From Hardlimit
I don’t think the issue is performance though. The unspoken part of this comparison is in bold:
“Linux vs Windows tested in 10 games. In the games we could find that work on linux, the performance was 17% faster on average. In all the rest of the games, Windows worked 100% better.”
I’d argue that the idea that most games don’t work on Linux is a flat-out misconception in 2023.
It’s hard to quantify, but Valve’s own Steam Deck (=running on Linux) verification stats have 70% of games either Verified or Playable (Playable generally means that it runs but text is small on the Deck screen, or it needs a lot of keyboard input – nothing that matters on the desktop). Crucially, “Unsupported” doesn’t mean it doesn’t run – it means untested, and in my experience at least, many of those just work too.
Protondb shows 80% of its catalog with a Platinum, Gold, or Silver rating – 70% are Gold. Silver generally corresponds to e.g. switching to Proton Experimental, which is a single-click process.
Anecdotally, after being gaming only on Linux for more than a year, with a catalog of 500+ games, I’ve had one (1) that gave me any more trouble than that Proton Experimental switch (Assetto Corsa, first one).
So there is no “unspoken part” here. The experience running Windows games on Linux isn’t what it was even 2 years ago. It is, for many people, an entirely seamless experience now.
PS: seeing Windows games running better on Linux isn’t a new observation either. Elden Ring was a great example where Proton shader precaching eliminated the stutter that plagued that game at launch, so it didn’t happen on Linux.
“gave me any more trouble than that Proton Experimental switch (Assetto Corsa, first one).”
Oh great, the only game I have any interest in playing…
I found these comparisons not useful. Nobody play on Linux for searching for performance, but to avoid switch os only for playing.
I do. Windows 11 is just a bunch of bloatware and ads stuffed in a trenchcoat. I want to be able to use all those rams and GBs I downloaded, without half them being tied up in tracking.
It’s rather important to understand the performance characteristics for people to know what to expect if they want to switch to Linux.
If games ran at half the FPS on Linux as they would have on Windows, then pretty much no one would be gaming on linux.
If you got 90% performance on Linux, only Linux enthusiasts would take the performance hit.
At 100% performance the choice is completely free, people that got fed up with windows could just switch.
When Linux outperforms Windows, this puts us in very interesting territory, as this might even entice a bunch of people to give Linux a try to see whether the switch is worth the performance. I’m personally quite interested in seeing whether this could be the tipping point for Linux on desktop and laptop to really start taking off.
