Elite dangerous for Linux

I am currently running on Xbox, but I am not satisfied. The lack of Journal files and mor trivially the lack of a keyboard to text & search is terrible.

I do not own a windows system.

Is there a chance that in the future we could have a Linux version?

It would be reeeeeallly a great thing.
 
You'd be better of dual booting your PC with Windows and Linux. I doubt FD will put the manpower behind making another version of the game to maintain.

I'm quite surprised FD are not dropping support for Apple, as it seems Apple do not want to actually make a useful PC ever. Lol

CMDR Cosmic Spacehead
 
Right now, ED is the only thing keeping me using Windows as my main OS. I'd love it if FDev would make a version of ED for Linux - I can't believe that the kickstarter stretch goal was for a Mac version, not a Linux version, I'm sure there are more people who would pay extra for a Linux version than a Mac version.

I know this has already been said many times, but converting ED for PC to use Vulkan instead of DirectX would mean the game could run on Windows and Linux with minimal extra effort (less than needed for PS4 or maybe even XBox).

If FDev did another crowd-funding round for a Vulkan/Linux conversion of ED then I'd put my money where my mouth is :D
 
Me, too.

I want to buy a new copy of ED, that is I want to spend my money... anyone listening?

There are plenty of games running on ubuntu. Even fedora has the same games, so... you have no excuses.

I already bought more than a copy of ED and the Lifetime Pass, so what have I to do more for David Braben's wallet? :)
 
IMHO: It will be easy 2 do, if change graphics API to OpenGL\Vulkan. This is not easy indeed. And graphics drivers 4 linux are only getting out from higly to simply (self-censored) state...
 
IMHO: It will be easy 2 do, if change graphics API to OpenGL\Vulkan. This is not easy indeed. And graphics drivers 4 linux are only getting out from higly to simply (self-censored) state...
It is not quite as simple as that. that is just one of the pieces to manage, then you have the whole sound API, keyboard/mouse API and so on.... so there is quite a few more bits to change than just go on than to change the Graphics API.


I do not know how much work they would have to put in to make it run with WINE.
 
I was wondering if there are ways to make it work in wine that would not take alot of effort. Since people say shaders are the problem waht about a retro graphics mode with older looking graphics(even from the original 1984 game) that can act as a startable mode for troubleshooting and compatibility to get it to work with as many systems as possible. And then build specific graphics options or other things later with more neutral settings that are either added or strippe down version of the current graphics if needed. Is there a simple way to do it? What else is missing to get unnoficial support through wine?

They could even use the graphics mode to let you play the other games in this universe as an added bonus and for easter eggs and whatnot either solo or with friends.
 
Last edited:
It is not quite as simple as that. that is just one of the pieces to manage, then you have the whole sound API, keyboard/mouse API and so on.... so there is quite a few more bits to change than just go on than to change the Graphics API.


I do not know how much work they would have to put in to make it run with WINE.
As far as sound goes... Linux has a very wide range of options available to it.

Keyboards are standard... controllers might be an issue.

And WINE is an emulator... not a platform. It would probably work in WINE as is. What we want is a full fledge PORT to Linux.
 
If we can get wine working faster then why not wine while we wait for a potential port. I'd like to play the game now. Maybe it could also be used as a test bed to get it working in linux to help a later port... In my experience, at least with ports on steam, you are better off sometimes with wine over the port. Sometimes the ports simply don't work and the run with your money.. and the company never does anything about it regardless of the illegal sale of non working programs...

I just read wine 3.0 is coming out. Maybe we'll get enough dx11 stuff to make this and other games work. I was technically in the demo and was in the cockpit with 2.21 staging in POL. But I couldn't figure out the controls. And the graphics were funny colors.

Edit: I was wrong. It requires dotnet4.0 and possibly xinput... And must be set to windows xp in wine(Edit: in a 32bit bottle. And I use PlayOnLinux.). And it will only let you play the demo and you must let it sit until it loads. Sometimes it will not load competely or crash. You can tell it's up when the sound starts going. I also installed from steam in pol for a windows download of the game first. So It might not install. Haven't tested.
 
Last edited:
As far as sound goes... Linux has a very wide range of options available to it.
Apart from uncle Nadella, the problem is always the same: money to spend on linux version.

I am too old to play videogames so I had little time to play ED but I am willing to pay for it, and I did.

