Oh... I tried renaming the library to something nonsensical and the problem came back. I conclude that WINE uses the DLL automatically no matter whether I explicitly tell it to do so or not. I suppose that the "override" directive is needed only if the library is placed in the WINE's equivalent of "system32" directory, whereas I put in directly into LFS directory to avoid any potential problems with other applications.
Nice, thanks.
Reverting to E15 doesn't fix anything which is rather odd because I'm absolutely sure it used to work before. It's however possible that I'd had the required fix already in place when I tested LFS in WINE last time. I replaced the hard drive in my computer some time after that so I might as well have forgotten all about it...
Putting "d3dx9_43.dll" into LFS directory and launching LFS like this:
WINEDLLOVERRIDES=d3dx3_43=n wine LFS
fixes the problem with no apparent side effects so I mark this problem as solved. It'll make a good WINE AppDB entry though.
The problem seems to be isolated to JPEG textures - as is somewhat confirmed by this "fixme" message
fixme:d3dx:D3DXCreateTextureFromFileInMemoryEx Generation of mipmaps for compressed pixel formats is not implemented yet
I guess that a newer revision of DX9 introduced a functionality that is not supported yet by WINE directly. I'm on a less than useless mobile connection at the moment so I cannot grab the latest DX9 libraries and start experimenting but if it takes as less as telling WINE to use a native d3dx9 library I don't consider it a problem.
Any chance you can reupload E15 so we can verify that a native d3dx9 library is necessary only for E17 onwards?
I attached a screenshot to show what my LFS looks like now.
EDIT: I tried to downgrade to 1.7.18 (what's reported to work by felplacerad) and it exhibits the same behavior so I suppose that WINE really is lacking some functionality used by LFS.
Last edited by MadCatX, .
The first DX9 enabled version (E15?) worked perfectly with the latest version of WINE at that time. Oddly enough I seem to be having an issue with E17. WINE 1.7.19 apparently fails to do some texture conversions and a lot of textures are missing as a result.
fixme:d3dx:D3DXLoadSurfaceFromMemory Format conversion missing 0x14 -> 0x31545844
fixme:d3dx:D3DXCreateTextureFromFileInMemoryEx Texture loading failed.
Google suggests that this can be worked around by installing native d3dx9_36.dll
It depends on the type of the FFB effect. Only constant force is supported by the Linux driver at the moment. LFS should work out of the box as long as FFB is enabled in the game and correct input device is selected. WINE might pick up gaming controllers twice, once as "/dev/input/jsX" which is a legacy joystick interface and again as "/dev/input/inputX". AFAIK the legacy joystick interface does not support force feedback so if you accidentally selected that interface in LFS it might be the reason why you're not getting any FFB.
Dead seriously. As much as you are aware of the position of your hands on the wheel you can be just as aware of the position of your hand on the table. It's arguably more difficult to master but it's mostly a matter of practice.
Mustafur is talking about a petrol engine. Turbocharged petrol engines usually run on a rich mixture ratio to keep the combustion temperatures low. If you lean the mixture out you're obviously going to get better fuel economy. This is obviously not the case with diesel engines.
I used to drive with a mouse back in the day and I didn't find the virtual wheel of any particular help. You can see what the car is doing and adjust the input accordingly. As much as you don't look at the real wheel to make the car go straight you don't have to look at the mouse or any other reference except than the car itself. The only thing I found really inferior to the wheel was the digital brake input.
You've got to realize that the notion that some dude in a garage can figure out a better engine maps than a team of engineers with experience in engine design and state-of-the-art testing facilities is ridiculous. Chiptuning increases power by cutting away some safety margins the manufacturer put there to make the engine take a lot of abuse and last long. If you do decide to have the engine modified, you definitely should take a good care of the engine and have it serviced regularly by the manufacturers instructions.
A properly done chiptuning hopefully won't end up destroying the engine, but if it's done sloppily it might. For instance if the fuel injection time is increased too much, the injected fuel doesn't have enough time to vaporize and the drops of cold fuel hit the hot piston head. This is not supposed to happen and pistons are not designed to deal with that so eventually the piston will crack due to thermal stress. Diesel engines in particular are prone to a "diesel runaway" kind of failure.
All in all the success of this process depends on the condition of the engine and proper maintenance. Chiptuning an engine with 200k km under its belt and dubious maintenance record is probably asking for trouble. Also realize that the car is not just the engine, higher power will put more stress on the clutch and transmission.
Since chiptuning pretty much is all about pumping more fuel into the engine I highly doubt that it will be more economical.
