The online racing simulator
To Devs: Linux Support
(16 posts, started )
To Devs: Linux Support
Hello
I've been a fan of racing games, especially simulators, for many years. I am also a GNU/Linux user, which makes it quite difficult for me to fully enjoy many games that are not supported by our operating system. For some years we can enjoy some good games like "F1 2017" or "DIRT Rally" on Linux, but in the linux community we miss a real simulator to enjoy.

For years I co-manage a community of Linux gamers (jugandoenlinux.com), where racing games are very popular, and where we also organize online races often.I know that at least in my community a sim-racing game would have very good reception.
As for the support of steering wheels and controllers, you should know that at least Logitech steering wheels work fully with Force Feedback (Linux users always buy this brand), and the vast majority of controllers works in Linux without problems.
I know this post probably won't even be answered, but at least I should try
I don't know any commercial simulator that has support, so if you decided to finally develop a build for Linux you would be the first ones.
Greetings
Isn't it working just with WINE? Seems like the OP wants a native Linux support without need to use emulator.
Quote from Eclipsed :Isn't it working just with WINE? Seems like the OP wants a native Linux support without need to use emulator.

I'm saying exactly that. Thank you very much for clarifying it. Thumbs up

I knoow that LFS works with Wine. In the other post I see that Scawen talks a lot of times about XP and Linux users, and this is the reason of they use DX9. They must know that Wine is DX11 compatible thanks to DXVK, a Wine plugin that uses Vulkan to convert DX11. Of course is always better to not use wrappers to play this game. I think that they must to move to Vulkan to easily build for Linux also. You should also know that even if you don't port to a native application, if you use Vulkan, wine will work much better.
Quote from Scawen :
I haven't been catering specifically to those systems in any way. I haven't spent even 5 minutes trying to cater to them.

last time i checked, LFS runs flawlessly on linux.

Quote from Scawen : It seems strange to go native Linux yet. Maybe OpenGL will be the way to go, Microsoft is basically driving me away.

#6 - Ped7g
Just recently updated my KDE neon distro, installed wine (that went a bit shaky due to 32b support which is not supported by the distro itself anymore, and I'm probably heading to a state when I will have to better understand how 64b wine operates and also drop all 32b native linux games), and LFS "just works", although I don't have any wheel, so I'm talking about rendering/network/keyboard/mouse.

EDIT: of course I would love to see LFS going open source somewhere in future, that would also allow for native port over time...
Quote from Ped7g : I don't have any wheel, so I'm talking about rendering/network/keyboard/mouse.

EDIT: of course I would love to see LFS going open source somewhere in future, that would also allow for native port over time...

Logitech Wheels support are included in the Linux kernel since a lot of time, and they works perfect with Wine and LFS.

About OpenSourcing LFS, it's obviously a decision their developers have to make. Opening the code would obviously give them a lot of help for their project, but they would have to change their business model.

Personally I would be thinking about updating the graphical API to a more modern and multiplatform.... Vulkan is the only way....
#8 - Ped7g
Quote from leillo1975 :Logitech Wheels support are included in the Linux kernel since a lot of time, and they works perfect with Wine and LFS.

... and? I mean I don't have any wheel at home, no HW. (I'm not talking about driver issues ... Once you don't have HW, you don't have driver issues, it's magical, everybody should try that some time...)

Quote :About OpenSourcing LFS, it's obviously a decision their developers have to make. Opening the code would obviously give them a lot of help for their project, but they would have to change their business model.

Why exactly? It would made cracking offline part of LFS slightly easier (removing the check in source is simpler than debugging the binary), otherwise not much else would change, you would still have to buy license to access S1/2/3 content in MP games.

Unless Scawen is actually selling LFS code base as his know-how into other projects, the business model would be quite unaffected, except all the new noise of pull-requests coming...

Quote :Personally I would be thinking about updating the graphical API to a more modern and multiplatform.... Vulkan is the only way....

It took long time to move from DX8 to DX9, and that's somewhat related API, and it's just *now* the content and rendering method does catch up and the LFS is finally taking advantage of those new DX9 features... Seems to me like this next update (with track graphics and new rendering method) will be done somewhere in 2019. Then the car graphics will become elephant in the room, plus there's still the new tyre physics to be finished and released. I don't see any practical reason to prioritize yet another render API change over these two major points, so I guess your request will be considered not sooner that at summer 2020 (after cars-patch, which I guess will be released somewhere in 2020..2022)...

At least vulcan will be quite mature at that point I guess, unless they will start to add new features every year like DX and OGL did lately, keeping the drivers in constant mess of not working precisely as specification says, or even situations where gfx drivers contain database of known shader logical bugs, and compiling such source into different shader code to fix the bugs in some games...

There's thread where I was suggesting moving to OGL about 5 years ago I guess (too lazy to search for it), but only DX8 -> DX9 migration happened in the meantime.

This is all JFYI (it sounds a bit like I'm patronizing you or something, sorry for the tone, pick up the facts please), as it seems you are somewhat new here, and not accustomed to speed of LFS development and you somehow are asking for fix for something what is not truly broken (wine works well so far, and Scawen is keeping LFS win-api usage intentionally minimal to levels where older windows and wine should be safe), while there are major game changes on-going for last 10 years (new tyre physics), blocking pretty much everything else, for a bit too long. Patience... (it still makes me a bit curious, why impatient people didn't meanwhile fix TORCS to be better simulation, all it needs it's just forking it and fixing everything you wish...)


