Update 27/06/19: Steam Play Proton 4.2-9 was released as well, to fix multiplayer issues with "Mordhau, SOULCALIBUR VI, and others with problems from 4.2-8".
Original Article:
Steam Play has been updated today reaching 4.2-8 along with DXVK also seeing an update to 1.2.3, let's take a look.
As a quick refresher: Steam Play is the system built into the Steam Client on Linux, that allows you to play games meant for Windows. As for DXVK, it translates D3D11 and D3D10 into Vulkan for use with Wine and it's part of Steam Play (but it can of course be used with Wine directly).
Firstly, the Steam Play update includes these changes:
- Fixes for games which embed web browsers using the Steam client. For example, Football Manager 2019.
- Fix an issue with Bloodstained: Ritual of the Night and other titles crashing with an error dialog on exit.
- Upgrade wine-mono to 4.9.0, which includes winforms support. This may help some game launchers.
- Various window management and alt-tab fixes.
- Fix for controllers losing force feedback when removed and re-added.
Full changelog found here as always.
As for DXVK, here's what's new there:
- Fixed bug that would cause some Unreal Engine 4 games to show error messages upon exit (PR #1104)
- Fixed regression which would break texture loading in World of Warships and potentially other games (#1096)
- More minor CPU overhead optimizations
- Implemented timestamp disjoint queries properly to make measuring time on the GPU more robust.
- Improved memory allocation behaviour under memory pressure. This may in some situations improve performance on lower-end Nvidia GPUs.
- Improved staging buffer allocation behaviour to limit the amount of memory that is permanently reserved for resource uploads.
I'm honestly amazed that even more performance is being squeezed out of DXVK. It's already very impressive and every optimization, even a minor one, brings us that little bit closer to perfection.
Pretty fun to see them both updated on the same day, quite unexpected but a nice surprise to be sure.
Article taken from GamingOnLinux.com.
Some you may have missed, popular articles from the last month:
The comments on this article are closed.
I'm slowly gliding over to the conclusion that Steam Play probably is the solution for gaming on Linux.
Don't get me wrong, I will always be eager to support native binaries, but let's face it: The companies/individuals who work on porting to Linux are way too few and the interest at the game developers too low.
Last edited by Beamboom on 26 June 2019 at 9:34 pm UTC
Don't get me wrong, I will always be eager to support native binaries, but let's face it: The companies/individuals who work on porting to Linux are way too few and the interest at the game developers too low.
Last edited by Beamboom on 26 June 2019 at 9:34 pm UTC
14 Likes, Who?
Quoting: BeamboomI'm slowly gliding over to the conclusion that Steam Play probably is the solution for gaming on Linux.
Don't get me wrong, I will always be eager to support native binaries, but let's face it: The companies/individuals who work on porting to Linux are way too few and the interest at the game developers too low.
I would tend to agree native games would of been preferable however steam play comes with no cost to the end user or more importantly from a developer point of view zero overhead cost we are back to relying on valve and i have to be honest i kind of hate that. I know people love to white knight them but no the track record of cancelling projects and features that they have invested heavily in the past when the demand was not as expected. I hope am just being negative and valve never turn off the money tap. However it must be working as they keep paying for development mmm I wonder if we could get hold of the steam play sales for the top ten steam games (excluding battle eye games)
0 Likes
Quoting: Whitewolfe80... I know people love to white knight them but no the track record of cancelling projects and features that they have invested heavily in the past when the demand was not as expected. I hope am just being negative and valve never turn off the money tap. However it must be working as they keep paying for development mmm I wonder if we could get hold of the steam play sales for the top ten steam games (excluding battle eye games)
I am not sure what kind of track record you are talking about, that kind of behavior is typical to Google... I am not aware of project that would eat lot of money and got canned (except rushed Steam machines, but that probably didnt cost that much).
But to be honest, I don't think they can stop supporting Linux at this point... Linux as gaming platform is the "future", because of the same reason as it is on servers and in other cases. Sure, it might not always be direct support, like Google Stadia is basically walling off the user from it. But that's where Valve will beat Stadia, because you still have to buy games on Stadia, so yet another library and one that you cant even play locally. Steam offers local play and once streaming service goes live you will be able to stream (almost?) any game from your massive library... no need to port thanks to Steam Play or already native games. If they offer some basic streaming for free like Stadia (fee taken from the 30% cut) then it's no brainer. So no, they really can't stop supporting Linux as it is their competetive edge and future-proofing.
3 Likes, Who?
Even Steam Machines isn't something I would quite consider a cancelled project. Sure they're not selling them anymore and they didn't meet the success they would have liked, but they are still updating the OS and nothing's stopping them from making a new generation of Steam Machines in the future with the experience and insight they've gained. In the meantime, as far as I'm aware, those who did buy one aren't left in the dark. They still have a working PC, an OS that receives updates (or whatever other OS they choose to install on it), and their Steam account isn't going anywhere.
5 Likes, Who?
My only gripe right now is that Vampyr for some reason have had it's performance seriously downgraded the last versions of DXVK.
First it was reported that this happened because the new DXVK required some new Vulkan extension that wasn't in Mesa 19.0 and the fallback was slower, but then Mesa 19.1 with the extension came and the performance dropped even more :( so it went from a very playable game to completely unplayable.
But then perhaps the problem lies in Mesa?! For with 19.1.0 also Wolfenstein II stopped working altogether due to some Vulkan init error and here DXVK isn't even being used so I can hardly blame it here).
This on a RX480 btw.
First it was reported that this happened because the new DXVK required some new Vulkan extension that wasn't in Mesa 19.0 and the fallback was slower, but then Mesa 19.1 with the extension came and the performance dropped even more :( so it went from a very playable game to completely unplayable.
But then perhaps the problem lies in Mesa?! For with 19.1.0 also Wolfenstein II stopped working altogether due to some Vulkan init error and here DXVK isn't even being used so I can hardly blame it here).
This on a RX480 btw.
1 Likes, Who?
gradyvuckovic Jun 27, 2019
1. PRAISE LORD GABEN AND LORD PHILIP!
2.
Absolutely. The way I see it, the battle strategy is thus:
Gamers can't game on Linux because there's no games?
Game devs won't port to Linux because there's no gamers?
Fix: Bring the games to Linux, so the gamers can come, then the game devs will follow.
"How do we know this won't kill native games?"
Here's my logic: Why doesn't Linux get more games?
Is it hard to port a game to Linux? No, not now, not with engines like UE4, Unity, Godot, or even the in house engines the likes Ubisoft and EA use, as their software is designed to be flexible so they can easily port to new game consoles and new platforms. We've seen that with Stadia.
Bringing a game to Linux doesn't mean porting it, it means supporting it, and there in lies the issue.
Support is more or less a fixed cost, despite how many customers you have. Support includes testing, checking compatibility, running tests on multiple hardware configurations, providing help staff, etc. That's not happening right now because "not enough Linux gamers to justify that cost".
Proton doesn't just offer porting, it also offers support, kinda. Valve takes the heat of support and supplies the commitment to porting (updating Proton), the game developer just gets sales.
The way I see it, the more people who use Linux for gaming, the more Linux gamers there are, the larger the potential marketplace gets. The more sales those game developers are going to get.
Sure, at first, for some game devs they will weigh up that market and say "Still not worth it to do a native port". But that revenue from Linux sales won't go unnoticed, and eventually it WILL make management of some game developers ask the question, "What are we doing to ensure we're compatible with Proton for those extra sales?".
Then we'll go through a period of time, where game devs are willing to 'half commit' to Linux, putting minimal effort into making their games run via Proton as long as that doesn't represent a large commitment, to enjoy some additional sales. That will result in many games working via Proton that currently don't, which will only increase the number of people who game on Linux.
So at what point does Proton cease to become relevant and suddenly get replaced with native games?
Hypothetically if Linux had 90% of the marketshare right now, would game developers continue letting gamers play their games via Proton? No.
Eventually, Proton ceases to become useful as a compatibility solution with minimal cost, and instead becomes a dependency burden. Just look at the number of times an update to Proton has caused a regression, if you become dependent on Proton, then you're subject to those regressions out of your control.
Supporting a game means testing, ensuring compatibility, performance, etc. Not because you want to, but because you want sales from your customers, and you want the game to run on their systems reliably, because you want those customers to tell their friends to buy the game.
Proton is an extra dependency, extra layer of complexity, an extra drag on performance, etc. When the costs of ensuring high quality Proton compatibility become larger than the cost of just doing a native port, then in my opinion it becomes cheaper and simpler to just do a native port and ensure a high quality result, with no Proton dependency to worry about.
Basically what I'm saying is, I think if Linux gaming gets to say, 10% of the marketshare, Proton will begin to naturally fade out and become replaced with natives anyway. And it will have a heroes funeral.
2.
