Gamasutra: The Art & Business of Making Gamesspacer
arrowPress Releases
April 24, 2014
PR Newswire
View All





If you enjoy reading this site, you might also want to check out these UBM TechWeb sites:


Carmack: Supporting Linux isn't worth the hassle at the moment
Carmack: Supporting Linux isn't worth the hassle at the moment
February 6, 2013 | By Mike Rose

February 6, 2013 | By Mike Rose
Comments
    24 comments
More: Console/PC, Programming, Business/Marketing



"I don't think that a good business case can be made for officially supporting Linux for mainstream games today"
- id Software co-founder John Carmack admits that he doesn't think Linux has a very bright future in video games -- at least not right now.

PC games giant Valve has been showing huge support for Linux games recently, adding Linux titles to its online catalogue, and stating that its upcoming living room PC will run Linux.

However, Carmack this week questioned whether it's actually worth bothering with Linux, when gamers can often use open source software Wine to run Windows games on the operating system.

"I wish Linux well, but the reality is that it barely makes it into my top ten priorities," he said.

He added that those businesses looking to offer Linux porting opportunities to big publishers will most likely get turned away.

"You probably can't even get an email returned if you are offering less than six figures to a top ten publisher," he noted. This may sound ridiculous - 'Who would turn away $20,000?' - but the reality is that many of the same legal, financial, executive, and support resources need to be brought to bear on every single deal, regardless of size, and taking time away from something that is in the tens of millions of dollars range is often not justifiable."

However, the industry veteran believes that there is a way that, sometime in the future, Linux could become a viable option for big publishers.

"Ideally, following a set of best practice guidelines could allow developers to get Linux versions with little more effort than supporting, say, Windows XP," he said. "Properly evangelized, with Steam as a monetized distribution platform, this is a plausible path forward."


Related Jobs

Io-Interactive
Io-Interactive — Copenhagen, Denmark
[04.24.14]

Gameplay Programmer
Io-Interactive
Io-Interactive — Copenhagen, Denmark
[04.24.14]

Generalist Tech Programmer
Crytek
Crytek — Shanghai, China
[04.24.14]

Mobile Programmer
2K
2K — Novato, California, United States
[04.24.14]

Senior Tools Programmer










Comments


Benjamin Quintero
profile image
The title sounds a bit Trolly to me. I can't help but agree with what he says though. Linux is only slightly higher than BeOS on my radar...

Lewis Wakeford
profile image
Well it's kind of a self fulfilling prophecy at the moment, isn't it?

Most PC games are windows only -> Most gamers use windows -> Most PC games are windows only -> Most gamers use windows...

Somebody (a powerhouse like Valve, basically) has to break the cycle for Linux to matter. I'm sure many people would be willing to switch over to Linux or at least try it if their games where waiting for them. That's basically what he says in the last paragraph as well. It isn't particularly hard to support Linux if you plan it from the start instead of embedding loads and loads of platform dependant code throughout your game.

Mike Jenkins
profile image
If Valve launched the "Steam Box" alongside Half Life 3 as a Linux exclusive, for example?

Lewis Wakeford
profile image
I don't think they would go that far, but if they released it two weeks earlier on Linux I can see people installing a distro just to play it early, they might stick around.

Michael Rooney
profile image
I'm not sure I agree. The average user cares more about novelty, usability and availability than games. As much as I would like to believe games have that kind of power, I just don't see it being the case. When I talk to people about why they don't want to use Linux, lack of applications is rarely in the list of reasons I'm given.

Merc Hoffner
profile image
If steam running under windows could safely create a partition and install a custom linux steam distro, that would take out a good deal of the barriers to entry, even if it is a bit redundant.

E Zachary Knight
profile image
Hey, if Carmack and his team actually contributed to Wine and helped resolve compatibility issues between their games and Wine, then I would be fine with that. Unfortunately, most developers take a complete hands off approach in that regard.

And as Lewis says above, if you consider Linux (and at the same time Mac) from the beginning of development and you don't use a bunch of platform dependent code, then you will find bring your games to Linux to be worth while and relatively simple.

But, lucky for us Linux users, more and more companies are making the choice to support Linux. This includes powerhouses like Valve as well as middleware providers like Unity3d. More will follow.

