• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

Ars Technica: The state of Linux gaming in the SteamOS era

http://arstechnica.com/gaming/2015/02/the-state-of-linux-gaming-in-the-steamos-era/1/

But there's one primary reason that Linux gamers can enjoy nearly 1,000 professional, commercially distributed games today, and it goes by the name of Valve. "At the end of 2013, when Valve released the beta of SteamOS everything changed," Dean said. "After years of promoting the various Linux distributions, we had a major gaming company not just porting their games to Linux, but actually creating their own Linux-based operating system. It was an incredibly exciting moment and a turning point for Linux users."

Holding up a steady target

Croteam had dipped its toes into the Linux waters as far back as 2001 with a port of Serious Sam: The First Encounter. Now, a Linux version is an included target in every project the developer works on. Getting to that cross-platform default took some incremental work over the years, Ladavac says, as the company phased out programming and graphics tools dependent on the Windows-exclusive DirectX and Direct3D. These days, Ladavac says making PC games that are truly platform agnostic isn't hard for the company. "Since we already have OpenGL on Windows, the same OpenGL works in 99 percent of the cases on Linux, so that’s not really much of an issue."

2K shows additional Linux support with new port of Borderlands 2.
The team at Aspyr Media is in the same boat. The company known for porting many popular Windows games to the Mac has released six titles for Linux since the SteamOS announcement, including ports of major 2K Games titles in the Borderlands and Civilization series. “The Mac guys really do 90 percent of the work for us, as far as getting everything on OpenGL," Aspyr Senior Linux Engineer Ian Bullard said. "The first [Linux version] we did was a lot of work, but the port time has gotten a lot shorter the more that we do. Every single title gets faster [to port] now."

Performance problems

"The official closed-source drivers for AMD graphics cards for one example are notoriously bad on Linux," Gaming on Linux's Dawe said. "Their open source driver effort seems solid, but [it] has a little way to go before it hits that golden performance target... Unity engine-based games also seem to get lower performance on Linux, and considering how many developers use it now, that is a concern of mine."

Without optimized drivers and development tools, though, those performance gains are going to be hard for the average developer to see. "If someone specifically went and optimized on Windows and optimized on SteamOS, I think they could get more frames per second on SteamOS, but it is a major undertaking that would require a lot of people to change I think," Aspyr's Bullard said.

Still a niche within a niche

For all the excitement and effort from Valve and others, and the explosion of available games over the last few years, the Linux gaming community is still relatively small. Those I talked to in the community estimated the entirety of the Linux gaming market is still one to two percent the size of the Windows gaming market. That's up from essentially "zero percent" a few years ago, but it's still a small sliver from a business perspective.

Aspyr VP of Publishing Elizabeth Howard says the Linux ports her company has worked on generally sell about half as well as those same games ported to the Mac. That can make getting approval to work on a Linux version an uphill battle, she said. "If [a publisher thinks] the Mac already isn’t worth the investment, then Linux is even tougher."

"Unfortunately until Valve fully releases their OS and quality Steam Machines start shipping [the availability of AAA games on Linux] is not likely to change." Dean says. "I suspect the decision whether to port a title to Linux/SteamOS is more down to market share rather than technical difficulties."

Getting through "Valve Time"

While Valve has unquestionably built a viable Linux gaming market from practically nothing, the company's lackadaisical development timeline might be holding the market back from growing even more. In the last year, the initial excitement behind the SteamOS beta launch seems to have given way to "Valve Time" malaise in some ways.

“It might look like that from completely outside perspective, but from what I see I don't think they've stalled, really," Croteam's Ladavac said. "It’s being worked on within different fronts... There’s still lots of small details you can't really announce, 'We now have 100 little bugs fixed,' [for example]. It’s not just things in Steam itself, but constant efforts with [independent hardware vendors] to fix drivers, things on the distribution side so the OS comes with the appropriate libraries. There are a lot of things going on, and everything is improving slowly, but it will take some time before you can see something [officially announced]."

More at the link
 
If you Half-Life 3 it... They will come.

Half-Life 3 wouldn't do anything without it being exclusive and Valve categorically ruled out SteamOS (i.e. Linux) exclusivity. Even if HL3 were announced as a Linux exclusive, you'd have to wonder just how many people would actually go to the effort of switching versus just waiting for the inevitable Windows/OSX/console versions.
 
