Valve Seems To Be Working On Tools To Get Windows Games Running On Linux

DooKey

[H]F Junkie
Joined
Apr 25, 2001
Messages
13,473
According to the folks at Ars Technica, Valve may be working on a tool called Steam Play that will help some Windows games run on SteamOS. They base this speculation on a Reddit thread that showed Steam GUI files that talked about Steam Play. This tool will have a list of supported titles and will also allow for testing of games that aren't on the list. This sounds like a pretty good tool, but nothing is firm yet and no one knows how it's going to work. However, if Valve can get this out in the wild it might move more people over to SteamOS.

With all that said, some caution is warranted before getting too excited about these possibilities. For one, we don't know what specific form Steam Play will take. Valve could simply be preparing a wrapper that lets users run existing emulation tools like Wine and DOSBox on top of SteamOS without actively advancing the state of that emulation directly. (Valve has not responded to a request for comment from Ars Technica).
 
About fucking time Valve does this. I've been saying this for years, in that you're not going to get everyone to port their games over to Linux. Some will, but most won't. The only way to get Windows games working on Linux is through Wine and Wine is a big piece of shit. If Valve open sources it, even better. Maybe Blizzard and Origin can think about using this tool to get their games working on Linux as well, cause God knows they ain't doing shit.
 
If they can get compatibility numbers up that would be nice. I'd love to be able to switch off of Windows for gaming. Definitely won't happen in the currently climate though. This would be a good step. Some improved display drivers would be another good thing.
 
Good thing for Linux gamers no doubt. Not sure how effective this would be in getting more people to game on Linux. Then there's the potential of demotivating developers to do native Linux ports if they can sell to Linux gamers without the effort of native development.
 
Good thing for Linux gamers no doubt. Not sure how effective this would be in getting more people to game on Linux. Then there's the potential of demotivating developers to do native Linux ports if they can sell to Linux gamers without the effort of native development.

I don't really care about the demotivation factor since they don't seem to be all that motivated in the first place. (as long as these tools actually work) :D
 
About fucking time Valve does this. I've been saying this for years, in that you're not going to get everyone to port their games over to Linux. ... If Valve open sources it, even better.

Yeah, I really hope they are coming up with an open source framework or package that can be deployed on other Linux distros. There would be very little reason for me to get a Windows system if they could make more games work on Linux.

Good thing for Linux gamers no doubt. Not sure how effective this would be in getting more people to game on Linux. Then there's the potential of demotivating developers to do native Linux ports if they can sell to Linux gamers without the effort of native development.

I don't see how that is even possible. There is demotivation now because of the lack of Linux gamers in general. If Valve makes this successful and you get a decent increase in Linux gamers, there would be even more reason for developers to pay more attention to developing on Linux. The problem with any tool is that it isn't always compatible with the latest games. Developers go off of the number of potential clients when making decisions about platforms. If a lot more people start using Linux, they have all the more reason to code for Linux.
 
Would definitely consider switching my gaming box over to Linux full-time if this actually happens. Guessing that Valve sees the writing on the wall for Windows and gamers (hell, Windows and IT enthusiasts in general) so making it plausible if not easy to play titles "for Windows" on an Linux OS could convert a big number over to loving everyone's favorite red-headed step child.
 
Microsoft won't charge a subscription fee for using Windows 10 maybe then =) Because everyone would flock to Linux
 
Here's a better idea. Valve could actually develop a killer game, you know, like they used to, and develop it to be native to Linux. If they really want to spur Linux game development, make it a Linux exclusive. Make HL3 a Linux exclusive, and see what that would do for Linux game development!

I'm not going to hold my breath.
 
I don't really care about the demotivation factor since they don't seem to be all that motivated in the first place. (as long as these tools actually work) :D

I don't see how that is even possible. There is demotivation now because of the lack of Linux gamers in general. If Valve makes this successful and you get a decent increase in Linux gamers, there would be even more reason for developers to pay more attention to developing on Linux. The problem with any tool is that it isn't always compatible with the latest games. Developers go off of the number of potential clients when making decisions about platforms. If a lot more people start using Linux, they have all the more reason to code for Linux.

I think it's obvious that the biggest issue for desktop Linux gaming is the size of the user base. People will argue just how big it is but it's certainly a niche thing. I would expect most Linux gamers to see something like this as a good thing but just how it helps to promote Linux gaming beyond Linux gamers, that's not at all clear.

