I don’t know how Windows games are counted when they are executed in a Linux environment but I do know that Windows games executed on macOS through GPTK are counted as Windows environment, which is dumb.
Anyway, good to see Linux holding strong. I’ve been waiting for this since Hardy Heron
Don’t they count what the Steam client runs on?
The Steam client running through GPTK is the Windows client
The steam survey doesn’t count individual games; the Steam client itself collects and sends the data after prompting the user for permission. So all the matters is the OS visible to the Steam Client.
Which is, in that case, Windows 10 in a Wine bottle.
Steam.app can’t run Windows games directly on macOS, like it does on Linux with Proton. So you have to install the Windows version of Steam in a Wine bottle to be able to install and run Windows games from Steam on macOS
Finally got a 4tb ssd installed and installing a flavour tonight, time to add one more tick to this list
Awesome stuff, a consistent percentage is what we want. Sadly, I don’t think we’ll see developers flocking to Linux as they did when Macs had a similar percentage.
honestly, linux native games often run worse than windows binaries through proton and dxvk. game developers really only need to get the anti cheat working, if there is any and fix potential issues.
Anti cheat is not a technically issue, it is a busniss decision
Also kernel level anti cheat is idiotic. but if people accept it, then it is on them. why anyone would accept a fucking snake into their bed just to play call duty of duty tho
Asking for a friend.
Had that happen to me with Last Epoch. At launch the native linux version had graphical issues where as proton ran the windows version almost perfectly.
YSK that Steam counts Macs running Windows games through Game Porting ToolKit as Windows machines.
Game developers will do the bare minimum. IMO they won’t bother with a native Linux build if their games run good enough through Proton, at least for the time being.
Linux needs to get more market share, like way more, to move things unfortunately. I know the feeling very well. It was the case on Mac when Bootcamp was still a thing, even though macOS had great OpenGL support.
I think we‘re in a very different situation right now. Proton has become so good that it‘s just not necessary for most developers to do anything to get their game running on Linux. When Macs peaked in the hardware survey, the compatibility tools were far less powerful and developers had to actually invest time and resources, if they wanted their game running on Mac.
I also think that the Steam Deck is absolutely being recognized by many developers. Even big publishers proudly announce their games being playable on it. And having games optimized for Deck often improves them on Linux in general.
So I really wouldn’t worry about developers not specifically targeting Linux. Even without that, gaming on Linux is in the best spot it has ever been and is steadily improving.
When a new game is released I usually check if it’s steam deck compatible, if it isn’t for no specific reason (like, a 2d platformer, I’m not going to expect a high fidelity 3d game to work) I’m way less inclined to buy it. The market is there and really should be picked up.
Even with steam deck verified I was skeptical but I finally made the jump to linux on my gaming pc and installed starfield and it booted right up, didn’t notice any difference it’s amazing. I imagined I was going to have to go into steam settings and do stuff and keep retrying but nope just worked right away
Does anyone know if the Steam Hardware Survey identifies your OS even in the Flatpak version? Or will it detect the Freedesktop SDK?
50% still on Windows 10 is crazy. What is Microsoft going to tell them next year when the support runs out?
What I like more is that it looks like only half the win10 users are switching to win11
Edit: -1% win10, +0.5% win11
Most Linux users on Steam use Arch btw.
Most as in SteamOS + Arch = 49.25%.
It’s interesting how fragmented the Linux user base is in the survey. Excluding steam deck from the equation, the visible versions of Ubuntu are getting roughly 18.6%, Arch is getting like 14% of the desktop and Mint 21.3 getting 8.5%. The Flatpak version does put confusion into the data (hiding 11% of desktop versions) and the missing “other” 22.94% group accounts for 39% of the desktops so there may be lots of other version fragments hidden away, but regardless no single distro version seem to dominate.
It’d be nice to see the whole list.
SteamOS 3 is Arch BTW.
most desktop users
Mostly because Arch is Arch while Ubuntu and such are divided into the different release versions.
They split manjaro out
In the end they all are Arch, Debian, RedHat
But don’t forget our friends at Gentoo and SUSE.
that’s true, they’re also most welcome
If you count different versions of Ubuntu separately, then yes.
@robocall @mr_MADAFAKA Steamdeck is arch based, isn’t?
its a different os in the survey, real arch linux is really this popular.
Why isn’t it listed here?
SteamOS Holo is the Steam Deck
@mr_MADAFAKA It was more just works than tell papa microsoft EVERYTHING? But probably just random luck off the sampling.
can’t wait for steam to release an ARM linux build.
Okay great. Call me when it’s 10%.
It’s sounds snarky but the reality is not much will change from software and hardware developers until it reaches that level. Right now the direct support we get is from developers that just happen to like Linux. After around 10% most other developers can no longer afford to ignore that market even if they aren’t adept or comfortable with it.
Why the fuck is windows 7 increasing
The hardware survey doesn’t ask every single user, it just gets a sample. So it probably just happened to hit a few more Windows 7 people this month.
I guess I’m really just questioning why there are any at all… hope they enjoy being part of a botnet…
From my experience there’s this weird subset of people who don’t like newer Windows versions, which is fair enough, but instead of learning to modify those or learning Linux, they believe they can turn back time, which isn’t something you can just do when connecting it to the ever forward-marching internet.
Does valve publish a margin or error for the percentages? Steam is such a huge sample I feel like the effect would be miniscule