In a word of businessmen where only money is always right, what is the problem? I already saved my money do not buying other games and I already bought my Lifetime Pass, so... any 130 pounds or more? I hope I could afford to start 2018!
 
Apart from uncle Nadella, the problem is always the same: money to spend on linux version.

I am too old to play videogames so I had little time to play ED but I am willing to pay for it, and I did.

In a word of businessmen where only money is always right, what is the problem? I already saved my money do not buying other games and I already bought my Lifetime Pass, so... any 130 pounds or more? I hope I could afford to start 2018!
So this brings in the viability of a Linux port, are there enough users to justify a full fledged Linux Port? And sadly most time from a business perspective, making Linux ports is not economical viable, so this creates a bad situation for gaming on Linux, as many who would play on Linux have to use Windows, and thus making Linux even less interesting as a gaming platform.
 
As far as sound goes... Linux has a very wide range of options available to it.

Keyboards are standard... controllers might be an issue.

And WINE is an emulator... not a platform. It would probably work in WINE as is. What we want is a full fledge PORT to Linux.

But who is going to pay for that Linux version? Where is the economics of spending the time to write it, and then to maintain it?

Look at how many clients they have already,
PC
MAC
XBOX One
PS4

And No, WINE is NOT an Emulator.


But you are totally missing the point about the API's. It is the stuff programs uses to talk to hardware and system. Because you will not write your own USB protcol driver to talk to a Keybaord you rely on the OS to do that for you. So they do. And still you do not write your own listening code to talk directly to your keyboard, the operating system does that too, so it decodes the data arriving from your keyboard and let your program and others get the refined version of the data. Ontop of this, we usually get some other API's that are used by games, on Windows we have DirectInput as one option, there other options to use as well.
The same story goes for Sound, multiple layers of software before you program can use it.
Controllers, the same

Graphics, we have two standards, DirextX, and OpenGL/Vulkan.

Now examine these two and where can play Elite today.

Windows:
DirectX (uses)
Vulcan

XBOX One :
DirectX (uses)

PS4:
PS4 specific API (problably uses)
Vulcan, but not from the start

OSX:
OpenGL, but lacks certain functions, so no Horizon (uses)



And for comparison
Linux:
Vulkan
OpenGL


So they most likely are using 3 API's: DirectX, PS4 native and a limited OpenGL implementation

So now you want to add 4th API, Vulcan or full OpenGL. And rewriting any existing client to use Vulcan/OpenGL would require even more work, and the only benefit they could get from that is possibly removal of the PS4 native API.

Now, onto the Vulcan/DirectX 12 issues. These APIs are simpler than they predecesors, ie OpenGL and DirextX 11, as they do not contain as mush calls to alot of stuff, this is because many thinks that they can write better code to things that you need specifically for you game, and not have to implement support other uses for that function. And this should give us more performance, And yet when comparing performance between OpenGL/Vulcan or DirextX 11/12 they are about the same. And most of this is because the OpenGL and DirectX 11 was pretty optimized to begin with, so it takes alot of skills to optimize this by yourself, and even if you can write more optimized version than what OpenGL and DirectX 11 had, is the rest your code as optimized? probably not, as you are not writing that code... So bad coding practices is in most cases having ALOT more impact on performance than choosing between OpenGL/DirectX 11 or Vulcan/DirectX 12.


So it is bit more complicated


But why do you want a full fledged port in the first place? What do you think you would gain from that as compared to a working Wine version?
If you say better performance, then most of the time that is not true. Have a look at Phoronix site, they have done some great work at comparing performance between Linux and Windows and Windows client running on Linux using Wine. And the normal outcome is that native Linux client is generally the same the Windows client either on Windows or on Linux using Wine! and the only times I have seen a huge difference is when FPS goes crazy high, over 170 FPS. But then what effective difference does this make in the real world? do you have a monitor that show 170+ FPS?

Another interesting find Phoronix did, was that the use of older DirectX versions in Wine or older OpenGL versions where noticeably slower by running the Windows version in Wine. But this does not current games. but an interesting side note.


And I have seen atleast one game that did have native Linux support to scrap that client in support of actually running the Windows client via Wine, as that was actually faster and new fixes, patches etc always came to the Windows clients first and later to the Linux client. Now the Linux crowd got the same updates as the Windwos crowd, and it was tested and verified to run via Wine.