Chris Proctor
profile image
Giving id a hard time for Linux support for their games isn't going to fly, it's only with Rage that they've stopped supporting them. Seems clear that they've given it a great go and the business case hasn't existed.

http://en.wikipedia.org/wiki/Id_Software#Linux

E Zachary Knight
profile image
Yes id has had a good run of unsupported Linux binaries. However, as he says, Zenimax has pretty much killed that concept.

[User Banned]
profile image
This user violated Gamasutra’s Comment Guidelines and has been banned.

Chris Hendricks
profile image
Fortunately, now that engines like Unity can export to Linux, it's a much easier decision for many developers whether or not they can/should support it.

Dylan Tan
profile image
Yes for Developer who uses Unity, which it might help a lot (save the headache by miles) but many developer like us, don't use Unity for various reason.

Kris Morness
profile image
While I've been a hardcore Windows guy for my entire career, it wasn't until a little over year ago that I was first exposed to the Linux world. And I run two computers at home -- a Windows 7 gaming rig, and a Linux web/file server.

I do see there are a lot of people that like the ability to install their own OS without having to pay for it, and Linux is quite popular for that, and the open-source nature of everything. I can see this growing into a non-insignificant number over time. Especially for developers, some really fantastic software such as STS (Eclipse) -- I know you can get it for Windows, but it just works better with Linux. A lot of open-source development, tools, and workflows are made for the Linux environment.

My experience with Wine is that it is a terrible and brute force Windows emulator. When any particular software feature works 90% of the time, it doesn't really give you piece of mind, not to mention, I would question it's support of graphics and audio drivers and maintaining any semblance of gaming performance.

That said, I don't use Linux to play games, and don't really plan on it and I would imagine most Linux builds are more geared to home servers. The way I see it, most proper gamers yield to Windows whether they want to or not... because frankly that's where the critical mass is and has always been. I do think over time, Linux will become more of a force to recon with. My feelings on Linux is that it is a hodge-podge of software that ranges from awesome to sketchy. At least with Windows, you know exactly what you are getting.

Unity is great, and making a Linux version for the +$500 per license would be worth considering on any desktop release, and of course, iOS.

james sadler
profile image
I think the big issue with Linux reaching the masses is the fragmentation of the OS. There are just too many versions out there that product x wont work on for this reason, but works on this version. Its the same problem a lot of apps have with android. Hardware support is also lacking amongst a wide array of components. An annoying thing to is that when I have tried to download or install a piece of software it then tells me I need to go download a bunch of libraries I'm missing. Those kinds of things should be packaged with the software, but because they aren't it leads me to believe that they should be there and the OS didn't natively supply it. I have no problem doing this, but I am an above average PC user. One can't expect the average or below average user to have to do this kind of stuff.

I have no problem developing for linux though as a branch to regular releases, but exclusives just aren't worth it yet. I think it is still a sort of novelty to say "I got it running on Linux." We're still a bit away from it being a norm. Steam will probably solve that though.

Ron Dippold
profile image
It really is a tech support nightmare with all the different distros having different conventions and naming, and all the different sound systems (alsa? oss? esound? klang?) as just one example. I still do new Linux installs with well maintained distros where getting the sound to work takes text editing config files. There's just so much Direct[X] does for you other than graphics that people forget about till it comes down to the nitty gritty. Managing input devices is another.

I'm really curious to see what happens with Project Eternity having committed to Linux support through Unity. If that turns out well and Unity handles most of the sound/input/network abstraction for you then it seems like a low effort way to get a Linux version out. As long as you don't have to handle tech support for getting sound working in the first place.

A S
profile image
I think the question really is, who will benefit?

Companies don't benefit because the majority of gamers are on Windows. Adding further platforms adds both fixed and variable costs (porting, support, respectively).

Gamers don't benefit because unless companies can retain optimization experts for each platform (and let's face it, each major "family" of distros if we go Linux) we'll get generalized code which will perform worse than code with platform specific optimizations.

In the end, the only people who benefit are Linux afficiados who for some reason either won't pay 100 bux every 2-3 years to upgrade Windows and dual boot, or have some anti-MS thing.

MS could totally throw this away by making Windows a closed platform, and if they do Linux (or some other OS) will surge, not just for games, but for everyone. In that case though, we can just follow the trend. And by the way, if games go to "Linux" what that means is "either Redhat or Ubuntu become the standard". Debian, Slackware, Scientific, Centos, Fedora and any other distros, even if they are based on the above 2, will be unsupported platforms.