Half-Life 3 wouldn't do anything without it being exclusive and Valve categorically ruled out SteamOS (i.e. Linux) exclusivity. Even if HL3 were announced as a Linux exclusive, you'd have to wonder just how many people would actually go to the effort of switching versus just waiting for the inevitable Windows/OSX/console versions.

I wonder if there will even be a console version - assuming HL3 ever happens at all.

I can see it being PC (not linux/Steam OS) exclusive.

Regardless, they've still have a long ways to go with Steam OS. Open GL next needs to be a thing, and more major releases need to be available on that OS.
 
Half-Life 3 wouldn't do anything without it being exclusive and Valve categorically ruled out SteamOS (i.e. Linux) exclusivity. Even if HL3 were announced as a Linux exclusive, you'd have to wonder just how many people would actually go to the effort of switching versus just waiting for the inevitable Windows/OSX/console versions.

If I were Valve and I wanted to make SteamOS take off while not invalidating my original statement that Half-Life 3 would not be exclusive to SteamOS...

I'd never release Half-Life 3. I'd rename it Half-Life Tribus... And then I'd make it exclusive to Steam OS...
 
If I were Valve and I wanted to make SteamOS take off while not invalidating my original statement that Half-Life 3 would not be exclusive to SteamOS...

I'd never release Half-Life 3. I'd rename it Half-Life Tribus... And then I'd make it exclusive to Steam OS...

Valve didn't say Half-Life 3 wouldn't be SteamOS exlcusive. They said none of their games are going to be SteamOs exclusives, period.
 
As a long time Linux user, I have to say that Valves contribution has been more than satisfactory. I can open up Steam under my Mint installation and find about 75 games to play out of my library of about 240-something games. That's not a bad percentage. I could not imagine seeing that many games running natively like 5 years ago.

Yeah, there's still issues with performance problems and driver support. Nvidia has always been solid on Linux, but Intel and AMD could still use work. AMD's open source drivers are making some progress though, but there could still be more done.

The way Valve is pushing glNext, I expect the Source 2 engine to highlight this heavily.
 
HL3 timed exclusive if they're smart. Even 1 month would be enough. People wouldn't be able to handle it... they'd break after an hour. They said no exclusives, but didn't say no timed. An exclusive in the wrong place at the right time can make all the difference in the world.
 
I think the key thing to remember here is Valve is playing the long game with SteamOS. As the article said, they are already gaining traction yet they've barely pushed SteamOS beyond the hardcore niche crowd. I would imagine as the years go by and Steam Machines evolve you'll see that number slowly go up and up. Ten years from now we could have a completely different landscape.
 
SteamOS is going to be a big deal in the long run, as a very casual PC gamer I'm so excited that Valve is trying to break PC gaming away from Windows. I can live without having all the Windows Steam games on SteamOS at the beginning, with Valve taking SteamOS seriously I'm sure they'll come over time. I'm looking forward to 2nd and 3rd gen Steam Machines with the Steam Controller bundled in.
 
Valve didn't say Half-Life 3 wouldn't be SteamOS exlcusive. They said none of their games are going to be SteamOs exclusives, period.

Well there goes that.
I'd question their business acumen, but you know... Steam kind of prevents me from doing that.
 
If you Half-Life 3 it... They will come.

I think this is fundamentally pointless. HL2 was only available on Steam in its day, and sure it help give Steam a little kick, but once the novelty wore off, many including myself did not return - it was only a temporary kick. Fast forward to the Orange Box (not exclusive) in 2007, and that actually got me to have a better experience of Steam since it had improved - but I still left after that, again. It was the Christmas sale that year that got me stuck in when I actually bought none Valve content and had a brilliant time.

That is a similar outlook that I see for SteamOS as it goes forward. Half Lift 3 would not solve anything. SteamOS is not going to be anything other than a niche for multiple years - but this is something entirely expected by Valve as well. The real question of if they can actually be successful over time will be interesting to see.

The hope that SteamOS shows now, is as a supplementary devices particularly for indie titles and some other PC only content. To be honest, I am really surprised at how well Linux has done to this point at almost a 1000 games, and many of the most played PC games a part of that.
 
