The online racing simulator
Searching in All forums
(976 results)
MadCatX
S3 licensed
Quote from Skinzinho :
Just trying to understand a bit more about the Library. To be honest, I never tried to integrate the Insim.NET library with an application.

Is it worth it? LFS_External is outdated, unmaintained and partially broken. InSim.NET will do a much better job for you and you won't have to hack it by yourself to make it work with recent versions of InSim and LFS.
MadCatX
S3 licensed
Is there any reason why you use LFS_External which is known to have multiple serious issues with newer versions of LFS? Use InSim.NET instead...
MadCatX
S3 licensed
Quote from Scawen :
Thanks for that code. I would be interested in looking at that but I not this time round. I have to work on the final VR updates today so we can release a full version on Saturday.

I understand. For a quite poorly coded example you can take a look here (https://www.lfs.net/forum/thre ... 27-LEDs-mod-%5Brenamed%5D), "Source\DInputDll\myDInput.[h, cpp]". One potentially annoying problem is that there is (was?) a bug in Logitech drivers that could make the LED commands and force feedback commands clash if both were being submitted at too high rate, resulting in jerky force feedback. I could be worked around by limiting the rate at which the LED commands were sent.

The Logitech SDK I played with was pretty much just a wrapper around DirectInput and a layer that allowed it to talk to the Logitech Gaming Software. The SDK was open-source so I just ripped out the parts I needed.
Last edited by MadCatX, .
MadCatX
S3 licensed
@Scawen:

Since you implemented some wheel-specific options in LFS, is there any chance you can look into supporting the LED lights on G27 & G29 so I can stop maintaining my horrible hack that makes them work with LFS?Smile There is no need to use the Logitech SDK and it all should be a matter of a simple


if (wheelModel == LG_G27 || wheelModel == LG_G29) {
ffbEffectHandle->Escape(&packet);
}

where "packet" is a pointer to a DIEFFESCAPE set to this

packet.dwSize = sizeof(DIEFFESCAPE);
packet.dwCommand = 0;
packet.lpvInBuffer = &wheelData;
packet.cbInBuffer = sizeof(wheelData);

"wheelData" is a structure looking like this:


typedef struct {
float currentRPM;
float rpmFirstLedTurnsOn;
float rpmRedLine;
} LedsRpmData;

typedef struct {
DWORD size;
DWORD versionNbr;
LedsRpmData rpmData;
} WheelData;

"size" in WheelData is to be set to "sizeof(WheelData)" and "versionNbr = 0x00000001". Contents of the LedsRpmData are hopefully self-explanatory...
MadCatX
S3 licensed
This most likely doesn't have anything to do with your GPU actually running out of VRAM. As far as I remember from what one of the i915 module devs said, i915 will allocate as much memory as needed up to the 2 GB limit. Reserving huge amounts of RAM for GPU directly in BIOS is actually a waste of memory, i915 can allocate it dynamically.

What actually happens is probably covered by this bug (https://bugs.winehq.org/show_bug.cgi?id=13335). LFS apparently tries to allocate so much resources that it runs out of the address space that WINE reserves for it. This comment in particular (https://bugs.winehq.org/show_bug.cgi?id=13335#c192) suggests two hackarounds that might fix the issue for you.

I can try to experiment with this on my machine with Sandy Bridge GPU once I'm off work...

EDIT: 0.6N works fine on my Sandy Bridge laptop (Arch Linux, WINE 1.9.6) even with Westhill. Fedora patches WINE with out-of-tree Gallium Nine patch series. It is not enabled by default and it shouldn't have any effect when Intel GPU is being used but it's the only difference related to 3D rendering I can think of now.
Last edited by MadCatX, .
MadCatX
S3 licensed
Automatic shift mode in LFS is not supposed to simulate an actual automatic transmission; racing cars don't have those. What you might be looking for is this: https://www.lfs.net/forum/thread/55636-Automatic-Transmission-Simulator-%28v5-6b%29
MadCatX
S3 licensed
You don't need such a system unless you specifically want to share your music with other people. Music is however mostly copyrighted intellectual property and you cannot distribute it without explicit approval from the copyright owner. I cannot see how LFS devs would get that, not to mention that recording companies would most likely require LFS to have some advanced DRM to prevent unauthorized distribution of the music.