Michael Wenk
profile image
I don't think you understand linux. It would not need to be optimized per distro, but to the library, usually libc or libsdl to a arch type. Same thing has to happen on windows, really.

And motivations are many. I don't bloody care about the OS to be honest. I care about the apps I can run, and how easy it is to accomplish what it is I want to do. In every case outside of playing video games Linux is much better to do so.

It is also not an issue of money, tho I don't keep up with windows versions because quite frankly I have been burnt in the past. Vista/Win98 I am looking right at you.

And lastly, dual booting doesn't work. It kills momentum.

As for your answer, many people will benefit. It might even enlarge the market a bit.

A S
profile image
Thanks for the comment. I know what you mean about the library optimizations, but you can't develop for a moving target. This means you have to assume one of the more static distros and you can't offer full support for an distro that the user has customized, which is really the point of Linux. For example, as of kernel 2.5 OS scheduling timeslices went from 10ms to 1ms. This was a good change, but it caused some real subtle timing bugs that I had to deal with on the application I work on. So do you write your application for kernel 2.5+ or previous versions? It's all well and good to say it should be independent of these things, but when you can't even rely on the OS timeslice to be stable it really makes things tough to guarantee, particularly in performance critical apps like games.

It's fine to say many people will benefit, but I mean, put it concretely. I actually contend moving to Linux doesn't really help anyone except people who sell Linux. So who will actually benefit from the gaming ecosystem?

Michael Joseph
profile image
Valve may very well do for Linux gaming what Microsoft did for Windows gaming. Linux never really had a champion for gaming. the GL ARB for years seemed like it was infiltrated by the enemies of the linux desktop.

A custom Steam Linux Distro for gaming could finally give a broad base of PC developers a single target that isn't jumping all over the map.

Valve can even create an equivalent of a GL SDK to rival the DX SDK to assist windows developers making the transition.

Steam going network API in this light seems a critical first step towards vastly increasing the installed base by encouraging every single game developer to support the platform and thus every single gamer to join the steam community.

I wonder if Valve thought about buying out Unity3d? Or maybe it'll be Microsoft looking for their XNA replacement, lol.

Jimmy Albright
profile image
Personally I think if Linus ever wants Linux to be taken seriously as a gaming platform doing things like publicly giving nVidia the finger and saying "f*ck you" probably won't move things in the right direction.

That being said I think it's great Steam is coming full force to Linux and eventually everyone can enjoy their steam library regardless of their OS.

Rob B
profile image
Linus has never been a campaigner, its one of the reasons why I like him despite the fact he comes off as smug at times.

He gets on with the technical work is pleased that its doing well and unless someone actively makes his life difficult, like Nvidia has, he doesnt really care about anything else. (and has been quite open about this fact.)

If anyone is expecting Linus to lead the Linux community to polished PR heights it is not going to be Linus that does it.

Bob Johnson
profile image
Linux needs to be distributed by Valve and be part of a dual boot solution for the masses to make headway into becoming a gaming OS.

Look at what Apple did with OSX. Boot camp was a safety net that nudged many consumers over the fence. Same with the virtual machine software like Parallels or Fusion. Maybe Valve could partner with them or Virtual Box/Oracle.

Roger Tober
profile image
Because of consoles and tablets, etc, more and more games will be written in Opengl. At that point it gets easier to port to Linux. They probably will standardize on Ubuntu. Also, with the diversity of platforms for games now, adding one more port isn't as big of a deal as it used to be, especially because most games are choosing engines rather than writing them in house. For the engine writer, it's one more feather in the cap if it ports to another platform, especially if that platform is showing growth.

Ruthaniel van-den-Naar
profile image
Sorry, boys but Valve is too lazy to make own Linux distro, its is only empty wet dream. It would be nice, but chance is very, very small. Yes they can buy some comercial Linux distro, but it never works before (some company, but another and they quickly integrate themselfs and make product).
If Valve dont would be band of faerytale teller, but realy do someting, they realy could invest into good linux project in tens of millions dollars, simple donate these projects or make price money on achieving certain goal for develepers.. but think that they are too weak to do that.


none
 
Comment: