Quoting: Eike

Quoting: ageresDoes anyone remember TF2 Tux hat?

Wasn't it the only incentive ever of Valve for Windowsers to try Linux? :)

If my memory serves me correctly, it was not a hat, but something attached to your belt?

View this comment - View article - View full comments

Quoting: meggerman

Quoting: liamdaweSo, take Phoronix results with a huge bag of salt.

Don't most people :P

Well, when people use it to think there's some huge regression in my benchmarks, I need to make it clear.

So not only is he using a more powerful processor, it looks like a ton of settings weren't actually being set.

View this comment - View article - View full comments

Also note, Phoronix benchmarks were wrong: https://github.com/phoronix-test-suite/test-profiles/pull/7 [External Link]

Feral sent a pull request to fix them. Essentially, a bunch of options were totally wrong.

So, take Phoronix results with a huge bag of salt.

View this comment - View article - View full comments

Quoting: fatriffFeral recommends using the NVIDIA 384 driver series for now when playing this game as there are some current regressions with the 387 series up through 387.22.

Phoronix is getting 133FPS on Ultra VS your 80FPS on the same card..

As already stated in the article and in the comments, I tested both driver versions and the difference was negligible.

I'm currently going through more tests to see if I can find any reason for me to possibly have lower than expected performance.

Keep in mind with Phoronix, his CPU is overclocked at 1.7GHZ faster than mine.

The linked Youtube be benchmark is also using a 1GHZ faster processor.

The base clock on mine is only 3GHZ, which could account for a lot of it.

View this comment - View article - View full comments

Quoting: jens

Quoting: Alm888

Quoting: ShmerlAspyr never properly explained, why they don't want to release it on GOG.

It seems Feral and Aspyr will never release anything outside of Steam just because.
This, coupled with the fact that their "ports" are of outrageous quality nearly makes these two companies my personal sworn enemies. Their actions are more harmful than beneficial to Linux ecosystem (IMO).

Thank you for stating that this is just your opinion.
May I ask how many Feral/Aspyr ports you have actually played to make such a bold statement?

It's also false, Aspyr also published Layers of Fear by the same developer, which is on GOG for Linux.

It's not always up to Aspyr and Feral where the games go, they will often be at the mercy of the original developer.

View this comment - View article - View full comments

Be aware, Feral also told me about the driver issues, but I was told the driver issues with the 387 were crashes, not performance related.

I also ran the benchmarks on the 384 series, the difference was almost nothing. We're talking 2-4FPS between the driver versions, sometimes better and sometimes worse.

You need to take into account what Distro other benchmarks are using, what exact settings they are using including the track, weather and so on. There's a lot of factors that can change benchmark scores.

The linked benchmark video from airspeedmph uses a completely different track, different camera and completely different system specs. However, do note that in their benchmark, they also show the Windows difference grow larger as the settings are lowered, much like mine does, so I imagine it's due to our CPU more than anything.

Also, if the game is not restarted each time the settings are changed, some of the settings don't completely take effect.

Edit: For those who don't see my more recent comment. Phoronix benchmarks were wrong:

Also note, Phoronix benchmarks were wrong: https://github.com/phoronix-test-suite/test-profiles/pull/7 [External Link]

Feral sent a pull request to fix them. Essentially, a bunch of options were totally wrong.

So, take Phoronix results with a huge bag of salt.

View this comment - View article - View full comments

Quoting: dubigrasu

Quoting: GuestThe benchmark values are really looking bad.
But if you take a look at the video comparision here:

https://www.youtube.com/watch?v=ojKC5kXopCs [External Link]

it looks much better for Linux, roughly 10% difference.
Note sure why the results differ that much.

Different hardware/system, and likely different track used for benchmark. My guess is Liam used the default Melbourne track, while in video is the Silverstone one.

Yep, different tracks, different hardware.

Lower-end hardware might show a very different story to my benchmarks.

View this comment - View article - View full comments

Quoting: wintermuteFrom the requirements [External Link]:

The Nvidia 387 driver series has known issues. We advise that you avoid using this driver for a better user experience.

Is this a factor in the performance Liam is getting?

No. The issues are related to crashes that I have not encountered.

View this comment - View article - View full comments

Quoting: AvelanHi Liam,

could you tell us which graphics-api was used in your windows benchmark? Was it also Vulkan? Or DirectX 11/12? Thank you for the review.

Vulkan is only in the Linux version.

View this comment - View article - View full comments

Quoting: EgonautThose games were added already a week ago when the Halloween sale started. The Halloween sale will end in ~10 hours, those games will probably disappear at the same time from GOG COnnect.

Correct, but we have not covered it yet - now we have.

View this comment - View article - View full comments