If I were Valve and I wanted to make SteamOS take off while not invalidating my original statement that Half-Life 3 would not be exclusive to SteamOS...

I'd never release Half-Life 3. I'd rename it Half-Life Tribus... And then I'd make it exclusive to Steam OS...

Exclusivity is not Valve's business strucuture. It's much more the complete opposite.

Doug Lombardi, Valve's longtime public relations guru, says that simply wouldn't be the company's style. "It's against our philosophy to put a game in jail and say it only works on Steam Machines," he told The Verge in an interview at the company's headquarters. "It works best on Steam Machines, perhaps ... but to get to exclusives or anything else just wouldn't be our style," he says.

When we bring up the fact that Steam got a huge wave of adoption when the company locked Half-Life 2 to the digital distribution platform, Lombardi admits the precedent, but hints that the company ended up regretting that decision. "That may or may not have been a good idea given the condition that Steam was in at the moment," he says.

Exclusivity by itself does not validate a platform, especially in a modern world where any form of digital media can be in theoretically infinite supply, and especially not when devs can make Linux ports in Unity without any knowledge of how Linux works.
 
'the SteamOS era'? How many people actually use SteamOS?

ymK1IMd.png


Not enough to show up on the hardware survey (though AFAIK SteamOS may be part of 'Linux 3.10' if it doesn't have lsb_release installed.)

Probably not that many compared to any other distro but all Linux users benefit from SteamOS.

Definitely. I'd even argue that competition in the PC gaming market is generally a good thing, I think even people on Windows benefit. Having an alternative that might even produce better framerates will hopefully motivate MS to improve their OS (and DirectX) as well.
 
They can work on it all they want, I won't use it. It does seem most of the hype surrounding it all has died down.. especially since MS seems to have gone the opposite way with Win10 and their attitude towards any kind of walled garden approach to their OS.

I think if anything SteamOS just put a "little" pressure on MS to keep things open.. but I really don't think it did much there even... I honestly don't think MS was ever planning on closing the OS to just the Windows App store anytime in the foreseeable future.. which really just means I guess if you are looking to build a gaming PC and don't have a copy of Windows or are too poor to buy a copy you'll have an alternative at some point.. but in my eyes I'd just rather stick to the windows environment and hope that there's more pressure on MS to make boot screens and support for things like the Alienware Alpha or Steam-type machines in the future that allow good controller support for doing basic OS functions. I'd like to see the TV/Living Room based gaming PC become more of a market..
 
As a long time PC gamer, I appreciate all that Valve has done, but I'm highly skeptical about the Steam OS and Steam Boxes. I just don't understand what market are they catering to.

Current PC users already have a "Steam Box", it's called a PC. For those that want to sit on a couch and play PC games on a TV, it's already possible to do so. Console gamers are not going to get a another $400-500 box running SteamOS to play indie games or old AAA ports.

I don't know. I guess time will tell.
 
For those that want to sit on a couch and play PC games on a TV, it's already possible to do so.

Yes, for people really into PCs it's possible. There are some people who want to order a box from Amazon, plug it into the TV, plug in a controller and for it to "just work." For now that is consoles domain but there is nothing wrong with Valve testing the waters in that area either. I know quite a few people I could have convinced to buy a Steam Machine had they been available. Telling them to go to NewEgg and put together a PC based on the NeoGAF thread is not something they are interested in. Most ended up getting a PS4.
 
They can work on it all they want, I won't use it. It does seem most of the hype surrounding it all has died down.. especially since MS seems to have gone the opposite way with Win10 and their attitude towards any kind of walled garden approach to their OS.

But they didn't, they're putting more into the Windows Store with W10 than before.
 
They can work on it all they want, I won't use it. It does seem most of the hype surrounding it all has died down.

Considering the OS isn't even in an easily available or installable state, and all the components remain officially unreleased and with so little info, and the small niche it intends to serve, I have no idea why you expect hype.

I dream of a windows-free gaming future. Hail Linux-SteamOS.

I'm kind of half way. I want them all to be competitive and pushing each other to improve and keep our custom.
 