From the player's point of view I also don't see why would I even want to listen to a LFS Radio DJ rather than to my own stuff.
MadCatX
S3 licensed
Quote from Racer X NZ :What part of "WORKS ON PC WITH SELECTED TITLES" is difficult to understand ?????


Please define "SELECTED". Here is the standard definition.

...

Isn't it ironic that you want to school us about the meaning of the word "selected", yet it is you who fails to understand what that sentence actually says? The sentence is Logitech's way to protect themselves against morons extraordinaire who might expect their wheel to work games that do not support that kind of controller at all. I mean games like first-person shooters, RTS strategies and other obvious examples.
MadCatX
S3 licensed
Are you trying to modify the project? If that's so, you should use more up-to-date CInSim and InSim.h. Rather than trying to exhume this ancient VS 6.0 project it might be easier to create a new VS 2015 project and add these files to it.
MadCatX
S3 licensed
It all comes down to you misinterpreting Logitech statements. And yes, I do have a G920 and it works perfectly fine, even on Linux (but without FFB for now).

So, for everyone who is not Racer X, the bottom line is: G920 is fine to use on PC as long as you have drivers that support G920. It will work with any game title what uses DirectInput (which is basically every game out there).

As for one of us being an idiot, I think I'll let the jury decide that. I rest my case...
Last edited by MadCatX, . Reason : Make it clear that no profiles or whatever nonsense is needed.
MadCatX
S3 licensed
What does this have to do with anything? As can be inferred from the text you quoted, "profiles" in games are merely a convenience thing.

If you had a tool to switch the G920 into the "PC" mode, it would work even without any additional driver because the wheel appears as a standard HID device. The Logitech driver is needed to switch the G920 from its default "Xbox" mode to "PC" mode. It also provides support for force feedback and firmware updates. The driver uses DirectInput which is the standard Windows way for applications to talk to input devices.

Just so we're clear, what I wrote above is not speculation, I know it for a fact. I'm posting this with hope that your inaccurate claims won't confuse any other G29 and G920 users out there...
MadCatX
S3 licensed
I don't care who I prove wrong, it's just that a considerable portion of nonsense I read on LFSF originated from you. The webpage says that the list of "Compabible games" lists games that are *known* to work with a particular wheel; known as in *tested*. This does not imply that games that aren't on the list won't work. There is a list both for G29 and G920. Also, Live for Speed is specifically listed as tested with the G920 but not with G29.

Since there is DirectInput on Windows there is no reason why any of the Logitech wheels wouldn't work with any game that makes use of it. Logitech doesn't have any proprietary interface of their own that games would have to use.

I consider the bullcrap one might read on Logitech forums mostly irrelevant as it mostly consists of rants coming from people dumb enough to not install up to date drivers and etc.
MadCatX
S3 licensed
Quote from Racer X NZ :A quick warning on the G920, this is from logitech.

"Driving Force is designed for the latest Xbox One™ console racing game titles. You may never race with a regular controller again after you experience G920 Driving Force.

G920 Driving Force also works on PC with select titles using Logitech Gaming Software."

This is NOT a PC wheel, it is an Xbox wheel with limited PC support. So if it doesn't work with some games on a PC that is why.
The forum is full of info on this.
https://forums.logitech.com/t5/Logitech-G-Controllers/G920-on-PC-is-a-disaster/td-p/1436049

Sorry, but you are completely wrong (I'm tempted to say "as usual"). G920 speaks both proprietary XBox protocol and Logitech HID++ protocol. The latter is implemented in drivers for PC. On Windows all you need is a recent driver.

