Prepare a glass for some more Wine as DXVK 0.62 is out with possible performance improvements

I had two games which still needed Windows, Dark Souls 3 and DCS:World. DS3 works now with DXVK, so one less reason to reboot. But DCS:World is whole different beast...

Is Wine and its derivated (DVK...) the future of Linux gaming?
Game developers could find here an easy/fastest, perhaps even cheaper way to port their games to Linux gamersplateform.
Are games devs aware of this possibility? If so why they don't work with Wine and DVK developers?


Last edited by legluondunet on 14 Jul 2018 at 10:38 am UTC

Is WIne and its derivated (DVK...) the future of Linux gaming?
Game developers could find here an easy way to port their games for Linux gamers.

In the end, the only thing that should truly matter is the actual end-result. Does it work, does it work well and is it supported? If a game developer packages it with some form of Wine and you can tick those three...then it shouldn't really matter much.

Is WIne and its derivated (DVK...) the future of Linux gaming?
Game developers could find here an easy way to port their games for Linux gamers.

In the end, the only thing that should truly matter is the actual end-result. Does it work, does it work well and is it supported? If a game developer packages it with some form of Wine and you can tick those three...then it shouldn't really matter much.

We're all aware of the games that has a Windows version working better in Wine than the native version. While it's kinda shame, if you think about this for a minute it only means that sometimes Wine is worth a shot. Native versions and purity is good to have but not always an option, really. Wine is mostly considered a necessary evil because it's often cumbersome to set up and mess with the library overrides and tweaks. If the developer/publisher/porter do that for you so that your experience is the same "click the play button", does it still worry you? I think I can happily live with that. And I guess Gaben is moving in exactly that direction.

Might seem funny but every time a new wine version is posted here in GamingOnLinux I actually get a gulp or two of red wine. Now I've just emptied my bottle.

"So here's my story how a Windows API implementation made me alcoholic and ruined my life"

I just look forward to a day where I don't dual boot to play with friends. It's the only thing I dual boot for. Plus as a bonus aged software that is no longer supported should live on for us using wine well after windows users stop being able to run it.

Well, stranger things have happened. DOSBox is around for a reason, perhaps one day Wine will be such a tool. It is an interesting thought, but we're likely thinking quite a long time in the future with that.

If I'm not mistaken, you need the nvidia drivers 396.24.10 (396.24.0 will not work) to get those actual speed improvement, and If I'm not mistaken those are beta drivers.

So I'll be patiently waiting till nvidia releases the newer stable driver.

Prepare a glass for some more Wine..

Wine and all "related" projects are moving so fast lately.. I think I am becoming an alcoholic..

Edit: oh.. I just realised from the comments above that I am not alone! Well.. thats something


Last edited by SadL on 14 Jul 2018 at 12:04 pm UTC

Is Wine and its derivated (DVK...) the future of Linux gaming?
Game developers could find here an easy/fastest, perhaps even cheaper way to port their games to Linux gamersplateform.
Are games devs aware of this possibility? If so why they don't work with Wine and DVK developers?

I think that the actual technical implementation of how a game runs on Linux is not that important as long as the purchase counts as a Linux purchase.
I don't mind at all if a game is native, Feral/Eon wrapped, wine/dxvk packaged or whatever as long as it runs well. What I do want/demand though is that developers/publishers know that I spend my money for Linux and not for some other platform. Only a (recognized) better market share will take gaming on Linux to the next level. How it is done is not important imho.


Last edited by jens on 14 Jul 2018 at 12:46 pm UTC

This DXVK shows version in hud


New parameter needed, in my case use this:

DXVK_HUD=version,devinfo,fps DXVK_FAKE_DX10_SUPPORT=1 WINEPREFIX='/home/linuxdesktopx86/.local/share/wineprefixes/Default-x64' WINEDEBUG=-all,+fps /opt/wine-staging/bin/wine64 steam

^_^

Is DXVK meanwhile safe to use? I'd really like to give it a shot, but I'm still afraid of getting (VAC) banned.

To those who are playing games with DXVK, which games do you play, how often and how long is your session and did you experience any problems so far?

I just look forward to a day where I don't dual boot to play with friends. It's the only thing I dual boot for. Plus as a bonus aged software that is no longer supported should live on for us using wine well after windows users stop being able to run it.

Well, stranger things have happened. DOSBox is around for a reason, perhaps one day Wine will be such a tool. It is an interesting thought, but we're likely thinking quite a long time in the future with that.

As far as I know it already happens. I remember hearing about people having to fiddle and struggle with games that I can play in Wine just fine.

Crash Bandicoot doesn't work for me at all on my nvidia card, it just freezes completely right at the start (don't even see the activision logo)

Anyone know a way around this?

There's a bug report about this, apparently the game requires some tinkering with command line options and disabling NVAPI (which you should do anyway if you use dxvk, since it won't work). That's not something I can fix though.


Last edited by YoRHa-2B on 14 Jul 2018 at 5:35 pm UTC

I've been using DXVK for a majority of my games since completely reformatting my Windows drive and switching to Linux and I seriously have no issues playing the same games I used to play on Windows so I would say its a huge success! Sure waiting for the shader cache to fill can be annoying but once it does fill its very much worth it!

Is WIne and its derivated (DVK...) the future of Linux gaming?
Game developers could find here an easy way to port their games for Linux gamers.

In the end, the only thing that should truly matter is the actual end-result. Does it work, does it work well and is it supported? If a game developer packages it with some form of Wine and you can tick those three...then it shouldn't really matter much.

We're all aware of the games that has a Windows version working better in Wine than the native version. While it's kinda shame, if you think about this for a minute it only means that sometimes Wine is worth a shot. Native versions and purity is good to have but not always an option, really. Wine is mostly considered a necessary evil because it's often cumbersome to set up and mess with the library overrides and tweaks. If the developer/publisher/porter do that for you so that your experience is the same "click the play button", does it still worry you? I think I can happily live with that. And I guess Gaben is moving in exactly that direction.

I think, though, that basically it's impossible for a thing like Wine to result in equivalent performance. This is not a problem for games where performance isn't an issue either because they're "smaller" games not demanding too much of the hardware, or older games that demanded a lot of old hardware. But when it comes to the "latest and greatest", Wine-based releases would mean Windows games would be consistently faster than Linux games, and the basic gamer position on Linux would then simply be "it's slow".
I still like Wine, though. But I don't like the idea of Wine as the default "release" method for games on Linux, no matter how good Wine gets. It would relegate Linux to a ghetto.
I guess the main reason this bothers me is that to a fair extent I see games on Linux as more of a means than an end--I want to see the Linux desktop grow to the point of being a recognized serious platform with a big share of desktop use. I want this because I support Open Source software, because I dislike monopolies, because I specifically dislike Microsoft, because more Linux desktop users means more Linux desktop development and bugfixing, and because many of the remaining problems of the Linux desktop are related to it being too small for outfits doing desktop-related things (software, hardware, standards) to have to take it into account. A Wine-based gaming solution would in theory solve the problem of game availability for existing Linux users, but with inferior performance would be less good in terms of the problem of gamers being willing to switch to Linux, so it wouldn't help grow the Linux desktop.

Mind you, there is a related issue where Wine could be a big help to Linux desktop adoption: The situation where Wine will easily play older Windows software, and newer Windows versions won't, either at all or not easily. This seems like it's going to become more and more common.

I just look forward to a day where I don't dual boot to play with friends. It's the only thing I dual boot for. Plus as a bonus aged software that is no longer supported should live on for us using wine well after windows users stop being able to run it.

Well, stranger things have happened. DOSBox is around for a reason, perhaps one day Wine will be such a tool. It is an interesting thought, but we're likely thinking quite a long time in the future with that.

As far as I know it already happens. I remember hearing about people having to fiddle and struggle with games that I can play in Wine just fine.

Happened to me once, three or four years ago. I was happily playing original Starcraft on Wine, and my daughter's boyfriend thought it looked cool so I lent him the CD. He took it home, couldn't get it to work. I think probably he could have, there was like some kind of compatibility mode or something, but still.

I guess the main reason this bothers me is that to a fair extent I see games on Linux as more of a means than an end--I want to see the Linux desktop grow to the point of being a recognized serious platform with a big share of desktop use.

The same applies to gaming on Linux itself too. To attract the mass of people, Linux needs to be indeed on pair with Windows regarding gaming experience. That will only happen with native titles that truly unleash the power of Linux. But these kind of titles won't come out of the blue. They will eventually start coming once Linux has reached a serious market share. So it's the chicken and egg problem. A cheap solution for developers based on wine/dxvk with a small performance penalty could be the in-between solution that slowly increases number of available titles and alongside attracts new users to increase market share. That with the assumption that games bought for playing on Linux, no matter how, are recognized as Linux purchases.

