Is your system overclocked? It sounds like either an overclocked processor/video card or a hardware failure.
While I think the current team is doing a bang-up job, I have to agree that certain parts of the sim might be helped by adding a few people. That doesn't necessarily mean someone coding with Scawen, but it does mean people contributing to the overall project.
For instance, there are a lot of talented modelers who play LFS and games like Racer. I think that most of them would kill to get one of their car models into LFS. And, similarly, many of us would kill to have those cars in LFS. For example, what if the guys who did the updated car models for GPL were given the chance to make a 60's F1 car for LFS? Would anyone NOT want to see that in the game? Or how about if a really great LMP car model were available to be put in LFS? Wouldn't we want to see that? Similarly, look at the community projects such as Electrik's textures and Dave's sound packs. These are the kinds of things that the LFS devs should be leveraging. They could easily use the work of these community members and it wouldn't cost them a dime or take much of their time. The game would look better, sound better, and have more variety. The only thing they'd really have to do is add a couple of names to the credits page. The content creators are happy, the users are happy, and the devs have more free time to work on other things. It's a win every way you look at it.
Let Scawen continue his physics work, as it's coming along quite well. Let Eric continue building and refining the tracks. Let Victor continue to maintain and improve the web sites. But take some of the talent that this community has and put it to work! We'll all be thankful in the end.
For instance, there are a lot of talented modelers who play LFS and games like Racer. I think that most of them would kill to get one of their car models into LFS. And, similarly, many of us would kill to have those cars in LFS. For example, what if the guys who did the updated car models for GPL were given the chance to make a 60's F1 car for LFS? Would anyone NOT want to see that in the game? Or how about if a really great LMP car model were available to be put in LFS? Wouldn't we want to see that? Similarly, look at the community projects such as Electrik's textures and Dave's sound packs. These are the kinds of things that the LFS devs should be leveraging. They could easily use the work of these community members and it wouldn't cost them a dime or take much of their time. The game would look better, sound better, and have more variety. The only thing they'd really have to do is add a couple of names to the credits page. The content creators are happy, the users are happy, and the devs have more free time to work on other things. It's a win every way you look at it.
Let Scawen continue his physics work, as it's coming along quite well. Let Eric continue building and refining the tracks. Let Victor continue to maintain and improve the web sites. But take some of the talent that this community has and put it to work! We'll all be thankful in the end.
I don't think you have to lift/blip in either of those cars. They use tiptronic-type systems which handle throttle control automatically. Most cars that have paddles on the wheel use tiptronic or DSG type gearboxes and manage rev matching by computer. I think most of them also have systems that prevent you from shifting down too early.
In LFS the BF1, FXR, and XRR have blipping controlled by the car. The FOX, the MRT, and the new BMW use "dumb" gearboxes which require the driver to take care of the throttle, and which use a shifter - not wheel mounted paddles.
These are two totally different types of systems, but for some reason people seem to think that since their DFP has paddles, the shifting in LFS should work like a Ferrari Enzo.
You can't use an H-shift with a sequential car because H-shift uses 1-2-3-4-5-6 while sequential uses UP-DOWN. If you have a G25, just turn the knob to set it to sequential and make sure that your script for the FXR and XRR are set to use the sequential settings ("set sequential" in the script).
Adjustable mirrors are easier said than done. LFS already has to dedicate a lot of cycles to doing the mirrors that we have now, as imperfect as they are. Rendering even more views would have a huge impact on frame rate and make the game unplayable for anyone without a very fast machine and video card.
This has been covered before, I think in the Improvement Suggestions section. You might be able to find one of the old threads with the details through a search, if you care to.
Nobody knows since the demo doesn't have the XRT anymore.
I think that the current behavior is very raceable, but it's still very incorrect and will have negative side effects. There's no reason to settle for the way it is now if Scawen can tweak it a bit to be more realistic and natural.
The grip of the clutch should be stronger than the grip of the tires or the power of the engine. We can still keep the clutch overheating when people slip it too much, but the basic behavior really does need to be fixed.
Can you run a few laps and upload a replay? You've GOT to be doing something wrong. But without even seeing you drive, my guess is that you are not lifting off the throttle when shifting up, or blipping the throttle when shifting down. There's also a post in the main forum about how to shift each of the cars properly. That might help you. But you NEED to lift/blip during shifts. Not doing so will fry the clutch.