If your G920 is not even recognized as an input device, open Device Manager (WinKey + R, type "devmgmt.msc"), find the G920 in the list, double click it and and open the "Details" tab. On the tab select "Hardware Ids" from the dropdown menu and check the output. There should be a line that says "USB\VID_046D&PID_C262". It can't check at the moment so the actual output might be a bit different, however, the IDs "046D" and "C262" are important. 046D is Logitech vendor ID, C262 is the ID of G920 when it's switched to the PC mode. If you see C261 instead, the wheel is still in the Xbox mode. If this happens, make sure that you have the driver installed correctly and maybe try to plug the wheel in a different USB port.
MadCatX
S3 licensed
I don't think that the graphics issue is fixable. I believe that the VMWare 3D acceleration exists only to provide basic support for applications that require it, it is not meant to be used for gaming. With LFS using pixel shaders now you will probably have to abandon the VM solution.

As for the sound routing issue, perhaps any of these tips will help? https://superuser.com/questions/258969/routing-applications-sound-to-different-sound-device-windows
MadCatX
S3 licensed
Are you sure that the problem can't be caused by an overzealous firewall? Some firewalls can detect that an executable has changed and disallow access to the network unless you tell them otherwise. I find it quite odd that the connection seems to drop out right after it's established. I can't recall at the moment but IIRC LFS threw an error when it received a malformed IS_ISI packet.
MadCatX
S3 licensed
I'm sorry to say this, but your "new" computer might have even worse specs that your old one. The CPU is over 5 years old and the GPU is the worst in the line. I wouldn't be surprised if the older CPU (assuming it's a Core2 Duo) performed better at LFS than E5800.
MadCatX
S3 licensed
This is the intended behavior. Nonsensical value of InSim version was a result of the mod not checking if the connection was actually opened successfully. Are there any error messages popping up in LFS because the error message returned by the mod suggests that LFS doesn't like the mod for some reason. Can you check that the mod still works with some older version of LFS?
MadCatX
S3 licensed
I updated the G27LEDs mod to hopefully spit out some useful error message if something goes wrong with InSim or OutGauge link. Maybe that'll give you some idea what's going on. I haven't seen the source for ages so I have no idea what I might have broken in the processSmile
MadCatX
S3 licensed
Well, this actually suggests that something in the InSim interface has changed. What I find particularly odd is the LFS version being read as zero. I can see a few problems with the mod's net code so perhaps something has changed in LFS and it's causing trouble now.
MadCatX
S3 licensed
The mod does not care about what version of LFS you have. Is should work with every version of LFS that uses the "new" OutGauge format.
MadCatX
S3 licensed
Quote from joordy599 :There are ways to prevent such a thing, if you look at hosting companies, a lot have an DDoS protection, Rack Service and MHServers, well that stands in their description, but I know that there are sources for DDoS, like myself I'm using for my sites CloudFlare and game server, there is also such a resemblance.

Instead of arguing about how the Internet works, could you suggest what you think LFS could do against DoS? Because I cannot think of anything that would be really effective.
MadCatX
S3 licensed
DoS attacks cannot be effectively prevented because of the very nature how the Internet works. If the attacker has a big enough gun, they can always flood your downlink. Mitigation of some lamely executed DoS attacks can be done by a properly configured firewall. I don't know why LFS should be concerned by this, not to mention that LFS itself cannot really do much once you are under attack.
MadCatX
S3 licensed
LOL, I assume that you were having a pretty serious hangover? What you said translates to "I want the stuff you spent God knows how many hours working on for free". The devs told any potential buyers right away that RO might be the only thing they get for £12. Thank you for being the first 2016's example that human stupidity is still alive and kicking...
MadCatX
S3 licensed
I didn't need to set any library overrides with 0.6K. Having D3DCompiler_43.dll and d3dx9_43.dll in the LFS directory seems enough for me with Wine 1.8. So far I checked only on my laptop with Inteh HD3000 GPU, but it all seemed fine.
MadCatX
S3 licensed
These are pretty strange results. I couldn't find any traces of stuttering in the strace log, there is only one little spike that seems to be done intentionally by WINE. As long as you can reliably reproduce it it might be interesting to ask around WINE devs mailing lists about this. MOMO uses the same lg4ff driver as G25 so I find it odd that it wouldn't work.
FGED GREDG RDFGDR GSFDG