Valve Continued Doing A Lot For Linux Gaming & Open-Source Radeon Drivers In 2020

[oiaohm]

An important distinction to make here is that Valve is not a publisher, outside of their own games. They are (primarily) a distributor. The 20-30% Valve takes is in addition to the share the games publisher takes. I've read stories that go as far as Valve taking 30%, publishers taking 50%, leaving only 20% of the game's revenue for the actual developer. This isn't so much a problem for companies like Rockstar, who are both developing and publishing their own games, but might be(come) a problem for devs like Gearbox Software ( famous for their Borderlands series ) who depend on publishers like Take-Two / 2K for their releases. It's at least a 3-way split ( sometimes more if multiple developers / publishers are involved ) and the developers get the short end of the stick every time.

Now, I'm not saying it's not fair that Valve takes a 30% cut, I am not a bookkeeper. I'm just saying that developers getting the smallest piece of the pie when they are the ones that actually make the product is definitively not fair IMMO.
The solution ? I wouldn't know where to start. People being d***s is a problem that has been around since the dawn of mankind…

https://arstechnica.com/gaming/2019/04/why-valve-actually-gets-less-than-30-percent-of-steam-game-sales/

When bookkeepers go looking the Valve cut is a max of 30% the key word is "Max" that is missed a lot. The average is around 20% to valve but that 20% is a average not the min Valve does take. Its really simple to miss that some of the special prices on steam is valve cutting into their own share not into the developer/publishers share and the keys Valve let game publishers sell themselves yes Valve let the publishers/developers sell those keys without paying a cent to Valve yet they register into the steam services system.

When you buy a game third party that is hosted in steam its anywhere between 30% to valve if you bought it on steam at full price to 0% to valve if it was a key your got from publisher/developer by some other means.

So its 0-30% not the 20-30%. The correct way it write it is: Valve get between 0-30% with average being 20% from game sales of third party paid for games on steam. (yes is a horrible line to write it all correct)

Of course there is some of games that is 100% to valve that is stuff valve developed or vacuumed up from deceased companies.

https://store.steampowered.com/app/599390/Battle_for_Wesnoth/

Yes you also have the true open source games given out by steam of course valve does not get a cut out of any donations to those games. Yes was why before it was important to say paid for games because if its all games on steam it is really a lot lower than 20% when you start allowing for other costs. The fun of running retail you have products there that you are making loss on that you are crossing fingers person will come in door and by something else.

[bridgman]

I want to know where the Linux has a 1% share of the market statistic comes from; near as I can tell that is just something you say when you think a platform is irrelevant. It has been my experience that when anyone says a percentage without a citation it's made up. For instance on my Discord where I have 1000+ members for my Minecraft mod packs. I could say that about 25% run on Linux. But where would I've gotten that number? Am I saying that at least 250 of them are running Linux? Clearly it's conjecture. Just going on conversations alone and whom I've suggested going to Linux when they complain about the insane updates from Windows based on memory alone that is at about 25%, even then it's still a made up number. Actively I would say out of 1000+ members there are about 15 people in total that talk to me and 25% of them probably are on Linux based on the conversation. This is like a certain viral statistic. So if the real sample is 25% of 15, and you base that against the total population rounded to 1000, the actual percentage might be (still conjecture you see):
15 * 25% = 3.75 / 1000 = 0.00375 * 100 = 0.375%

Which is completely made up. Looks good but it isn't.

What happened to articles "and the like" that had citations?

AFAIK the primary sources of OS usage info have been the same for a while, to the point where citing them on every post gets tedious. One of the challenges with measuring Linux usage is that it depends a lot on which particular domain and source you are looking at, eg:

gaming - Steam stats - just under 1%
gaming - independent publisher stats - 2-3%

general - Statcounter - just under 2%

developer - w3schools - just under 5%
developer - stack overflow survey - 25%

embedded - haven't seen any good stats recently; my impression is that Android has been eating into desktop Linux share but unconfirmed

server - generally accepted as >50% but no specific sources that I am aware of

workstation - mostly anecdotal at the moment via app developers; used to be 30-50% but lower these days

The "developer" domain varies a lot depending on which OS will be the eventual target of the developer's work, eg people developing for Windows desktop/laptop target generally use Windows, while people developing for server deployment are much more likely to use Linux.