Once we have the gaming experience and serious number of users it is indeed time to take over the (desktop) world ;)

Is WIne and its derivated (DVK...) the future of Linux gaming?
Game developers could find here an easy way to port their games for Linux gamers.

In the end, the only thing that should truly matter is the actual end-result. Does it work, does it work well and is it supported? If a game developer packages it with some form of Wine and you can tick those three...then it shouldn't really matter much.

We're all aware of the games that has a Windows version working better in Wine than the native version. While it's kinda shame, if you think about this for a minute it only means that sometimes Wine is worth a shot. Native versions and purity is good to have but not always an option, really. Wine is mostly considered a necessary evil because it's often cumbersome to set up and mess with the library overrides and tweaks. If the developer/publisher/porter do that for you so that your experience is the same "click the play button", does it still worry you? I think I can happily live with that. And I guess Gaben is moving in exactly that direction.

I think, though, that basically it's impossible for a thing like Wine to result in equivalent performance. This is not a problem for games where performance isn't an issue either because they're "smaller" games not demanding too much of the hardware, or older games that demanded a lot of old hardware. But when it comes to the "latest and greatest", Wine-based releases would mean Windows games would be consistently faster than Linux games, and the basic gamer position on Linux would then simply be "it's slow".
I still like Wine, though. But I don't like the idea of Wine as the default "release" method for games on Linux, no matter how good Wine gets. It would relegate Linux to a ghetto.
I guess the main reason this bothers me is that to a fair extent I see games on Linux as more of a means than an end--I want to see the Linux desktop grow to the point of being a recognized serious platform with a big share of desktop use. I want this because I support Open Source software, because I dislike monopolies, because I specifically dislike Microsoft, because more Linux desktop users means more Linux desktop development and bugfixing, and because many of the remaining problems of the Linux desktop are related to it being too small for outfits doing desktop-related things (software, hardware, standards) to have to take it into account. A Wine-based gaming solution would in theory solve the problem of game availability for existing Linux users, but with inferior performance would be less good in terms of the problem of gamers being willing to switch to Linux, so it wouldn't help grow the Linux desktop.

Mind you, there is a related issue where Wine could be a big help to Linux desktop adoption: The situation where Wine will easily play older Windows software, and newer Windows versions won't, either at all or not easily. This seems like it's going to become more and more common.

Hmm, I do dual boot and have noticed equivalent performance running some windows games using wine

Take DOOM 2016 for example, I have monitored gpu and cpu usage for it using hwmonitor on windows 10, and then the same on linux mint using psensor and cpu and gpu usage seems to check out about the same.. and an extra bonus, doom 2016 actually loads faster on linux for me

And . . . framerate?

For NVIDIA Vulkan+Wine games run pretty close to windows performance (some people claim better performance) but with AMD and RADV there is a perf hit due to the windows vulkan driver being better. You can hack in the AMDGPU-PRO vlk driver for better performance.

I think, though, that basically it's impossible for a thing like Wine to result in equivalent performance. This is not a problem for games where performance isn't an issue either because they're "smaller" games not demanding too much of the hardware, or older games that demanded a lot of old hardware. But when it comes to the "latest and greatest", Wine-based releases would mean Windows games would be consistently faster than Linux games, and the basic gamer position on Linux would then simply be "it's slow".
I still like Wine, though. But I don't like the idea of Wine as the default "release" method for games on Linux, no matter how good Wine gets. It would relegate Linux to a ghetto.

In the current situation of Linux gaming market, ain't we a ghetto? Also, I think that doing the right optimizations, many games will get the almost the same performance in both platforms. But even if we there is a performance hit, that shouldn't bother us if it lets people play the games they want in our platform with publishers support and one click installers.

I guess the main reason this bothers me is that to a fair extent I see games on Linux as more of a means than an end--I want to see the Linux desktop grow to the point of being a recognized serious platform with a big share of desktop use. I want this because I support Open Source software, because I dislike monopolies, because I specifically dislike Microsoft, because more Linux desktop users means more Linux desktop development and bugfixing, and because many of the remaining problems of the Linux desktop are related to it being too small for outfits doing desktop-related things (software, hardware, standards) to have to take it into account. A Wine-based gaming solution would in theory solve the problem of game availability for existing Linux users, but with inferior performance would be less good in terms of the problem of gamers being willing to switch to Linux, so it wouldn't help grow the Linux desktop.

Something is better than nothing. With nothing no one will make the switch, as they don't get "the software they need" in our platform (that's ones of the main complains I get when I offer a Linux), with something at least most of the people that dual boot will stop doing so, and that's a big win from my point of view.