Yep, a great site and it has an interesting wiki and stats sections:
https://www.gamingonlinux.com/wiki/
https://www.gamingonlinux.com/index.php?module=statistics&view=trends
https://www.gamingonlinux.com/wiki/
https://www.gamingonlinux.com/index.php?module=statistics&view=trends
By colossal waste of money I meant the amount of money you'd have to invest in the PC and the small amount of big budget games doesn't make it a good investment for the majority of people out there. Again, it's a niche OS that appeals to a very small user base. That's why most publishers and devs don't give a damn about it. Besides your point of needing good hardware to run games that run poorly on Linux compared to Windows is not exactly a great selling pitch for the system..
I could see myself installing Linux on some old laptop and play indie games on it, but nothing else really.
I could see myself installing Linux on some old laptop and play indie games on it, but nothing else really.
Not really, you don't need official ports for everything, like I said above, a huge amount of games is playable in Wine. So getting high end PC is a worth while investment if you want to play on Linux.
It's not about the system. If you take a non native game and perform static binary translation, it's always expected to have some performance overhead, no matter what system it's translating to, that's just how it is. And translating graphics code has its own specialty, since graphics pipeline feeds the GPU, not the CPU. So it's using its own logic.
You are probably familiar with shaders, that's how GPU is programmed essentially. Shaders are compiled into some intermediary bytecode representation when shipped with games (DXBC in case of DirectX and SPIR-V in case of Vulkan). That's needed to allow them to run on any GPU, since you can't just ship binary machine code for them - each GPU has its own architecture and instruction set. That bytcode is translated at runtime into GPU machine code by the compiler that's part of the graphics backend in Direct3D and Vulkan drivers.
If you follow one pipeline, optimizations line up as developers expected. But if you take DXBC and at runtime translate into SPIR-V and then into machine code, some optimization potential is lost. That's why it's always better when developers who are porting, have access to the source code not just for the engine, but for shaders as well. They can compile it into optimal SPIR-V at the development stage and get performance which is as good as on Windows.
Anyway, Wine itself is a wonderful tool to play games that don't have native ports (especially together with dxvk and vkd3d), and some performance overhead is completely tolerable to have a playable result. It's not as big as you think.
Last edited:
Nope, that's not why. Legacy publishers simply don't care about anything new and emerging. Their mentality doesn't fit it. Luckily not all developers have such backwards thinking approach and in fact recent developers conferences had some surveys that showed that more and more developers are releasing for Linux (the graph I linked above actually confirms it as well). So there is growing interest. And legacy ones always come last, so that's not news.
Tried googling and didn't find any upcoming big game being on its way to Linux so while there might be some increased interest on the OS, it clearly doesn't show, neither in numbers or games coming to the OS.
And with Wine. Sure, I guess those games are playable, but it's hardly optimal. And you need to have official support to have the best experience. Besides, how much time does it take to get a DX11/12 game running on Wine comparable to WIndows version, without any bugs or glitches etc? Because I just googled Witcher 3 and even after 3 years it seems it's still not perfect and has graphical glitches etc.
And Linux is just not worth investing into, from a publisher's perspective. Better ways to utilize those limited resources, like bringing extra content on consoles or (windows) PC.
And with Wine. Sure, I guess those games are playable, but it's hardly optimal. And you need to have official support to have the best experience. Besides, how much time does it take to get a DX11/12 game running on Wine comparable to WIndows version, without any bugs or glitches etc? Because I just googled Witcher 3 and even after 3 years it seems it's still not perfect and has graphical glitches etc.
And Linux is just not worth investing into, from a publisher's perspective. Better ways to utilize those limited resources, like bringing extra content on consoles or (windows) PC.
In the graph I linked before, you can actually notice faster growth in the recent time. That indicates more interest from developers.
Depends on the game, but that's the bonus of Wine being an open source project. If something doesn't yet work, developers will eventually fix it, while something "officially" supported will likely lose such support not long after the release since developers who released it will focus on other games.