It seems to me as though the clutch heat was put in to prevent people from driving unrealistically, and to punish those who are hard on the car. Personally, I would rather see the clutch more strong and lifelike, and punish bad shifting in the same way that it's punished in real life - locking tires and car instability.
Go practice your shifting technique, which should make you faster and make your car last longer. Then also cross your fingers that X32 has fixed clutch behavior.
While I do believe that there is something fundamentally wrong with the current clutch modeling, I really don't see how people are having so much trouble with this. I raced last night and was able to finish 33 laps around Westhill and 25 laps around South City Classic in road cars with H-shifters. In both cases I finished the race with barely any heat at all in the clutch.
I honestly don't know how or why people are having such trouble with clutch overheating. Seriously - if you can't finish a race without burning up the clutch, even with the currently flawed modeling, you need to take a good hard look at your driving style.
No offense, but if you can't figure out how to apply the test patch, you should probably just wait a few days until Patch Y is released.
I have a feeling (though it may just be wishful thinking) that what you describe is on the agenda for S2. Blackwood now has lights and marshals all around the track. I'm guessing that eventually those will be used instead of the test messages.
Even if they implemented 2D sprites waving the flags (ala GPL), that would be an improvement, and enough to make most people happy.
How are we supposed to know what garage is assigned to us if there's no qualifying enabled on the server, since you start on the grid and not in the pits.
Also, even if you DO start in the pits, the stalls all look the same and the numbering is quite difficult to see on your way out.
I'll have to run a test and check, because I'm fairly certain that the AI don't exit, pit, and return to the same stall throughout a race. It doesn't seem to be consistent.
Edit: Okay, I've done some testing on this, and here are the results I've seen. AI cars, after the race, enter the pit stall that corresponds to their number. ie: AI1 will stop in the first stall, AI2 in the second, and so forth. For some reason, the later AI cars have a hard time getting into their stalls regardless of their finishing order. For instance, on a race around SO Classic with 8 AI LX4s, cars 1-5 successfully entered a stall after the race, but cars 6, 7, and 8 all failed and did a drive through of the pits, stopping at the end of pit late. Car #7 tried to enter the stall, but hit the wall. He then backed up (hitting the wall behind him), then finally gave up and exited pit lane, plowing into the other two cars that were already sitting at the end.
The pits that the cars use for refueling do not seem to correspond to the car number, the position in the race, or anything else that I can determine. In the race I tried, AI#1 who was running in first place pitted in the 9th pit box (out of 16 boxes total). AI#2, running in second place, pitted in the 12th pit box. AI#7, running in 6th place, pitted in box #5. This is made even more complex by the fact that two stalls share one pit spot. For instance, South City has 16 places to pit, but 32 stalls/garages. Because all of the AI pit at the same time, there are not enough pit spaces for them all. I haven't tried running with >16 cars yet, but I'm going to give that a try. I'm sure it's going to cause havoc.
In essence, it seems like the AI need to be made aware of which pit box is "theirs", and whether or not the car they share it with is already there. This is quite complex, because the car they share it with is likely to be a human player, not another AI car. I'm actually surprised that this hasn't caused more problems in the past, given that we can have more people on a server than there are available pit stalls. On short races with forced pit stops, I can easily see people coming into the pits only to find that there are no open stalls. Something that makes this even more complex is that at least some of the tracks do not have numbered stalls or pit areas. How someone would be expected to know which pit box was theirs, with no number, no markings, no pit crew, etc. is beyond me.
I really, really would love to see LFS make pitting more realistic, but there are obviously several issues that will have to be addressed before that can happen.
Last edited by Cue-Ball, .
I think that if the tracks had signage showing which way to leave the pit that problem would be resolved. People have done this with signs and marshals standing there, but those things need to be there by default when you install LFS. That is something that I miss from GTR2.
It seems as though better yellow flag handling is on the way. We've now got lights and marshals scattered around the tracks (new Blackwood, for instance). So, if they would just work, we'd be in business there.
Knowing if someone is racing for position is one of the reasons I loved LFS Companion. Hopefully Becky will release a new version or someone else will build something similar to handle this sort of thing.
