Not necessarily. The problem can be anywhere deeper down the way in the HID or USB stack. The lg4ff driver is rather simple and although it does grab two spinlocks since kernel 4.2, it cannot really make an userspace application stuck unless there is concurrent write access to the wheel. The only way how to get this on a G25 would be by constantly changing the range through sysfs and playing some FFB effects. It would also show up in strace as write() or ioctl() calls on the wheel's file descriptor taking a long time.
This is a very weird output. Are you really running at least kernel 4.1? Is the support for Logitech wheels (CONFIG_LOGIWHEELS_FF=y) compiled in? Can you boot the kernel with "hid.debug=1" parameter and post the output of dmesg after you have plugged the wheel in? There is absolutely no way that the sysfs entries would be missing unless something fails during the initialization.
The missing buttons and shifter would most likely be consistent with the wheel staying in the compatibility mode. I don't recall how LTWheelConf worked but if my memory serves me right LTWC would not switch the wheel into native mode if it had been switched already by the kernel driver. In fact LTWC cannot work properly anymore because the kernel driver overrides any changes made by LTWC when libusb returns the wheel back to the kernel.
What interface did you check? 4.1+ kernels should pick up DFGT correctly. Can you check the output of "/sys/bus/hid/drivers/logitech/<dev>/alternate_modes" and "/sys/bus/hid/drivers/logitech/<dev>/real_id" when the wheel is connected? Value of <dev> will vary.
This sounds like a hardware problem to me because the driver has no control over the autocalibration.
When does this happen? I'd image this to be a problem with a game and not the wheel of drivers.
It's next to impossible to answer your question when you don't post any code that is supposed to do what you want. You should check that "time changed" event in your application is actually triggered and that you are sending the correct button in the event handler.
Are you looking for something that will alert you when somebody is using speedhacks?
Print the current distance from the location and car's X,Y coordinates to console to check what values are you actually getting.
He's talking about the Intel 4500MHD GPU that is a part of Q4x chipset. This is a pre-Sandy Bridge era GPU and as such it isn't any good for gaming. I would also imagine that the drivers haven't received any updates in years.
Set the FPS limiter to something your machine can keep up. New patch added some optimizations that might result in sudden FPS jumps at times on weaker machines.
Also, does it happen everywhere or just on the new Westhill?
One of the obvious mistakes you're doing is that you're not checking for exceptions. If something goes wrong, your program just ignores the problem and continues.
That code as it's written wouldn't even compile, so what's the problem you're having exactly?
You cannot do this with just InSim alone. You can use WinAPI's FindWindow() and SendMessage() functions to get the HWND of LFS window and send simulated keystrokes to it.
You can still use LTWC's "alt_autocenter" option to adjust centering spring force.
./ltwheelconf -b N -d /dev/input/eventXX
You can also control the gain in the same way. Note that this will work with any force feedback device that supports gain and centering spring, it is not limited to just Logitech wheels.
As of now there is no such option...
People have been complaining about this for years. Bottomline is that there is nothing you can do about this. You could perhaps negotiate some bans with the owners of individual servers but there is no general solution. I suggest you ignore it, it's just a skin...
By that logic you should also stare in amazement at the coffee maker in your kitchen because it makes water hot at the push a button - and without any fire no less!
It's not like I agree with the occasional moaning though, because I never understood the logic behind it. Usually if one is dissatisfied by a product they go and get another one - it is called voting with one's wallet. Instead the moaners keep crying about what they dislike about LFS and they come up with all sorts of odd reasons why the devs should accommodate them such as "we have the right to know". Others just keep restating something along the lines of LFS being dead. Just face it already friends, you got what you paid for with your S1 or S2 license, anything else is just a bonus given to you for free by the devs...
Since when do you not know what is happening? First the engine got updated to use D3D9 and gained VR headset support. Then there was the new Westhill. After the Westhill came a series of engine patches to reduce CPU load. As far as I understand there are some internal changes being done to meet Eric's needs for some new content. Is this not enough to know? It's not like we were getting any more news back in the old days but because the patches were being released at a faster pace nobody moaned about it.
I think that worse than this is easily possible. LFS might not be in the best shape but this is a opinion shared mostly amongst some part of the current LFS community. If the development picks up the pace, the tables might turn in a some time. Putting in on Steam in its current state and having everybody complain about its high difficulty, outdated graphics and a vacuum of additional features would literally be a suicide for the project. Gaming market doesn't give second chances easily.
The whole idea of selling LFS through Steam is based on the assumption that putting LFS on Steam would automatically translate into greater revenue and more players. I have however never seen a good series of arguments to support this claim.