Not really, innovative publishers invest in it and are getting returns. Only legacy ones don't. So today it's not about business opportunities (because Linux is profitable for those who invest in it), it's about attitude.
Hehe, so limited that some of their games already have internal Linux ports (like Bethesda's) and it would cost them nothing to release, yet they still don't. It's not about covering expenses, it's about platform mentality and approach to innovation.
Yeah I noticed the increased interest but didn't see any evidence it's coming from AAA developers publishers. I don't really find games made by three blokes in an apartment building that interesting. I'm sure Linux is a great OS if you're into niche Indies. As a system for AAA games, there are just too many inconveniences compared to Windows. Not worth it.
Legacy publishers are followers, not pioneers. And they are also quite uncomfortable with anything they can't put a finger on because it's not owned by any single vendor. Quantifying Linux market is not easy, and it takes some push from distributors who are interested in it to move slumbering publishers forward. Before Valve were working with publishers more actively, so they woke some of them up, but Valve's marketing cooled down somewhat, so legacy publishers went back to sleep. I wouldn't worry about that though. Natural growth of the market will wake them up eventually.
Valve also aren't just sitting relaxed, they have some plans as well. See https://www.gamingonlinux.com/articles/valve-confirms-their-continued-support-for-linux-gaming.11530
We'll see. Afaik, Gaben has proclaimed Linux being the future of gaming for ages but I've yet to see that promise materialize into anything solid. Maybe sometime in the future with VR but right now I don't see it even becoming a competitor against Windows, let alone overtaking it. Afaik, it's even far behind Mac users in user base.
Besides, I'm really looking forward to MS and their "shared platform" with Xbox support on W10. It means games like the next Fable and Halo are coming to the Windows store and if the ports are as well done as Gears 4 on W10, we're in for a treat.
Besides, I'm really looking forward to MS and their "shared platform" with Xbox support on W10. It means games like the next Fable and Halo are coming to the Windows store and if the ports are as well done as Gears 4 on W10, we're in for a treat.
I wouldn't consider macOS to be even in the same league. Apple don't really care about gaming, at least not about demanding games anyway, since Apple's hardware is not capable of handling them. So amount of macOS gamers has little importance.
Valve's comments about Linux gaming in the past were correct, but expectations of some for them to materialize instantly were premature. It took a major effort (not just for Valve but for many involved parties) to bring open graphics ecosystem into the modern time (which also forced MS to rush with their DX12 lock-in). Today we have Vulkan and OpenXR projects moving along nicely. And Mesa reached major milestone only this year. So what Valve were talking about starts falling in place really only now. So indeed, we'll see how things will go from here, but it all looks quite good. That effort investment is paying off, since developers are getting good tools now and users can have good experience.
If you are talking about UWP, that's actually something I'm not looking forward to, because it's MS's ploy to increase their lock-in. As far as I know, UWP has no Vulkan support, which MS will try to use to force developers to use DX even on Windows.
Tim Sweeney from Epic was very critical of UWP because of it.
Last edited:
I think AMD has potential, and has recently gotten back to being almost competitive, but they still aren't really there. Especially not if you throw power consumption into the equation. Oh yeah, about that... a lot of AMD's value advantage goes away once you factor in the cost of added cooling.
As for gaming on Linux, I'm pretty neutral there. And I suspect that the AMD-vs-nVidia deabte would change dramatically if nVidia supported Linux as well as AMD does. But the drivers arent' there, nor is the demand for a native version,and WINE has some severe limitations, so it's all moot.
As for gaming on Linux, I'm pretty neutral there. And I suspect that the AMD-vs-nVidia deabte would change dramatically if nVidia supported Linux as well as AMD does. But the drivers arent' there, nor is the demand for a native version,and WINE has some severe limitations, so it's all moot.
Nvidia supports Linux in a way "well enough". Far from perfect from integration perspective, but their drivers have same performance as their Windows drivers, in fact they basically share the same code. Their problem is not lack of support, but lack of cooperation with Linux kernel developers. While AMD (and Intel) have open source kernel drivers and graphics stack, Nvidia in contrast keeps them closed, which on Linux causes some irritations for users. So while they are usable, I wouldn't call them best user experience. AMD is surely more pleasant to use on Linux these days.
