Linux breaks the 5% threshold for the Hardware survey first time ever on @steam@lemmy.ml
Cross posted from: https://lemmy.world/post/45783520?sort=Top
Great to see linux finally getting some notice in the gaming world. It is about time :)
@frodololo @mesamunefire I absolutely love it! It’s funny but native Windows games actually perform better on Linux because Linux is vastly more resource efficient than Windows. Windows suffers from huge code bloat. Plus Linux is more fun and satisfying. 😉
Proton is such a good tool. I play Windrose on my windows pc but it also runs well.on my steamdeck. I’m prepping a new pc for the kid and i think it will be linux out of the box, just to see how far it can go.
Tbh, another reason is that you can’t change the boot logo anymore. That was one of the earliest “hacks” and i still have fond memories of the XP Pirated Edition Logo. It also sparked my fascination with tinkering with the OS - and with Win 11 that is all gone
@jagermo Smart move! The more people we can get on Linux, the better. My finacee is not an IT pro by any means and she really likes Linux when compared to Windows. One of my missions is to get everyone I can on a Linux distro.
Well, ofc it does better, no telemetry built-in so you ain’t waiting for your PC to spy on you
So It does depend on the situation, there have been some issues with significantly reduced performance for heavily multi threaded games through wine/proton. It seems like that is getting cleared up by a direct implementation of the relevant stuff in the kernel, rather than being a distro specific thing. There are also some performance losses due to linux giving all application equal priority for “video random access memory” on a GPU (so if a game needs more than what is available it gets forced to use other memory which slows things down a lot) but there are some fixes coming down the line to make a system default to giving games higher priority for VRAM than the browser or desktop. In general stuff through a translation layer will always cause some lose in performance though.
Realistically, on a super high end machine (multi thousand dollar GPU, nonsense amounts of ram, and absurd core count over clocked CPU) the bloat of windows doesn’t use up enough resources to impact performance, but on a lot of mid and low end machines Linux will outperform. The issues with translation layers are getting solved at an astonishing rate right now and… windows is just getting more bloated and buggy as Microsoft tries to push more subscription services and data harvesting.
@megopie I am sure it is also highly dependent on the age of the game and even the level of graphic detail needed. Since I consider myself to be a retrogamer, Linux cruises!
there are some older games that rely on multi threading and even on a modern system this can cause some serious performance issues, again, this is getting solved though.
@megopie Yeah, I guess you’re right. I know that GTA III, San Andreas, and Vice City play better on Linux than Windows. I’m enjoying those very much.
2027 will be the year of the Plan 9 desktop!
My money is on GNU Hurd instead.
Heresy!
Damn just got my new linux desktop going like last week. Its my headcanon that I was the very last one in to push that (I most definitely wasn’t)
Linux breaks 5% threshold for the first time again?
Or is this a repost of the previous time?
Yeah, right, I swear I see this post every month. Same comments too. It’s a little trippy.
I think the post might be talking about last month again.
Interesting but the absolute figures are not helpful due to the way the survey is conducted. But the macro trend over a long period is clearly up; we just don’t know where is actually lies as a true %.
Valve will have far more accurate idea of actual numbers as they will be able to correct the raw survey data using other information and data they hold and collect. But they’ll never release that as it’s a commercial advantage to them to know so much about the PC games market.
@BananaTrifleViolin @mesamunefire How might they correct the fact that only a small percentage of Linux users agree to the survey? It seems like there are too many variables. I am not very familiar with data science.
What’s an example of something they could do with more data than what we see?
I guess it would be more precise, and they could sell it to game developers.
Determine if sample set is representative. And if people switch. Valve know which user change setup but not release because privacy.