Each of the above sources has its own challenges, eg Steam defaults web purchases to Windows if not used on Linux in first two weeks; web browsers sometimes fake the OS string to Windows for better compatibility etc...

https://store.steampowered.com/hwsur...tform=combined
https://gs.statcounter.com/os-market...ktop/worldwide

https://www.w3schools.com/browsers/browsers_os.asp

[Kreezxil]

AFAIK the primary sources of OS usage info have been the same for a while, to the point where citing them on every post gets tedious. One of the challenges with measuring Linux usage is that it depends a lot on which particular domain and source you are looking at, eg:

gaming - Steam stats - just under 1%
gaming - independent publisher stats - 2-3%

general - Statcounter - just under 2%

developer - w3schools - just under 5%
developer - stack overflow survey - 25%

embedded - haven't seen any good stats recently; my impression is that Android has been eating into desktop Linux share but unconfirmed

server - generally accepted as >50% but no specific sources that I am aware of

workstation - mostly anecdotal at the moment via app developers; used to be 30-50% but lower these days

The "developer" domain varies a lot depending on which OS will be the eventual target of the developer's work, eg people developing for Windows desktop/laptop target generally use Windows, while people developing for server deployment are much more likely to use Linux.

Each of the above sources has its own challenges, eg Steam defaults web purchases to Windows if not used on Linux in first two weeks; web browsers sometimes fake the OS string to Windows for better compatibility etc...

https://store.steampowered.com/hwsur...tform=combined
https://gs.statcounter.com/os-market...ktop/worldwide
https://www.w3schools.com/browsers/browsers_os.asp

Thank you for these resources.

[qarium]

Each of the above sources has its own challenges,

all the numbers are only: Past or present... not the future.
i even know hardcore windows users (who do not use linux at all) say that linux is the future.
AMD should invest in the future and not in the past or present....

[oiaohm]

AFAIK the primary sources of OS usage info have been the same for a while, to the point where citing them on every post gets tedious. One of the challenges with measuring Linux usage is that it depends a lot on which particular domain and source you are looking at, eg:

gaming - Steam stats - just under 1%
gaming - independent publisher stats - 2-3%

general - Statcounter - just under 2%

developer - w3schools - just under 5%
developer - stack overflow survey - 25%

embedded - haven't seen any good stats recently; my impression is that Android has been eating into desktop Linux share but unconfirmed

server - generally accepted as >50% but no specific sources that I am aware of

workstation - mostly anecdotal at the moment via app developers; used to be 30-50% but lower these days

The "developer" domain varies a lot depending on which OS will be the eventual target of the developer's work, eg people developing for Windows desktop/laptop target generally use Windows, while people developing for server deployment are much more likely to use Linux.

Each of the above sources has its own challenges, eg Steam defaults web purchases to Windows if not used on Linux in first two weeks; web browsers sometimes fake the OS string to Windows for better compatibility etc...

Great list.

https://chuttenblog.wordpress.com/2020/11/05/data-science-is-hard-alsa-in-firefox/

Does miss another absolute known fact. Linux users in many segments really don't like being counted and do a lot of things either directly or indirectly that makes counting them impossible. Like firefox telemetry having zero Linux Alsa users and them finding out after they removed the features there was still really tones of ALSA users at the time just using distributions that had disabled telemetry.

Yes when the Linux world is talked about being like herding cats its no understatement that its a very uncooperative place in many areas. Yes those wanting to count Linux users is absolutely uncooperative.

https://www.embedded.com/wp-content/uploads/2019/11/EETimes_Embedded_2019_Embedded_Markets_Study.pdf

This is about the best recent embedded numbers. Yes this is another area that really does not like being counted that much.

Thing is Linux desktop users are insanely hard to count. Linux distributions provide a lot of games in general that are easy to install that would see people go to steam as well under Windows. If you look at the games included with the OS selling games on Linux you are facing a lot more competition from the OS/distribution provided games than Windows.

I would not say Android has eaten into the Linux desktop. We are not seeing any major market movements that say that. Android appears as it chomped into Windows market share fairly much leaving mac os and Linux alone. Its not like those who would be using mac os or general Linux distributions can be really serviced by android offering. The interest in pinephone and Librem 5 is also because there are things the wall garden android cannot do that people want to do.

[swoorup]

Linux gaming has to be promoted much more. Above all in gaming fairs demonstrating how is better the Linux desktop proposal in gaming such as simple is the installation of a Linux operating system. The major issue actually is Linux operating systems stability environment in the huge heterogeneous hardware settings, the different way hardware gets software stack support, and the obsolescence of X11 as well. The first aim is to get stability and efficience in managing the hardware resources, the second one is to switch to wayland abandoning the graphical stack obsolescence as soon as possible. SteamOS is a great improvement because Linux has a specialized OS in gaming.

I am sorry for the tone, but I used to think like that. Now I rarely use linux on my desktops, but I game occasionally. Putting a gamer mind, I am thinking why should I give a shit about gaming in linux, when lot of things still barely work, worry about hardware support and game performance and support, when I can just worry minimally and get straight to gaming in windows at least most of the time.

Besides when the screen is fullscreen, I rarely care what's powering underneath or the OS that is running. If valve has a business interest so be it, but as a gamer the current state of art does not impress me. There has to be a strong enough incentive for game companies to provide support on linux, currently there is virtually none.