Edit: I would like to see the track map go as well. Actually, I'd like to see pretty much all of the overlays disappear unless you're spectating. But tire temp and such need to stick around, at least in the GTR cars, until we have some sort of AI pit crew that can manage that stuff for us.
Last edited by Cue-Ball, .
I'm not sure if this is planned to be addressed when the AI become more aware of human players, but it's an issue I saw last night so I wanted to mention it.
When returning to the pits after their cool down lap, the AI don't realize if there is a player in the stall. I raced 8 AI players and passed all but two. We all ran a cool down lap and returned to the pits. I went into the stall after the first two AI guys, and then the next car that came into the pits tried to fit into the same stall, pushing me out the back. This was on South City, but I'd imagine the behavior is the same everywhere.
That said, I really, really like that the AI run a lap then pit. It prevents a massive pile-up at the finish line and teaches good race etiquette to newbies.
When returning to the pits after their cool down lap, the AI don't realize if there is a player in the stall. I raced 8 AI players and passed all but two. We all ran a cool down lap and returned to the pits. I went into the stall after the first two AI guys, and then the next car that came into the pits tried to fit into the same stall, pushing me out the back. This was on South City, but I'd imagine the behavior is the same everywhere.
That said, I really, really like that the AI run a lap then pit. It prevents a massive pile-up at the finish line and teaches good race etiquette to newbies.
You mean FXR. The FZR you already listed as using H-shift, which is correct.
Concerning the clutch, I think we have people on both extremes of this issue. Half of the people are saying that the clutch lasts just fine if you don't abuse it (which is true), the other half says that the clutch slips too easily (IMO, also true).
I think there is something fundamentally wrong with the way that LFS is modeling clutch slip. It's fairly easy to see even in the lowly XRT. During a 1->2 powershift, the tires chirp a bit, then grab. As soon as they grab the clutch starts slipping until the speed differential between how fast the engine is turning and how fast the rest of the drivetrain is turning is resolved. In real life the clutch would be the stronger of the two (having more "grip") and the tires would be forced to spin. For the race clutches, this would be even more true since they often have thousands of foot pounds of clamping force.
The way that you usually burn up a clutch is from slipping it too much. You can burn up a clutch with just a few improper launches (slipping the clutch to keep from bogging the engine off the line) or from "holding" the car on a hill using the clutch. But the clutch should not slip when the pedal is not depressed. Either the tires slip or the engine changes RPM to match the drivetrain.
I think that modeling clutch heat is very important to simulate reality, but the clutch - right now - doesn't have nearly enough force. It should easily overpower the tires and/or the engine, and that's not currently the case. Right now LFS is modeling the clutch as the weak link when, in reality, it's the strongest (compared to the grip of the tires or the torque of the engine). If Scawen can fix this I think both of the groups I mentioned would be happy. Clutch wear would be appropriately modeled, cars could stall, but clutches wouldn't burn up just from a few mis-shifts.
I think there is something fundamentally wrong with the way that LFS is modeling clutch slip. It's fairly easy to see even in the lowly XRT. During a 1->2 powershift, the tires chirp a bit, then grab. As soon as they grab the clutch starts slipping until the speed differential between how fast the engine is turning and how fast the rest of the drivetrain is turning is resolved. In real life the clutch would be the stronger of the two (having more "grip") and the tires would be forced to spin. For the race clutches, this would be even more true since they often have thousands of foot pounds of clamping force.
The way that you usually burn up a clutch is from slipping it too much. You can burn up a clutch with just a few improper launches (slipping the clutch to keep from bogging the engine off the line) or from "holding" the car on a hill using the clutch. But the clutch should not slip when the pedal is not depressed. Either the tires slip or the engine changes RPM to match the drivetrain.
I think that modeling clutch heat is very important to simulate reality, but the clutch - right now - doesn't have nearly enough force. It should easily overpower the tires and/or the engine, and that's not currently the case. Right now LFS is modeling the clutch as the weak link when, in reality, it's the strongest (compared to the grip of the tires or the torque of the engine). If Scawen can fix this I think both of the groups I mentioned would be happy. Clutch wear would be appropriately modeled, cars could stall, but clutches wouldn't burn up just from a few mis-shifts.
I don't think that most pit lane speed limits are monitored down to 1 mph increments. I would guess that in real life there is a small margin of error allowed. If so, it would make sense to do the same in LFS.