If you think about it, LFS does not have much to offer. The graphics is dated, sound effects are not anything special and the entire gaming experience is almost sterile. LFS lets you drive around in circles in (mostly) fictional cars and that's it. There is no career mode, the AI is boring to race against, the damage model is quite simplistic, there is no dynamic weather, etc. You also get no actual driving aids and as far as my experience tells it takes most people a good long while to even complete a lap when they try LFS for the first time. Add the fact that you cannot even hope to be competitive without a decent car setup into the mix and you end up with a game with very high entry barrier and little to offer if you overcome it.
This is not to say that LFS is a bad game, it just targets a very specific audience. Even if you put it against AC or iR it still appears to be a clear looser; iR and AC at least offer real cars and tracks and nicer graphics. Why should I consider getting LFS that lets me drive FZ50 around Westhill when I can drive Ferrari F40 around Monza in AC? The actual driving part in LFS is still top notch but I would imagine that there is just a handful of people who recognize and prefer that over real content, better looks and multitude of additional features.
As for the notion that greater revenue would help LFS move forward, I suggest you look up some scientific studies done on the topic of using money as motivation. Simply put, more money tends to overmotivate people who then to do sloppier job. The better the money is, the worse it gets. Contrarily, people who have a chance to work on things they like and enjoy usually produce great results.
If you think about it, LFS does not have much to offer. The graphics is dated, sound effects are not anything special and the entire gaming experience is almost sterile. LFS lets you drive around in circles in (mostly) fictional cars and that's it. There is no career mode, the AI is boring to race against, the damage model is quite simplistic, there is no dynamic weather, etc. You also get no actual driving aids and as far as my experience tells it takes most people a good long while to even complete a lap when they try LFS for the first time. Add the fact that you cannot even hope to be competitive without a decent car setup into the mix and you end up with a game with very high entry barrier and little to offer if you overcome it.
This is not to say that LFS is a bad game, it just targets a very specific audience. Even if you put it against AC or iR it still appears to be a clear looser; iR and AC at least offer real cars and tracks and nicer graphics. Why should I consider getting LFS that lets me drive FZ50 around Westhill when I can drive Ferrari F40 around Monza in AC? The actual driving part in LFS is still top notch but I would imagine that there is just a handful of people who recognize and prefer that over real content, better looks and multitude of additional features.
As for the notion that greater revenue would help LFS move forward, I suggest you look up some scientific studies done on the topic of using money as motivation. Simply put, more money tends to overmotivate people who then to do sloppier job. The better the money is, the worse it gets. Contrarily, people who have a chance to work on things they like and enjoy usually produce great results.
Last edited by MadCatX, .
You can, but I have some doubts that the extent of the possible modifications will allow you to get reasonably close to characteristics of an electric drive.
You can use LFSTweak to modify a lot of parameters related to the car dynamics. However, I don't think it will be very useful to simulate electric cars because the underlying model assumes a car with internal combustion engine.
LFS makes use of Direct3D9 Ex since 0.6J, perhaps that triggered a bug in the graphics driver, especially when you are on Win10 where the drivers might be a bit glitchy. Unless there is an update available you might want to talk to AMD's techsupp.
This looks a lot like the CPU governor mishandling power management and sending the CPU into some power saving mode even when it shouldn't. I don't know what tool on Windows can track CPU's C and P states but I suppose this is something you should look into. There have also been reports of the Hyper-V hypervisor causing serious performance problems with LFS.
What happens when the wheel "disconnects"? Does it die completely and recalibrate itself when it comes back online or does it just stop responding in the game?
The brief history of Logitech wheels support in Linux can be summarized like this:
- 3.2 Initial support for force feedback and range change, DFP, DFGT, G25 and G27 supported
- 3.5 G27 LEDs support added
- 3.10 Support for a newer revision of DFGT added
- 3.13 Various improvements to the handling of autocentering and "constant force" force feedback effect
- 3.14 Support for newer revision of G27
- 4.1 Reworked wheel identification and mode switching code, all current and future revisions of supported wheels should be recognized properly
- 4.2 (currently in -rc7) Various cleanups and improvements, most of them irrelevant for a general user
... Future plans
- G29 support: could land any day
- Support for other force feedback effects that "constant force": In the works, proof of concept code working, lots of testing required
Force feedback is mostly a kernel thing. Besides patching the kernel there is not much more that can be done to improve the experience. Universal tools to set the range etc. won't happen for a long while - if ever - because this stuff is all vendor-specific with no standardized interface such a tool could leverage.
I noticed that the Wiki page concentrates only on Debian at the moment, maybe a bit of restructuring so that instructions for other distros can be added easily is a good idea?
This has been asked like a hundred million times. LFS_External itself is broken, you can't fix it.
FGED GREDG RDFGDR GSFDG