Yes, for people really into PCs it's possible. There are some people who want to order a box from Amazon, plug it into the TV, plug in a controller and for it to "just work." For now that is consoles domain but there is nothing wrong with Valve testing the waters in that area either. I know quite a few people I could have convinced to buy a Steam Machine had they been available. Telling them to go to NewEgg and put together a PC based on the NeoGAF thread is not something they are interested in. Most ended up getting a PS4.

Fair enough, but if that's their target audience, then they will need to set and enforce precise hw specs and maybe even lock down part of the OS. Otherwise I cannot see how they can maintain a consistent, trouble free, user experience between Steam Boxes sold by many different vendors with all kinds of different hw variants.

Also, the same people you mentioned, will not want to deal with tweaking, drivers updates, etc. like the average PC gamer does. I'm no expert, but the only way I see it possible is to pretty much make the Steam Box like another console, which is what Valve does not want to do, as I understand.

This is why I'm not sure where they are going with this.
 
I always thought Valve would develop some kind of super-intuitive Cider wrapper for devs to use, but I guess that's kind of naive. Also, what happened to the "performance gains" SteamOS was supposed to bring? I think it'd be a great second OS for gaming if they could deliver on that.
 
Fair enough, but if that's their target audience, then they will need to set and enforce precise hw specs and maybe even lock down part of the OS. Otherwise I cannot see how they can maintain a consistent, trouble free, user experience between Steam Boxes sold by many different vendors with all kinds of different hw variants.

Also, the same people you mentioned, will not want to deal with tweaking, drivers updates, etc. like the average PC gamer does. I'm no expert, but the only way I see it possible is to pretty much make the Steam Box like another console, which is what Valve does not want to do, as I understand.

This is why I'm not sure where they are going with this.

Hmmm.. last time I checked I've had zero problems with my PC as a gaming device for... well for a long ass time.

If they have some base spec to work toward, that's all that's needed to achieve very good compatibility stats.

Now, developers can surely drop the ball, so it might indeed require Valve to get it's hands dirty and do at least a cursory QA check.
 
Fair enough, but if that's their target audience, then they will need to set and enforce precise hw specs and maybe even lock down part of the OS.

I think the way they are doing it is that SteamOS by default just boots into Big Picture Mode, and you have to change some settings to access the underlying desktop. So you are not locked out, but you have to choose to fiddle with things that might break compatibility. Also, if you really want to break the system, you need root access, which at least currently requires a terminal.

Otherwise I cannot see how they can maintain a consistent, trouble free, user experience between Steam Boxes sold by many different vendors with all kinds of different hw variants.

Also, the same people you mentioned, will not want to deal with tweaking, drivers updates, etc. like the average PC gamer does. I'm no expert, but the only way I see it possible is to pretty much make the Steam Box like another console, which is what Valve does not want to do, as I understand.

I don't think the average PC gamer does all that much tweaking. I'm on Linux, and I certainly don't (unless I want to, because sometimes I enjoy tweaking).
I don't think different hardware or driver updates are that much of a problem either. I've used the same LiveCD to install Linux on different systems without problems. Of course, most Linux distros don't use proprietary drivers, but that's easy to change to get the better performance. And updates just pop up a notification that an update is available. Same as consoles, except I can choose to install the updates later. Having a package manager makes this easier, since all software and updates run through the same interface. So you don't have to go to nvidia to install your nvidia drivers, go to MS to install OS updates, etc. Valve can just put all that into their repository.
 
Half-Life 3 wouldn't do anything without it being exclusive and Valve categorically ruled out SteamOS (i.e. Linux) exclusivity. Even if HL3 were announced as a Linux exclusive, you'd have to wonder just how many people would actually go to the effort of switching versus just waiting for the inevitable Windows/OSX/console versions.
I think that a lot would actually go and 'switch' - meaning that they would install SteamOS on a separate partition in addition to their main Windows one.

But saying 'a lot' on PC doesn't really mean that much for games like HL. And that is the main reason why I don't see Valve doing anything as a SteamOS exclusive. It just won't help much at all.
 
I don't know how they will ever gain big market share, I can't think why I would switch away from Windows when I could happily kill many birds with one PC stone or have a console under the TV. Microsoft aren't sitting still either. Performance gains would have to be astronomical, VR special treatment. Maybe it will be a very slow burner and carve out a small slice eventually but I can't see that happening any time soon or a tipping point. Microsoft would have to implode.
 