So from a business perspective, I think it would be more efficient to actually take the time to adapt the Windows versions to run via Wine on Linux than to write a "whole" new client. And after the initial learning curve for the limitations of Wine etc, any further maintenance would be more cost effective, as they are only maintaining one client, that runs on two platforms. And testing done from Phoronix suggests that there should not be any performance hit by doing it this way.



And I have totally avoided the dark pit of DRM and why that crap is in most cases totally incompatible with running stuff on Linux... and sadly DRM hurts the honest players on all platforms.
 
Actually I don't mind which implementation, I just want to play in GNU/Linux. Wine, POL, native, Vulkan, OpenGL, whatever, I don't mind, I just want to play ED in GNU/Linux.

The reason? The only reason I use Windows is for ED. If I could play ED in GNU/Linux I'd drop Windows no problem.

EDIT:
Considering the market for a GNU/Linux version:

The kickstarter stretch goal was ~300k to make the MacOSX port. I imagine it would be ~the same for a GNU/Linux version.

Who would pay? Customers. Like I suggested, they could do another kickstarter, or crowdfund it like they kind of did with Planet Coaster.

Too small a customer base? Er, there are probably more GNU/Linux users who will buy ED than MacOSX users! Consider the stereotypical user: MacOSX - arty, whimsical, cool; GNU/Linux: geeky, tech-savvy, "post-cool", interested in science, particularly space!

So, I think there's sufficient market to make it worthwhile for FDev, the question is whether or not they have the appetite (I doubt it).
 
Last edited:
But who is going to pay for that Linux version? Where is the economics of spending the time to write it, and then to maintain it?

So now you want to add 4th API, Vulcan or full OpenGL. And rewriting any existing client to use Vulcan/OpenGL would require even more work, and the only benefit they could get from that is possibly removal of the PS4 native API.

Now, onto the Vulcan/DirectX 12 issues. These APIs are simpler than they predecesors, ie OpenGL and DirextX 11, as they do not contain as mush calls to alot of stuff, this is because many thinks that they can write better code to things that you need specifically for you game, and not have to implement support other uses for that function. And this should give us more performance, And yet when comparing performance between OpenGL/Vulcan or DirextX 11/12 they are about the same. And most of this is because the OpenGL and DirectX 11 was pretty optimized to begin with, so it takes alot of skills to optimize this by yourself, and even if you can write more optimized version than what OpenGL and DirectX 11 had, is the rest your code as optimized? probably not, as you are not writing that code... So bad coding practices is in most cases having ALOT more impact on performance than choosing between OpenGL/DirectX 11 or Vulcan/DirectX 12.

But why do you want a full fledged port in the first place? What do you think you would gain from that as compared to a working Wine version?
If you say better performance, then most of the time that is not true. Have a look at Phoronix site, they have done some great work at comparing performance between Linux and Windows and Windows client running on Linux using Wine. And the normal outcome is that native Linux client is generally the same the Windows client either on Windows or on Linux using Wine! and the only times I have seen a huge difference is when FPS goes crazy high, over 170 FPS. But then what effective difference does this make in the real world? do you have a monitor that show 170+ FPS?

So from a business perspective, I think it would be more efficient to actually take the time to adapt the Windows versions to run via Wine on Linux than to write a "whole" new client. And after the initial learning curve for the limitations of Wine etc, any further maintenance would be more cost effective, as they are only maintaining one client, that runs on two platforms. And testing done from Phoronix suggests that there should not be any performance hit by doing it this way.

And I have totally avoided the dark pit of DRM and why that crap is in most cases totally incompatible with running stuff on Linux... and sadly DRM hurts the honest players on all platforms.
Actually, for Vulkan you DON'T need to "add" an API. You rewrite the whole thing and it will be working under ALL platforms without anything, because Vulcan is such a low level API, that it doesn't care on what system it runs, as long as the interface is the same. Which it is on any system that supports Vulkan, which btw includes MacOS. So you would not only have potential Linux users, you'd have more MacOS users, because they finally can use Horizons.

Also, personally i - like madbilly - don't care if it's a Wine port. The launcher is working quite good with Wine 3.0, but the rest just doesn't, because Frontier has been extremely sloppy with he Cobra engine. Personally, i think it's a great engine that uses more than 4 Threads and so on, but cross plattform support is awful, or rather not even built in.