Maybe we see an explosion of interest in Linux gamers with significantly more playable content on Linux that gets devs more interested? That I would tend to doubt but maybe. I thinking it might generate a few more sales to the Linux gaming base that wouldn't play the game under Windows.
 
If they really want to spur Linux game development, make it a Linux exclusive. Make HL3 a Linux exclusive, and see what that would do for Linux game development!

The financial loss from that wouldn't even make the game worth developing though.
 
I think it's obvious that the biggest issue for desktop Linux gaming is the size of the user base. People will argue just how big it is but it's certainly a niche thing. I would expect most Linux gamers to see something like this as a good thing but just how it helps to promote Linux gaming beyond Linux gamers, that's not at all clear.

Maybe we see an explosion of interest in Linux gamers with significantly more playable content on Linux that gets devs more interested? That I would tend to doubt but maybe. I thinking it might generate a few more sales to the Linux gaming base that wouldn't play the game under Windows.
[/quote]

I know a lot of people that would switch over gaming systems because of this. Anything that moves more people to the format will increase awareness to bring more on board as well as get the attention of developers. The fact that more and more developers have already chosen to develop for Linux as well already shows they are thinking about it.
 
Here's a better idea. Valve could actually develop a killer game, you know, like they used to, and develop it to be native to Linux. If they really want to spur Linux game development, make it a Linux exclusive. Make HL3 a Linux exclusive, and see what that would do for Linux game development!

I'm not going to hold my breath.
You know you can forget about valve making a killer game. Most the talent that made the game have long gone and moved on.
 
I know a lot of people that would switch over gaming systems because of this.

For folks already using Linux sure. For gamers that already have perfectly functioning Windows games running under Windows not so much. Something like this just adds another layer on complexity and issues and is unlikely to get support from the developer.
 
For folks already using Linux sure. For gamers that already have perfectly functioning Windows games running under Windows already not so much. Something like this just adds another layer on complexity and issues and is unlikely to get support from the developer.

You are forgetting those that use Windows only because they can't run some games on Linux. That is a much larger base than you may think, even though percentage wise to the total gaming base it may still be small. It is significantly more than the Linux gamers now, and with the number of Linux gamers now, there is still consideration for developing games for Linux. If you double or tripled that number, there would be even more interest. Also if you had a much easier way to game on Linux, it would be even easier to convince more people to switch to Linux from Windows.
 
You are forgetting those that use Windows only because they can't run some games on Linux. That is a much larger base than you may think.

I've long said that Steam is a HUGE asset for Windows and keeps people on Windows. Tools like this have long existed and the results are at best inconsistent. Maybe Valve has something unique and much better in store.
 
I've long said that Steam is a HUGE asset for Windows and keeps people on Windows. Tools like this have long existed and the results are at best inconsistent. Maybe Valve has something unique and much better in store.

How? Steam also facilitates gaming on Linux a lot more and has prompted more people to use Linux gaming. So how is that such a huge asset for Windows specifically?
 
I think it's obvious that the biggest issue for desktop Linux gaming is the size of the user base. People will argue just how big it is but it's certainly a niche thing. I would expect most Linux gamers to see something like this as a good thing but just how it helps to promote Linux gaming beyond Linux gamers, that's not at all clear.
While proportionally, the number of people who want Linux native is small compared to those who don't care, the actual number could provide a good opportunity. 10 million customers (a number made up off the top of my head) is still 10 million customers, and at $20-$65 per title, represents significant actual sales. Proportionally, you may do better coding for Windows, but you leave revenue on the table if you don't code for Linux. It might be good to look at companies that do make Linux and Windows native versions of their games, and see what the sales numbers are.

You are forgetting those that use Windows only because they can't run some games on Linux. That is a much larger base than you may think, even though percentage wise to the total gaming base it may still be small. It is significantly more than the Linux gamers now, and with the number of Linux gamers now, there is still consideration for developing games for Linux. If you double or tripled that number, there would be even more interest. Also if you had a much easier way to game on Linux, it would be even easier to convince more people to switch to Linux from Windows.

I think the number is much less than you think. But outside of tech forums, which are a small representative of the computer using population and where Linux advocates are especially vocal, I don't think it is a quantifiable number when looking at the cost of development.
 
