: wtf should my steering be speed sensitive? Why is the dfp preset set to 360°? Why does the Sauber steering wheel only turn 260°?

Well anyway after sorting these things out I went to the track. The first thing I noticed is that, as expected, the car was undrivable without TC. It wasn't quite as bad as I had expected in terms of instability but in the first two gears you really have no chance to catch the car using opposite lock.

Then I put on medium TC and I found that driving was actually quite fun. Still a bit hard but what I noticed is that driving on real tracks is really a very enjoyable thing to do for me, and, as I have often stated, is one of the major shortcomings of LFS.

Another feature I liked was the auto-FPS setting, which adjusts graphics detail to keep a decent frame-rate. Of course LFS doesn't need this but it was a hell of a lot better than Simbin games, which variate from 10 to 100 fps in a single lap.

Then I started LFS to compare. I turned off the TC and and went to Blackwood. As expected I felt more in control of the car and it wasn't quite as hard to catch slides but the main impression I had was that rFactor and LFS were a lot more similar than I would have thought. One thing I tried were the Alonso style slides that he does on warm-up laps or on start-finish on the last lap of Suzuka. In both games it was not easy and the car would get out of control after a couple of left-right maneuvers. Of course I don't believe that I have Fernando's car control but I was still surprised how similar LFS and rFactor reacted.

All in all I would say that LFS is more believable to me though. It is slightly easier than rFactor and considering how seldom it is for F1 drivers to lose control of their cars (even n00bs like Vettel) I think this is more realistic.

And the ASS article was not very convincing. Examples:

Quote from ASS :

ISI’s BMW, in particular, is much more
‘critical’ to drive, and the data confirms what we ‘felt’ was he
case: That the ISI BMW is far more nervous than its LFS twin.

Since when do F1 cars look nervous?

Quote from ASS :

… it offers more grip, and, subjectively of
course, it responds more like what our imagination thinks a
800BHP 600kg racing single seater would do in real-life.

Too subjective to be of any value.

Quote from ASS :

For instance, both cars seem to offer way too much
grip—3.5 lateral G at 183km/h seems, at least to us, to be
a little optimistic. But we may be wrong.

Either you have data to compare or you don't. This statement is pointless.

Another interesting quote in ASS was that of Doug Arnao, a person who is well respected by the ISI/RSC/ASS crowd, who states that he thinks the rFactor F1 tyres have too much grip drop-off with slip angle, as Todd Wasson and others have been telling us for ages.

Quote from Hyperactive :

It is really far more complex than that. Basically you have got it correct but you are missing a lot of stuff that makes the tires handle like they do in real life.

Yes, but maybe we can still get a basic idea of what's going on by makeing some simple assumptions?

Quote from Shotglass :

but a different kind not the longitudinal weight shift induced type but the suddenly the front wheels are going where theyre pointing at type which is probaly more of a feint/scandinavian flick type of car behaviour

Good point.

Well basically whenever you change the longitudal force at the front wheels by either braking or getting on or off the throttle (FWD) two things happen:

1) weight is shifted between front and rear
2) the reduction (or increase) of longitudal force will lead to an increase (or reduction) of lateral force potential.

Now these two phenomena oppose each other: when you brake 1) will increase the potential front cornering force and 2) will decrease it. So the question is how do 1) and 2) relate to each other and does one of them always "win" with real cars and real tyres? After giving it a bit of thought I think that wheelbase, CoG height and load sensitivity of μ should be important factors.

Hmm, maybe I'll do a few simple calculations to try and get a clearer picture. Of course experience would be the most significant information in this case so I trust Tristan may be right, considering he has driven quite a few nice cars.

Last edited by J.B., Tue, 10 Oct 2006 00:02.

I'd say it depends on many factors such as brake balance, tyre properties, speed and force of the "jab" and setup.

For instance we all know that FWD's like to oversteer when lifting off the throttly suddenly. There is no reason why the same thing cannot be achieved using the brake pedal. After all, the tyres don't know if you're using the brakes to apply longitudal force to the tyres or if you only lifted off the throttle quickly. If you hit the brake very suddenly and hard OTOH you will probably go off in a straight line.

In LFS trail-braking seems to be more likely to induce understeer than oversteer, which was for example highlighted by the brake-off-oversteer which was very pronounced in the old S2 Demo version.

Personally I don't think this is quite right but who knows? Without a lot of experience with different types of cars and especially tyres it's hard to tell if this can be realistic or not.

If you're faster the choice is yours: Leave the race or overtake. Staying behind a slower driver doesn't make sense.

Quote from Vain :

That's what I thought too. Must be annoying for Alonso to only win the championship because of MSC's blown engine. Now all his driving and effort is worthless, because everyone will say that the reason for his championship is this one blown engine. Poor guy.

Vain

I trust that was tounge in cheek?

Just comparing this season, MS has retired 3 times (AUS, HUN, JPN), twice by his own mistake and once for technical reasons. FA has retired twice (HUN, ITA), both times for technical reasons. I can't see any advantage for FA in terms of luck there. And that's without even getting into the whole mass damper and Massa blocking debate.

In fact no driver ever ever has had the reliability that MS has had at Ferarri. Does this make people remember him as the guy who won all those championships by luck?

And anyway, let's wait until the season's over, anything can still happen.

You shouldn't have posted that video. My bank account will not appreciate the consequences.

Quote from Me. :

Those pics don't look too bad. So now all we need is a program that converts this data into a track. And a track-moddable sim that's worth playing, of course.

...

Those pics don't look too bad. So now all we need is a program that converts this data into a track. And a track-moddable sim that's worth playing, of course.

Love the idea. What kind of hardware would I need for this? Is GPS without fancy error correcting technology accurate and fast enough? To create a track a simple lateral G sensor would probably be better but it would somehow have to be hooked up to the speedo too. No idea if this kind of simple data recording is available for roadcars.

Come to think of it, modern cars with electronic stability systems probably have the sensors built in anyway. You'd "just" need to create an interface for the recording.