It wasn't always the case however. Until about a year ago, AMD drivers (Mesa) weren't competitive enough, and Nvidia was the only well performing option. I switched to AMD GPUs from Nvidia only around a year ago, when Mesa finally caught up to Nvidia blob in performance. That's the reason majority of Linux gamers are still using Nvidia today. It was just a lot better until recently. Not anymore though, today AMD is the obvious choice for Linux gamers.
Last edited:
I think that even if the support were there Linux would simply fail to be popular just by the fact that it's more complicated to use than Windows. I mean, PC gaming isn't exactly simple compared to consoles in the first place so people would naturally choose the OS that's easier to use. And yeah sure, Linux has way more freedom and options etc. but that only appeals to power users, not the average user who just wants to game.
I expect most PC gamers to be power users, so they are actually good candidates for switching to Linux. But today the remaining problem is not in usability. Linux distros are quite usable and I wouldn't consider Windows easier to use.
The problem is the classic market capture by MS. They have near monopolistic grip on PC manufacturers. Sure, you can get PCs with Linux if you go out of the way to look for them (or you can assemble it yourself and install whatever you want obviously), but common users don't see them in regular stores when they buy computers for their everyday use. Most simply don't care what OS they get, and if Linux would have been preinstaleld there, they would have used it just fine. That's why Windows remains the majority. Not because it's better, but because no one broke the monopoly with Linux yet.
So natural growth of Linux user base is so far the only driver of the Linux gaming growth, and it's not fast. However there is some feedback loop that helps here. The more developers release Linux games, the more likely gamers will be using Linux. And Wine is a major factor here too. I see such comments all the time like "I've just ditched Windows, becasue my favorite Windows game XYZ now works perfectly in Wine".
Have you considered that people aren't willing to switch to Linux since Windows 10 works so well? I've had no issues with it and I've been using it since it was released. By far the most stable and trouble free Windows I've ever used. Why would I switch? I see no reason to go through the hassle.
The only reason I could see people looking into other alternatives would be if MS messed up their next OS, and released another Vista. But nothing else really.
The only reason I could see people looking into other alternatives would be if MS messed up their next OS, and released another Vista. But nothing else really.
I've actually seen many people who switched to Linux precisely because they couldn't stand Windows 10 for various reasons. From privacy abuse to uncomfortable interface or whatever else they didn't like. They were OK with Windows 7, but Windows 10 got on their nerves. MS increased Linux adoption with Windows 10 release.
Last edited:
That lack of cooperation and the fact that the official drivers are proprietary code is what I meant by "lack of support". I mean, it is less support than open-source from a cooperative hardware vendor, isn't it?
The problem I see is a cycle; fewer people use Linux, so fewer peopel want to use Linux because fewer people.... you get the idea. Seriously, there are MANY people out there who believe that popularity=quality, and Linux doesn't have the advertising/marketing budget that Apple and Microsoft do so they can't really get PR on their side.
I've seen a lot of folks who lack computer skills (for instance, my wife) use Linux, so I don't buy that argument. There are some distros that are so simple to instal, configure, and use that if you can't do it then you aren't able to install games under Windows either.
The problem I see is a cycle; fewer people use Linux, so fewer peopel want to use Linux because fewer people.... you get the idea. Seriously, there are MANY people out there who believe that popularity=quality, and Linux doesn't have the advertising/marketing budget that Apple and Microsoft do so they can't really get PR on their side.
Right, catch 22 is part of the issue. But it's being eroded with more developers releasing for Linux these days, so the progress is positive here. And if CDPR could come back and release for Linux again, it could give Linux gaming market a good boost too.
Last edited:
I doubt Linux has any chance of becoming mainstream at this point anymore. Maybe when there's a big shift from traditional gaming to VR there might be some, but that will probably depend much on what Valve and Gaben bring to the table. But until then it'll remain in the obscurity it's always been. And most likely it'll stay there even after that.