I think the number is much less than you think. But outside of tech forums, which are a small representative of the computer using population and where Linux advocates are especially vocal, I don't think it is a quantifiable number when looking at the cost of development.

Actually I think the number is larger than I have even mentioned here. If you see what I stipulated, largely I said it was people who use Linux, but only have Windows for games. I then allowed for people who also may choose to use Linux instead of Windows specifically because they could also them game on it. This being the crowd that isn't currently using Windows, but more so because they like to play games, than the overall ability of the OS. I then imply this number is large proportionally to those currently using Linux to game. I stipulate that is it small when compared to the whole subset of gamers.

However, the more you grow a resource and capability, the more people take notice and try it out. Also as for developers, you already have developers making games compatible with Linux, so not sure why you would think a larger increase in Linux gamers wouldn't have much effect...

I am also confused why you think my experience is limited solely to those in tech forums. I actually have many ties to game development as well as a fairly large reach to users who have no interest in tech forums but a large interest in gaming.
 
How? Steam also facilitates gaming on Linux a lot more and has prompted more people to use Linux gaming. So how is that such a huge asset for Windows specifically?

To date just this year, just seven and a half months into 2018, Windows has already added more game titles in Steam than Linux has for ALL of its life on Steam in five and half years. These kind of content numbers are clearly part of the motivation for Valve to support Windows clients on Linux under Steam directly.
 
Good thing for Linux gamers no doubt. Not sure how effective this would be in getting more people to game on Linux. Then there's the potential of demotivating developers to do native Linux ports if they can sell to Linux gamers without the effort of native development.
This will actually get more developers to port their games to Linux. Why? It's called support. If a lot of people are using Linux to play your game, and I'm sure Steam is showing this info to them, then that means to proper support those players it would be better to make a proper linux version than to deal with Steams Windows compatibility layer. If someone has crashing issues running this game on Linux through this tool, it would be much harder to fix it.
 
While proportionally, the number of people who want Linux native is small compared to those who don't care, the actual number could provide a good opportunity. 10 million customers (a number made up off the top of my head) is still 10 million customers, and at $20-$65 per title, represents significant actual sales. Proportionally, you may do better coding for Windows, but you leave revenue on the table if you don't code for Linux. It might be good to look at companies that do make Linux and Windows native versions of their games, and see what the sales numbers are.

If this added 10 million more active Linux gamers willing to spend money on games then sure. It's not going to be anyway near that though. The Linux fan sites Phoronix and GamingOnLinux actually have run a number of posts over this years talking to cross platform developers. The numbers for Linux can range to sometimes 10% but more often around 1%. At just a couple of percent of sales the overhead is going to eat that up quickly.
 
If a lot of people are using Linux to play your game, and I'm sure Steam is showing this info to them, then that means to proper support those players it would be better to make a proper linux version than to deal with Steams Windows compatibility layer.

And that's the key, if a LOT means more than a few percent generally speaking.
 
What I'm wondering is why Microsoft; instead of circling the wagons; they would easily come up with their own distro of Linux and make much of there software available for that.
MS was never much for innovation; it has always bought other software companies. This would be a easy way to gain a huge foot hold in Linux space.
 
What I'm wondering is why Microsoft; instead of circling the wagons; they would easily come up with their own distro of Linux and make much of there software available for that.

Considering the disdain for Microsoft in the Linux community, at least when it comes to desktop software, it's just not worth it on the desktop. And what's the point especially in the area of gaming where PC gaming means Windows gaming?
 
Lawsuit by ms commences in 5..4..3..2..
They couldn't handle the countersuits. Implementation variants are too common these days and everyone's business model is entrenched in one. e.g. DXVK is the one we're talking about today.
 
Steam os survey suggests they shouldn't bother and are just wasting resources. But the bigger issue is knowing Valve it will be a half ass attempt no matter what. Fuck it gaben, just give us HL3.
 
You are forgetting those that use Windows only because they can't run some games on Linux. That is a much larger base than you may think, even though percentage wise to the total gaming base it may still be small. It is significantly more than the Linux gamers now, and with the number of Linux gamers now, there is still consideration for developing games for Linux. If you double or tripled that number, there would be even more interest. Also if you had a much easier way to game on Linux, it would be even easier to convince more people to switch to Linux from Windows.


ya but those people aren't going to switch to Linux they want to use macs.
 