Also, considering the Steam Hardware Survey is just absolute garbage, the market for Linux is HUGE. Just look at Hitman, DeusEX and other stuff Feral Interactive put out. For me, Elite Dangerous is also the ONLY reason i've got Windows still installed as physical installation. Literally everything else works either in a VM or on Wine - or there are suitable replacements on Linux. I actually think the Linux market is hugely underestimated. Total War: WARHAMMER was a HUGE success for Feral. They doubled their value in the last year. Also, i actually know quite a lot of people that either buy the game for a Console (mostly PS4), or for Linux.

Actually I don't mind which implementation, I just want to play in GNU/Linux. Wine, POL, native, Vulkan, OpenGL, whatever, I don't mind, I just want to play ED in GNU/Linux.

The reason? The only reason I use Windows is for ED. If I could play ED in GNU/Linux I'd drop Windows no problem.

EDIT:
Considering the market for a GNU/Linux version:

The kickstarter stretch goal was ~300k to make the MacOSX port. I imagine it would be ~the same for a GNU/Linux version.

Who would pay? Customers. Like I suggested, they could do another kickstarter, or crowdfund it like they kind of did with Planet Coaster.

Too small a customer base? Er, there are probably more GNU/Linux users who will buy ED than MacOSX users! Consider the stereotypical user: MacOSX - arty, whimsical, cool; GNU/Linux: geeky, tech-savvy, "post-cool", interested in science, particularly space!

So, I think there's sufficient market to make it worthwhile for FDev, the question is whether or not they have the appetite (I doubt it).
There is a huge market. If they'd port the Cobra Engine to Vulcan, they'd have two games they could sell at least for PS4, Linux, MacOS and even a mobile Version. Also, the "prestige" in the Tech Industry for using Vulcan would be rather high.
 
Out of curiosity is the launcher still working for you? I've been testing it and it broke for me with the lastest version. Not sure why.

And, yea, if we could even minimally play in wine it would be nice until, or if, they ever do anything with vulcan. It's probably the fastest method to get us in the game. But I don't know enough to know what that takes besides messing with wine a bit through trial and error. The CRC thing sounds like it's a file checker problem from googling. And there is a problem with what seems like checking regions.. don't know if they are connected. I'm hoping something just starts working with a wine update and we can get in, even with funny graphics.

I hope vulcan becomes a normal thing for games in the future. Hopefully in the not to distant future.
 
Last edited:
Actually, for Vulkan you DON'T need to "add" an API. You rewrite the whole thing and it will be working under ALL platforms without anything, because Vulcan is such a low level API, that it doesn't care on what system it runs, as long as the interface is the same. Which it is on any system that supports Vulkan, which btw includes MacOS. So you would not only have potential Linux users, you'd have more MacOS users, because they finally can use Horizons. ...
AFAIK there is no official support for Vulkan in macOS. Only Metal (if we're talking new-generation rendering APIs only).

I have no idea if it's possible to build a Vulkan API/driver for mac without Apple's direct cooperation. Every 3rd-party library I know of that does graphics or drawing either uses built-in Cocoa (if that's it's current name, I can't remember) or OpenGL (perhaps through X11.app or something homebrew), meaning they can use Apple's built-in APIs.
 
Last edited:
AFAIK there is no official support for Vulkan in macOS. Only Metal (if we're talking new-generation rendering APIs only).

I have no idea if it's possible to build a Vulkan API/driver for mac without Apple's direct cooperation. Every 3rd-party library I know of that does graphics or drawing either uses built-in Cocoa (if that's it's current name, I can't remember) or OpenGL (perhaps through X11.app or something homebrew), meaning they can use Apple's built-in APIs.
Keyword is "Molten". Runs on Metal, it's kind of a bridge between Metal and Vulkan. apparently it runs pretty well. Also, it's been developed by the Khronos guys - those that actually developed Vulkan.
 
That's quite surprising to hear. I was under the impression that Metal was developed basically in-house by Apple. I know AMD's Mantle was the basis for Vulkan. In fact, Mantle was the first of the new generation of low overhead APIs, at least to be publicly usable. I don't really know the meat of the development timelines.

Edit: sorry I totally misread your post. There is a compatibility layer then, awesome!
 
Last edited:
Top Bottom