Quoting: BeamboomI'm slowly gliding over to the conclusion that Steam Play probably is the solution for gaming on Linux.
Don't get me wrong, I will always be eager to support native binaries, but let's face it: The companies/individuals who work on porting to Linux are way too few and the interest at the game developers too low.
Absolutely. The way I see it, the battle strategy is thus:
Gamers can't game on Linux because there's no games?
Game devs won't port to Linux because there's no gamers?
Fix: Bring the games to Linux, so the gamers can come, then the game devs will follow.
"How do we know this won't kill native games?"
Here's my logic: Why doesn't Linux get more games?
Is it hard to port a game to Linux? No, not now, not with engines like UE4, Unity, Godot, or even the in house engines the likes Ubisoft and EA use, as their software is designed to be flexible so they can easily port to new game consoles and new platforms. We've seen that with Stadia.
Bringing a game to Linux doesn't mean porting it, it means supporting it, and there in lies the issue.
Support is more or less a fixed cost, despite how many customers you have. Support includes testing, checking compatibility, running tests on multiple hardware configurations, providing help staff, etc. That's not happening right now because "not enough Linux gamers to justify that cost".
Proton doesn't just offer porting, it also offers support, kinda. Valve takes the heat of support and supplies the commitment to porting (updating Proton), the game developer just gets sales.
The way I see it, the more people who use Linux for gaming, the more Linux gamers there are, the larger the potential marketplace gets. The more sales those game developers are going to get.
Sure, at first, for some game devs they will weigh up that market and say "Still not worth it to do a native port". But that revenue from Linux sales won't go unnoticed, and eventually it WILL make management of some game developers ask the question, "What are we doing to ensure we're compatible with Proton for those extra sales?".
Then we'll go through a period of time, where game devs are willing to 'half commit' to Linux, putting minimal effort into making their games run via Proton as long as that doesn't represent a large commitment, to enjoy some additional sales. That will result in many games working via Proton that currently don't, which will only increase the number of people who game on Linux.
So at what point does Proton cease to become relevant and suddenly get replaced with native games?
Hypothetically if Linux had 90% of the marketshare right now, would game developers continue letting gamers play their games via Proton? No.
Eventually, Proton ceases to become useful as a compatibility solution with minimal cost, and instead becomes a dependency burden. Just look at the number of times an update to Proton has caused a regression, if you become dependent on Proton, then you're subject to those regressions out of your control.
Supporting a game means testing, ensuring compatibility, performance, etc. Not because you want to, but because you want sales from your customers, and you want the game to run on their systems reliably, because you want those customers to tell their friends to buy the game.
Proton is an extra dependency, extra layer of complexity, an extra drag on performance, etc. When the costs of ensuring high quality Proton compatibility become larger than the cost of just doing a native port, then in my opinion it becomes cheaper and simpler to just do a native port and ensure a high quality result, with no Proton dependency to worry about.
Basically what I'm saying is, I think if Linux gaming gets to say, 10% of the marketshare, Proton will begin to naturally fade out and become replaced with natives anyway. And it will have a heroes funeral.
9 Likes, Who?
Quoting: Salvatos...Steam Machines . . . Sure they're not selling them anymore...
You are, of course, correct in general principle, but would it be the Internet if there were no one around to split hairs?
https://www.cyberpowerpc.com/page/syber/syber-s/
https://www.cyberpowersystem.co.uk/page/syber/steam-machine/
Obligatory Monty Python and the Holy Grail "I'm not dead" quote. CyberPowerPC's Steam Machine offering is still available for sale. On the UK site it can be configured from a selection of current generation hardware (and apparently can be shipped internationally). On the US site the hardware configuration options are somewhat dated, and it can only be shipped to Canada or the US.
4 Likes, Who?
Quoting: vectorI'm only operating on hearsay, so I'm happy to be corrected :)Quoting: Salvatos...Steam Machines . . . Sure they're not selling them anymore...
You are, of course, correct in general principle, but would it be the Internet if there were no one around to split hairs?
2 Likes, Who?
Comandante Ñoñardo Jun 27, 2019
As I said before: This work justify the 30% cut.
We will never see a feature like this from EPIC, Ubisoft or EA.
We will never see a feature like this from EPIC, Ubisoft or EA.
7 Likes, Who?
Quoting: Comandante ÑoñardoAs I said before: This work justify the 30% cut.
We will never see a feature like this from EPIC, Ubisoft or EA.
*20-30% cut
4 Likes, Who?
Patreon
PayPal
See more from me