Herein lies the second largest problem, next to the compatibility issue.


The thing is, I am a Linux user. 95+% of my computer use is on Linux. I own some games in steam for which I have received the Linux version because I own the Windows version.

I don't play even the games I do own that have native Linux versions in Linux, and I probably would never even consider a game in a compatibility layer.

The thing is, games perform much worse in Linux, even when they have native versions. I don't know if it is a matter of poorly tuned Linux drivers for 3D performance or if it is the games themselves but this seems to hold true for every title I've ever tried.

This may not be a big deal for some casual platformer, but in most of my titles I am trying to get 4k60hz ultra settings. I need every last ounce of performance I can get, and thus I am not willing to make any sacrifices at all that impact performance.

The graphics quality is usually worse too.

And then for FPS games we have the mouse feel which is highly subpar under Linux

And there's the constant bugs and compatibility issues. Last time I tried running native Linux titles many of them would start on one of my side monitors instead of my large center screen, despite it being set as the primary monitor, with no apparent way to change this.

Now, this is for native Linux titles. Try running Windows titles via WINE or some other compatibility layer and it's only going to get worse.

The thing is, rebooting between operating systems in our modern fastboot pcie NVMe SSD world only takes a few seconds.

If I - a huge Linux fan and predominantly a Linux user - don't think gaming in Linux is worth it, why would anyone else?
 
Last edited:
Steam os survey suggests they shouldn't bother and are just wasting resources.

They must not have seen the Steam survey. You should probably call them and let them know!

Or maybe they know things you don't...

With Microsoft doing everything they can to ruin Windows and backwards compatibility I'm glad to see that Valve has taken it upon themselves to fix the problem. I'd rather have broad Linux support than another forgettable shooter game I'll play once.
 
They must not have seen the Steam survey. You should probably call them and let them know!

Or maybe they know things you don't...

With Microsoft doing everything they can to ruin Windows and backwards compatibility I'm glad to see that Valve has taken it upon themselves to fix the problem. I'd rather have broad Linux support than some forgettable shooter game I'll play once.

They know it, problem is they have so much money to burn they don't care. They tried to fix the problem with SteamOS/hardware. Didn't work but I'm sure this time it will, amirght? Like I stated previously, the bigger problem is they will half ass it. But just think of me as a Linux hater if it makes you feel better I guess.
 
At this point its pretty clear DXVK is valve.

That is the rumor... and based on the progress the project has made I would say there are 3 possibilities.

1) its a Valve project
2) there is one genius level dude with lots of spare time working on the project. As it gets MAJOR updates weekly.
3) more outside chance... its a google project

The talk of 20-30% performance hits are WRONG. Anyone that has been using DXVK knows that. I am seeing 5-10% performance hits OR GAINS. Depending on the title. Yes... in a few games I'm seeing small performance bumps not hits. Someone is seriously funding rapid Wine/DX over Vulkan development... and in less then a year DXVK has went from an idea to being able to run damn near every current AAA windows DX 11 title with next to no performance hit. If Valve adds a reliable one click front end which is what they do for these newer conversion techs... the case for Linux/SteamOS gaming is going to get very strong.

DXVK just expanded to cover DX 10... and they are planning DX 9 -> Vulkan in the near future. They also just recently added per game configuration files.
If it is a valve project. (and even if it isn't its Open source, Valve can use it even if they didn't directly create it) The pitch to developers is very simple. Instead of creating a Linux port, just create a SteamOS config file for your windows binaries.

Games like overwatch witcher ect are all running very reliably for Linux gamers with DXVK right now. Add an easy button for the not so technically inclined and Valve may have something much more compelling.
 
With Microsoft doing everything they can to ruin Windows and backwards compatibility I'm glad to see that Valve has taken it upon themselves to fix the problem.

Whatever issues Windows 10 has, game compatibility I don't think is one of then. The articles that [H] did recently on older games on new systems was kind of cool.
 
I'd jump at the chance to run my games under Linux. I'm one of those who keep W10 only because of the games, I don't do any productivity (MS) software, I do email & youtube/netflix. So if the performance hit isn't more than, say 20%, I'd go for it. I can make that up with a hardware upgrade.
 
Back
Top