EDIT: BTW, the DX11 support in wine is hot news (in the LFS development time-scale), when Scawen was last time touching the rendering code, the DX11 support in wine in prototype stage, basically not working, and you had to have development branch of wine, building it from sources, so that's why he did target DX9 and not DX10/DX11, to keep LFS somewhat linux compatible (plus the API change between DX9 -> DX11 is substantial and it would remove also WinXP users, not just wine, and DX9 has enough features to provide reasonable graphics, so there's no hard push for particular feature of DX10 or DX11 API, which would suddenly magically make LFS look/work much better, DX9 is quite sufficient for the moment).

EDIT2: hmm.. and just in case you missed it, Scawen is the only programmer on the project, so maybe do search for all his forum posts, to get better idea how LFS is being developed, and what are the future plans. Also it's often quite informal read, and one can learn a thing or two. So while your request would make quite some sense for somewhat larger team, where one part may start to work on new 3D engine, in LFS case working on Vulcan port would probably stop everything else. And if it ain't broken...
First, English is not my native language. Maybe I wouldn't have expressed what I wanted to say as well as I'd like to.

Quote from Ped7g :... and? I mean I don't have any wheel at home, no HW. (I'm not talking about driver issues ... Once you don't have HW, you don't have driver issues, it's magical, everybody should try that some time...)

It's just an informative comment, I didn't mean you had to have a wheel.

Quote from Ped7g :I don't see any practical reason to prioritize yet another render API change over these two major points, so I guess your request will be considered not sooner that at summer 2020 (after cars-patch, which I guess will be released somewhere in 2020..2022)...

I'm not saying Scawen has to make the switch to another API NOW. I just wanted to clarify that the most reasonable step to support as many platforms as possible is to use Vulkan. Obviously the aspects you are commenting on are more interesting and prioritary.


Quote from Ped7g :Patience... (it still makes me a bit curious, why impatient people didn't meanwhile fix TORCS to be better simulation, all it needs it's just forking it and fixing everything you wish...)

https://sourceforge.net/projects/speed-dreams/
Quote from leillo1975 :...Opening the code would obviously give them a lot of help for their project...

Quote from leillo1975 :
https://sourceforge.net/projects/speed-dreams/

Ironic!
Anyway, I don't think adding people to the team would be productive.
https://en.wikipedia.org/wiki/Brooks%27s_law

Quote from Scawen :
Our motivation is more on the side of what interests us, rather than what makes the most money. We're not that type of capitalists at all and don't want to manage a team of developers. That is totally out of the question.

Quote from leillo1975 :
Personally I would be thinking about updating the graphical API to a more modern and multiplatform.... Vulkan is the only way....

Why? So that I can run LFS on a device that doesn't support my wheel?

Quote from Scawen :
Anyone who knows anything about DirectX understands that DX9 has enormous power that we are hardly using yet. It is absurd to suggest that it is not good enough for LFS.

hm, I would *love* to see LFS code base being more portable too, so migrating to Vulcan, if it would be for free? Hell yeah!

I just don't see how it can happen in foreseeable future with current LFS development and situation, as there are clearly lot more important things on the agenda, and Scawen does not plan to enlarge the dev team or to open source it...

And at this moment everything works (also in linux), so I guess this is unfortunate request at *this* moment. Would the situation change, I'm all for it too... Especially as having "wine" working on my box is always a hassle, I basically stopped playing any games which don't have native linux support, the LFS installation was just to test if it still works (I mean "wine" installation, and it took 2-3 hours to get it into state which I wanted ... using distro without official 32b support will probably become a bit of pain in 12-24 months, at least for closed-source SW from 3rd parties...).
This suggestion comes up every few years, and it's always nice to see Linux enthusiasts showing interest in LFS.

I doubt LFS will ever go open source. Unless, perhaps, ScaViEr decides to abandon the project for some reason. Porting it to- or supporting open platforms and APIs is a different matter, though. Perhaps that'll happen eventually, but as already pointed out in this thread that is currently not a priority.

Meanwhile ...

I know that the original poster was asking for a native build of the game and not instructions on how to run it in Wine, but if someone else is interested there are articles on the LFS Wiki that describe how to set it up.

Please bear in mind that these articles are a couple of years old and that they only cover Debian-based distros.
Vulkan also isn't a silver bullet and support is lacking for users of older GPUs (which is a reasonable part of LFS' target market).

You need an AMD GPU that supports GCN or a pretty modern nVidia GPU. Intel AFAIK has 0 Vulkan support.

Even on my fairly recent laptop, I had to recompile my kernel to enable experimental support for the AMDPRO driver to be able to support Vulkan.


Open Source is also not a silver bullet to improvign a project.
Today I was testing Ubuntu Budgie Linux distribution, opened the Gnome software centre and I saw LFS in it, got so much excited, rushed on LFS site right away to see news about Linux port, but there is no any news about it. So I downloaded this app to see what is this, seems like some kind of a launcher with Wine preloaded, it downloads the game, sets up wine and you ready to play. The game works exactly like on Windows
Attached images
Clip2net_190104180913.png
Clip2net_190104181010.png
Quote from numbazZ :Today I was testing Ubuntu Budgie Linux distribution, opened the Gnome software centre and I saw LFS in it, got so much excited, rushed on LFS site right away to see news about Linux port, but there is no any news about it. So I downloaded this app to see what is this, seems like some kind of a launcher with Wine preloaded, it downloads the game, sets up wine and you ready to play. The game works exactly like on Windows

Nice to hear that it works exactly like on windows! And yes, you're correct, the game is not native, but is run with wine. Some more information on why LFS is available in the software center is available in the LFS in the news thread.
Seems like the author of package could have been more vocal about doing this on his own. "The game works exactly like on Windows" - and that's partially also thanks to how Scawen does develop LFS, taking into account older versions of Windows and wine users. I have to admit, while I would rather love to see native port, his approach actually *does work*, and allows him to focus on development of the simulator itself.

To Devs: Linux Support
(16 posts, started )
FGED GREDG RDFGDR GSFDG