Hacker News Reader

Comment by rounce | Hacker News Reader

Perhaps it’s just evidence that anecdata isn’t some universal truth.

I’ve had the complete opposite experience for the vast majority of games, where in most cases performance for me has been better on Linux than it was on Windows (can’t compare like for like now as I no longer have a Windows install outside of a VM). Friends of mine experience weird mid-session crashes and hangs on Windows that I’ve never had on Linux. I’m running an Nvidia GPU which is supposedly some kind of Linux boogeyman, but have had only one issue with EDID of a specific monitor and that’s it. Just my experience YMMV.

View on HN

> Perhaps it’s just evidence that anecdata isn’t some universal truth.

This isn't though. I have hard numbers. I've actually measured the performance. You get 5-20 FPS less and often more input latency and stutters (1%, 0.1% lows). If the machines doesn't well with Linux, it can be much worse.

Basically on HN whenever you express an opinion based on a significant amount of experience. You get someone basically saying "this is anecdote". There is a difference between "an anecdote" and "I've actually have a huge amount experience with this stuff.

Have you produced an exhaustive survey across a wide range of hardware and driver and display manager combinations? I’m happy to be an outlier here but my own experience doesn’t match with what you described hence my reply.

If I admit to anything less than doing a gamer nexus style benchmarking suite you will just claim it is an anecdote.

I have actually tested on a number of different distros and display managers and at least two different video card chipset manufacturers. No it isn't exhaustive, but it decent enough sample size to determine that the claim that Linux performs better than Windows isn't true. Even if it is the case,the results are so variable you are better just using Windows because things are more consistent.

I am saying this BTW as someone that first started using Linux in the early 2000s. I think gaming now is really good on Linux. Is it better than Windows? Well I don't have to run Windows now to play games and that is good enough for me.

Here are good videos with benchmarks: https://youtu.be/u4a2pDMXLAE https://youtu.be/77LBtP3nZwY

And I get totally different results. It not just the distro. It the version of the Kernel, the version of proton, whether you are using X or Wayland etc. Etc. Etc.

The very point I am making is that it is so variable. So posting benchmarks pretending that it proves anything is asinine.

I won't even get into all the other issues with the mouse getting lost on some games, text being too small/to large. Having to fuck around with LD_PRELOAD flags and loads other gumpth that is never mentioned on a YouTube video.

Your claim was that you tested and "You get 5-20 FPS less and often more input latency and stutters (1%, 0.1% lows)", "it decent enough sample size to determine that the claim that Linux performs better than Windows isn't true". That you tested so well, that it can't be considered to be anecdata. You claimed that it is universal truth.

So I provided you with solid data from testers, who found many cases in which Linux was on par or faster than Windows.

And now you are trying to change the goal.

Reply on news.ycombinator.com