Fair enough, but if that's their target audience, then they will need to set and enforce precise hw specs and maybe even lock down part of the OS. Otherwise I cannot see how they can maintain a consistent, trouble free, user experience between Steam Boxes sold by many different vendors with all kinds of different hw variants.

Also, the same people you mentioned, will not want to deal with tweaking, drivers updates, etc. like the average PC gamer does. I'm no expert, but the only way I see it possible is to pretty much make the Steam Box like another console, which is what Valve does not want to do, as I understand.

This is why I'm not sure where they are going with this.

These are indeed issues that they need to solve, but I think from what has been suggested by them previously, these are things they intend to have solutions for when they launch. Not sure how they will implement them, but interested to find out and see if they can do something similar on Windows
 
I don't know how they will ever gain big market share, I can't think why I would switch away from Windows when I could happily kill many birds with one PC stone or have a console under the TV. Microsoft aren't sitting still either. Performance gains would have to be astronomical, VR special treatment. Maybe it will be a very slow burner and carve out a small slice eventually but I can't see that happening any time soon or a tipping point. Microsoft would have to implode.

I don't think it's necessarily about beating windows market share among PC gamers. It's about making complementary devices more viable and cheaper.

I personally wouldn't be opposed to spending console level money for a gaming PC that worked right out of the box on my TV the way a windows machine probably never will.

If Valve can convince develoeprs and third party engines to support OpenGL Next, well, that cover sLinux/Mac Os AND windows in one fell swoop.
 
I've been using GNU/Linux since 1996, and there are more high budget games coming out in this sliding eighteen month window than there has been on the platform in the entirety of its existence up to this point.

Yeah, some of the releases, like Dying Light, have shaky performance or significant bugs. But if this pattern of major game releases on my platform of choice winds up being a trend and not a fad, I'll be absolutely thrilled, even if the releases aren't quite as well supported here as they are on Winbloze.

As a partial aside, watching a Microsoft CEO-- Satya Nadella in this case-- giving a presentation with a giant "Microsoft heart Linux" logo in the background, followed by a release of pretty much the entire .Net stack as Free Software, is pretty surreal. Then when you combine it with the gaming situation... well, let's just say that it's a dream come true really.

The slow buildup of SteamOS reminds me of Android's first couple of years; people kept saying "Android?? Is that even out yet?", just like people are about SteamOS now. Then all of a sudden, Android started to snowball, and before you knew it, it had flown by iOS in market share. I don't know if this SteamOS thing will wind up the same way, but the situation feels eerily similar to me.
 
Don't get it. I mean if PC gaming is that important to you, are you really not going to install a Windows partition? Maybe in some countries that's not possible but overall it seems like an easy decision to make.
 
They better be aiming for feature and performance parity, otherwise....

We need an API that isn't reliant on a particular OS in order to function. DX9 is old as shit, but it's still the defacto standard because so many people (particularly in places like China) are still using it.
 
I hope HL3 is one year timed exclusive for SteamOS.
 
I don't know how they will ever gain big market share, I can't think why I would switch away from Windows when I could happily kill many birds with one PC stone or have a console under the TV. Microsoft aren't sitting still either. Performance gains would have to be astronomical, VR special treatment. Maybe it will be a very slow burner and carve out a small slice eventually but I can't see that happening any time soon or a tipping point. Microsoft would have to implode.

They won't, ever.

Even with HL3 as a timed exclusive I don't think people would want a new OS. SteamOS is and has always been a really bad idea.
 
Last time I tried Linux for gaming AMD driver issues were driving me crazy.

I really like Windows and I even got used to Windows 8, but I like the idea of an OS like Linux much more, but it needs to work well. I get my Windows for free from my university, it must suck to have to pay for it.

SteamOS as a secondary OS optimized for games have a lot of appeal to me.

So time until Linux gaming becomes a thing = (Linux_Optimized_Drivers_Time + Valve_Time)/Half_Life_3_Release_Date?

Did you just divided by 0?
 
I think that a lot would actually go and 'switch' - meaning that they would install SteamOS on a separate partition in addition to their main Windows one.

I'm sure, of all Half-Life fans, some of the more diehards would. But the vast majority wouldn't. That's the point -- exclusivity wouldn't actually cause any sort of meaningful uptake.
 
Top Bottom