A properly done chiptuning hopefully won't end up destroying the engine, but if it's done sloppily it might. For instance if the fuel injection time is increased too much, the injected fuel doesn't have enough time to vaporize and the drops of cold fuel hit the hot piston head. This is not supposed to happen and pistons are not designed to deal with that so eventually the piston will crack due to thermal stress. Diesel engines in particular are prone to a "diesel runaway" kind of failure.
All in all the success of this process depends on the condition of the engine and proper maintenance. Chiptuning an engine with 200k km under its belt and dubious maintenance record is probably asking for trouble. Also realize that the car is not just the engine, higher power will put more stress on the clutch and transmission.
Since chiptuning pretty much is all about pumping more fuel into the engine I highly doubt that it will be more economical.
The string and damper effect strength probably shouldn't be down to zero.
Port 29339 does not have to be forwarded because it's the outgoing port, only 63392 does. "Forwarded firewall ports" makes no sense to me, did you actually mean "allowed those ports on the firewall"? Do you have a static IP address set on your computer? Some routers might fail to forward ports to devices with dynamic IP address. Does your router have any sort of firewall of filter that might block the connection?
And what exactly is that? Tell us precisely what you did and how and hopefully we'll spot something you've done incorrectly...
You'd have to patch the kernel with some improvements that have not made in into the official Linux kernel yet. It's not particularly difficult to do but you should look up some tutorials how to build a custom kernel in Ubuntu before you try...
FWIW, the patched lg4ff driver and the additional stuff that's required for it to work can be found here (https://github.com/edwin-v/linux-hid-lg4ff-next).
FWIW, the patched lg4ff driver and the additional stuff that's required for it to work can be found here (https://github.com/edwin-v/linux-hid-lg4ff-next).
The wheel seems to be picked up properly by the kernel so you should have force feedback in games that use "constant force" effect. I don't have a first hand experience with ETS2 but as far as I know it mostly uses "damper" and "spring" effects which are unsupported by the driver present in the Ubuntu 14.04 kernel...
LTWheelConf should not be needed at all. In fact, because the entire FFB support has been moved into the kernel, LTWheelConf cannot work properly anymore. You should be able to set the range via the kernel interface by writing the desired range into
.
Are you sure that the device appears as "event0" in /dev/input? You can find out the correct node name from dmesg. Are you getting force feedback in other games?
ETS2 makes use of effects that are not supported in current kernel drivers. A series of out-of-tree patches that enable full support exists, but you'd have to build yourself a customized kernel.
/sys/module/hid_logitech/drivers/hid:logitech/*:*/range
.
Are you sure that the device appears as "event0" in /dev/input? You can find out the correct node name from dmesg. Are you getting force feedback in other games?
ETS2 makes use of effects that are not supported in current kernel drivers. A series of out-of-tree patches that enable full support exists, but you'd have to build yourself a customized kernel.
DFGT should be supported natively by the kernel. Unless you're running a very old distro, it should work out of the box. If it doesn't, please post more detailed description of the problem you're having.
People with experience in managing projects where intellectual power is essential. Just read a summary of the book Dustin linked. It's easy to get money through a loan, startup, etc, but you cannot buy a good idea.
Again, not necessarily - something that has been investigated thoroughly by economists.
No matter how it's called MMORPGs were one of the first to bring this business model to gaming. Cloud computing is very different to gaming, please stop making such nonsensical comparisons, it doesn't move the conversation anywhere.
How would you know, you haven't seen the devs' accounting books, have you? You're again making assumptions that support your argument but you have no evidence to back them up with. I assume you're talking about Vic? What's wrong with him taking part in another project if he feels underutilized with LFS?
This is pure speculation and doesn't have much in common with computer software. It has been explained like gazillion times that more money does not equal better product or faster development. The devs said so, practical experience from various SW projects supports this claim. iR hasn't invented anything, its just an MMORPG business model applied onto a racing sim.
The devs have never ever suggested that financial difficulties were an issue. If they had, they could have sold the whole product and move on. I imagine that LFS physics engine would still be worth a lot of money.
But numerous test patches with the 3D engine ported over to DX9 and VR goggles support might...
Brilliant! Apparently I was wrong thinking that the conspiracy theories about frozen LFS development cannot get any crazier
Sailfish 1.0.5.16 arrived a few hours ago...
To get back on topic, can this be some odd power saving issue? I know for certain that HD3000 can average ~ 20 FPS in LFSBench at 1600x900 with 4x AA with and i5-520M when running through frikkin' WINE. Perhaps the Windows driver doesn't ramp the GPU clock up when LFS is running? In any case I'd check that all drivers are up to date.
Unless you're suggesting that LFS can tax GPU just as much as these games I think I made my point.
FGED GREDG RDFGDR GSFDG