[Azrael5]

I am sorry for the tone, but I used to think like that. Now I rarely use linux on my desktops, but I game occasionally. Putting a gamer mind, I am thinking why should I give a shit about gaming in linux, when lot of things still barely work, worry about hardware support and game performance and support, when I can just worry minimally and get straight to gaming in windows at least most of the time.

Besides when the screen is fullscreen, I rarely care what's powering underneath or the OS that is running. If valve has a business interest so be it, but as a gamer the current state of art does not impress me. There has to be a strong enough incentive for game companies to provide support on linux, currently there is virtually none.

the major problem on linux is graphical. Mesa drivers are not comparable to owner drivers which have issues on X11 besides Wayland is not properly integrated yet. X11 is good for 30 years legacy hardware so it is the main bottleneck nowadays. The main aim of activity based management process is to detect the useless activities and the bottlenecks of the productivity chain in order to fix them. Problem on linux is that X11 is still almost essential, so all the software has became X11 dependent. On the other side, the major hardware and software companies have not so interest in Linux desktop environment because they don't consider the sector profitable, also if is evident that Microsoft is evolving in Linux integration as I had to tell during a course joined in 2016. Probably when Microsoft windows will be linux based, Linux will get the deserved support in desktop environment when it excels in smartphone and smart TVs and in other embedded devices where the hardware producers are the same software linux producers adapting linux to their devices. Although Microsoft will avoid that the free operating systems will get benefits of its own activity in order to contrast competitors from free OSes, also if Linux has a minimal position in market share. As you state, the major incentive is the way Linux operating system environment spread itself in the market. So the wise user uses Linux operating system and Microsfot windows as well so to be that incentive for its own interest.

[oiaohm]

Mesa drivers are not comparable to owner drivers which have issues on X11 besides Wayland is not properly integrated yet.

There are GPUs that are not Nvidia where the vendor driver is the Mesa driver. Cases where the Mesa driver is the vendor driver the stablity with X11 and Wayland is way higher.

The main aim of activity based management process is to detect the useless activities and the bottlenecks of the productivity chain in order to fix them.

Problem on linux is that X11 is still almost essential, so all the software has became X11 dependent.

Xwayland means this is a minor problem in the path to Wayland. XWayland has most of the worst X11 trouble making parts of the X11 protocol removed. Yes this results in X11 windows managers and X11 compositors not being functional. Also breaks applications attempting to run GUI as root user by default. You can set Xwayland to allow root applications to work.

On the other side, the major hardware and software companies have not so interest in Linux desktop environment because they don't consider the sector profitable, also if is evident that Microsoft is evolving in Linux integration as I had to tell during a course joined in 2016.

Even in 2016 this is a mixed bag. AMD does have a interest in the Linux desktop environments but they have had the fun of almost going under so even their windows drivers have not been great. Lot of major hardware supports the Linux desktop just its not really Nvidia.

Probably when Microsoft windows will be linux based, Linux will get the deserved support in desktop environment when it excels in smartphone and smart TVs and in other embedded devices where the hardware producers are the same software linux producers adapting linux to their devices.

Really its not that simple. Windows will not have to be Linux based to really help Linux desktop support. WSL supporting wayland and xwayland applications will make possible to make a Linux binary that works on Windows and Linux so making software developers effort possible simpler.

Its one thing to say software companies have no interest in supporting the Linux desktop but you need to look closer. There is a problem you run into with some software companies where they do support the Linux desktop as long as you are willing to run their program in wine. Yes there are some of those companies who will pay codeweavers to fix particular issues so their applications work so they do have a interest in supporting the Linux desktop. Why do this the companies want to have a single build in house because they really don't want to spend the extra resources in house and some cases don't have the extra resources in the first place(just because a company is big does not mean they are not having major money troubles) . Remember this do not want to spend extra resources is a double sided factor. WSL supporting Linux graphical applications will cause the companies using wine as their Linux support to consider using WSL as their Windows support.

Windows coming Linux based is not really required. Windows coming Linux compatible this will alter the environment in the software side this could cause a lot of interesting changes as in increasing the Linux native software at the same time decreasing the Windows native software. Fun little problem right lot of companies only have the resources to make 1 desktop GUI application so there has been choosing. The wine choice has also been effecting Mac OS as well.

[bridgman]

the major problem on linux is graphical. Mesa drivers are not comparable to owner drivers which have issues on X11 besides Wayland is not properly integrated yet.

There are GPUs that are not Nvidia where the vendor driver is the Mesa driver. Cases where the Mesa driver is the vendor driver the stablity with X11 and Wayland is way higher.

You beat me to it

There are three major PC graphics vendors, and two of the three use Mesa as their primary vendor driver. Two of the three vendors also play well with Wayland.

[Azrael5]

You beat me to it

There are three major PC graphics vendors, and two of the three use Mesa as their primary vendor driver. Two of the three vendors also play well with Wayland.

Compare them to proprietary drivers and observe the differences in terms of performance.