Well, i think the best solution would be if the clutch were modeled properly in the first place, so that DXTweak wasn't needed.
I think that the removal of auto-blip and auto-cut is probably the best thing to happen to LFS in a very long time. Certainly the best thing since the improved tire physics way back in version T or whatever it was.
I'm wondering if that is partly due to the longitudinal grip issues in LFS. During a powershift from first to second, there are really only two places for the power to go, assuming no parts break: the clutch or the tires. The LX, XRT, etc all have enough power to break loose the tires during a hard 1->2 shift, but instead the clutch seems to slip. I have a feeling that the coefficient of friction for the clutch is lower than that of the tires to the pavement, which would seem to be incorrect.
Yup. I've got no problems at all. The speedo in the XRT is about 5" diameter.
I thought about that, but didn't have time to go back and check the gearing. I'll try to do that later, but since I was trying this on South City Town, I'm pretty sure I was using a fairly low geared setup.
Something is definitely off in the way that the engine/flywheel responds. It takes a very long time for the engine to wind down, and the engine will resume running if you depress the clutch when it's barely spinning. I don't know if changing the inertia when under 500rpm is a good fix or not. I think that would just introduce more problems in the future. I'd rather see compression, flywheel weight, etc. accounted for. That way, in the future, we could lose compression on a cylinder which would cause one effect, or lose spark on a cyclinder which would cause a totally different effect. Similarly, it would be easier to change car behavior using different flywheel weights, grabbier/softer clutches, higher compression, etc (via Scawen on the back-end, not user editable).
Of course, this is all a lot easier said than done. And I've really no idea how complex the current engine/flywheel/clutch model is or what factors it takes into account. All I know is that right now the behavior is quite odd. A huge improvement over X10, but weird enough that it takes away from the immersion.
Ah! Thanks for that. I looked down below where it was before, but didn't notice that it had simply moved.
But the clutch behavior is still funky.
After doing some testing with the XRT, I'm of the opinion that something is still not quite right with the way stalls and clutching are handled. Trying to do a start on an uphill is damn near impossible without either stalling the car or spinning the tires. There's almost no way to cleanly take off without slipping the clutch VERY slowly (and even then, it's extremely difficult). The same thing with trying to take off in 2nd gear on flat ground. The clutch grabs very quickly (at the very beginning of pedal movement) and the revs drop to zero. Pulling away from a dead stop in a normal road car, in first or second gear, is usually quite easy in real life. There's no need to rev it over 3K, and you don't have to be unusually slow in letting out the clutch either. LFS doesn't seem to model this correctly, IMO.
Additionally, stalling doesn't seem to act quite right either. Several times while trying to start off on an uphill the car would drop to 0rpm (or very nearly so, as indicated on the tach), but cluching would allow the engine to idle back up to normal. It's as though the car has a huge flywheel that stops the engine from stalling completely, while simultaneously having a flywheel that is so small it can't get enough momentum to move the car without really spinning quickly. Something is not quite right.
I've also noticed that calibration lock disappeared from the setup menu (this is X31. I dunno about X30 since I never checked). If you previously had your calibration locked and you just installed the patch over your old setup, I don't see any way to un-lock those settings.
Additionally, stalling doesn't seem to act quite right either. Several times while trying to start off on an uphill the car would drop to 0rpm (or very nearly so, as indicated on the tach), but cluching would allow the engine to idle back up to normal. It's as though the car has a huge flywheel that stops the engine from stalling completely, while simultaneously having a flywheel that is so small it can't get enough momentum to move the car without really spinning quickly. Something is not quite right.
I've also noticed that calibration lock disappeared from the setup menu (this is X31. I dunno about X30 since I never checked). If you previously had your calibration locked and you just installed the patch over your old setup, I don't see any way to un-lock those settings.
I'm very satisfied with the current 45* look. It lets you look far enough to avoid collisions and give up places when necessary, without being as over the top and disorienting as the 90* look.
I vote to set it at 45* (or maybe up it to 60*, as a concession to the people who want to see further to the side) and leave it at that. I would also like to see the backwards look removed completely.
I vote to set it at 45* (or maybe up it to 60*, as a concession to the people who want to see further to the side) and leave it at that. I would also like to see the backwards look removed completely.
FGED GREDG RDFGDR GSFDG