What I find most amazing about it is that this has been about a decade in the making - Valve started their Linux compatibility initiative back when Microsoft threatened to hurt Steam with Windows 8 and Microsoft Store requirements.
How dedicated they are to this effort is impressive - they first failed big with the Steam Machines when support was very half baked, but now they'll try again with the Deck just as compatibility is coming along much better.
I guess Linux gamers can largely thank Microsoft for this one!
As a Mac user I'm a bit disappointed they stopped their efforts on that front, since the new processors are quite capable.
Not only that, but it has a lot of pain points, including ARM CPUs in a world where a lot of games used to target 32 bit windows.
Proton supported macos at first[1]. I think support is still more or less there, though not integrated with "steam play". Proton "glue" scripts are also mostly python, and macos does not (did not?) support python3.
I think the community could step in and help support it, but valve pretty much stopped efforts on that front (some rationale in [2]), although I think community efforts on wine would be welcomed.
Different graphics APIs aren't a reason not to develop for a platform. Game consoles have unique APIs, yet both AAAs and Indies happily develop for those platforms. What matters is whether the prospective income from those platforms outweigh the additional dev costs.
Additionally, most Macs simply aren't gaming machines; they're mostly thin laptops or AIOs with very high resolution screens, meaning the GPU has less cooling headroom and has way more pixels to draw than pretty much all Windows machines that are used for gaming.
That makes the potential pool of users even smaller than it already is, so most of the time it doesn't make sense to develop games for the Mac.
Those different APIs are usually wrapped by game engines like Unreal Engine or Unity. Porting a game to a new platform is almost always a business decision, and the actual porting cost is minimal compared to marketing and (customer) support cost.
Disclaimer: I have done some amateur game development, nothing professional.
> 1) Porting costs are not per game but per engine, which can be shared between multiple games.
That's not true. Majority of the costs maybe, but every platform needs its own tweaks, if for nothing else input often very different in a way engine abstractions are usualy not good enough. And then there is DRM, and anti cheats etc, that are quite platform dependent.
In pretty much any engine there are games that work beautifully and games using same engine that are like trash on fire[1]. There is a reason nvidia and amd often have to release new driver with fixes/optimizations for new games.
[1]: I meant cross platform support (where they work decently on one platform but are trash on others), not that game as a whole itself being trash and or buggy.
> 1) Porting costs are not per game but per engine, which can be shared between multiple games.
True but they're still big enough that that cost is significant. Even when diluted across game budgets.
> 2) The majority of costs is ongoing support, not writing to a different API.
Ongoing support is still required for writing to a different API. Ongoing support involves (a) updating in line with API version changes and (b) updating API-integration abstractions in line with game engine API updates. Supporting multiple APIs adds at least some amount of friction to every code-change.
There's also the increase in cost of testing every engine code change across multiple graphics APIs.
I’ve never worked on a game where we didn’t have at least one rendering programmer who would be assigned to a particular platform, even with Unreal or Frostbite or whatever. Major titles will still write their own platform-specific shaders, they’ll still tweak and modify the game middleware, and they still incur a per-platform, per-api costs.
Yes, this is the big one. Microsoft, Sony, and Nintendo have dedicated integration teams just for getting your specific game working on their platform. They might even release updates to their platforms to cater to what you want to do
Apple will update it's apple tv with a beefy graphics M chip. And enter the console gaming market well it's already in it but it will really start competing against Xbox and Playstation in the next couple of years.
By far most games on Windows don't run on OpenGL or Vulkan but on Direct3D. Game consoles also don't use OpenGL or Vulkan but have their own APIs. Metal is "just another" platform-specific 3D-API, it's not an exception but the norm.
In reality, "cross-platform" 3D APIs don't matter much (for games). What actually matters are distribution platforms (for instance: Steam), and if a distribution platform promises good business, games will be created for it.
Most AAA studios don't care about such standards, as proven by game consoles and Metal already being supported by Unreal, Unity, Ogre3D, Xenko/Stride, among others.
Games include music, so the answer for that part is "not at all". Many games offer music at additional cost, so there is a market for that. Also certainly some games made their music iconic in the same way movies used to do it.
Can't speak for everyone but I sure struggle to engage as much with long-form non-interactive media these days. If I'm not affecting a narrative in some meaningful way it's really hard to stay interested.
What do you like in terms of "interactive media"? Are you talking strictly games or something?
When you said "interactive media", my mind immediately went to tell tale games, "Wolf Among Us" [0]. It's more of an animated story with limited interactivity, but I really enjoyed it. Would love more of this kinda low key gaming.
Hmm. It set in back when I was playing WoW and realised that the effort I was putting into gaming the auction house was exactly the same as the effort I'd need to make money flipping items on eBay or something. After a while I couldn't make myself care about the in-game goals because I could just be using the same effort IRL for far more significant reward.
Then a few years later I spent a bunch of time playing Factorio, to similar effect. Downside: Suddenly flipped from "wow this is awesome" to "man this game feels like a job, it's no longer fun" to "uh, my job is basically this game so... yay?"
Now between the two it's really hard to get excited about something that isn't managing a real life business and building useful IP. Am I old? :P
It's not really an accident, Apple strongly dislikes AAA games (note: not AA games like fortnite or .. "games" like candy crush). If they support any kind of game development it's purely indie or puzzle style in nature and not high fidelity.
The reasons are.. well incalculable, there are so many:
To be competitive you must work closely with driver manufacturers; you would likely need to segment your desktop class hardware platforms based on what games they can play, you might bef forced to update hardware on a significantly shorter cadence- probably support for third party upgrades (like graphics cards). Also, Apple strongly dislikes nvidia who have a stranglehold on AAA games. (ever since they sold them a bunch of cards that desoldered themselves)
Apple wants to be a computing platform, and for "creatives". That's the image they're going for.
They can't win on gaming, even if they wanted to, but it's against the brand image to go for that market anyway.
You can't really do AAA gaming on a mac, and that's ok, not everything has to be about playing games.
The biggest rub on desktop GPUs on macOS is Nvidia insisting on full control of the drivers, which doesn’t jive with Apple’s attitude that GPU drivers should ship with the OS due to how deeply integrated they are with the kernel. AMD shares its driver source with Apple which works around this and allows RX 6000 series cards to work under macOS, but Nvidia would never consider doing something like that in a million years because too much of their business model revolves around total control.
It’s actually closely related to Nvidia long being a problem for Linux users, where open source drivers for Nvidia cards are permanently throttled to minimum performance levels thanks to Nvidia enforcing use of their proprietary drivers in its hardware. Incidentally AMD is also better behaved here — the FOSS AMD drivers have been quite decent for a long time now because a few years ago AMD got out of the way and started collaborating with the Linux community.
Apple loves AAA games if they're on the iOS store and bringing in billions USD per year of microtransaction revenue from teens and compulsive gamblers. They can't get that by supporting native APIs for non-App-Store desktop games.
While I was mainly talking about MacOS (the desktop); Does iOS support AAA games?
I haven't been promoted any AAA games on iOS- fortnite is not a AAA game.
Edit: argue your point phantom downvoters.
Fortnite is not AAA; God Of War, Horizon Zero Dawn, Tom Clancy's the Division are AAA. Find a source that calls fortnite AAA. Find any source of a AAA studio working collaboratively with Apple and I will retract everything I said.
Just Dance Now[0] is not a AAA game. It was made by Massive Entertainment, who makes AAA games.
So, a game being produced by a AAA studio does not make it a AAA game.
What "is" and "is not" AAA is debatable, sure, but some things are clear:
1) they're not mobile games.
2) They have a lot of marketing budget
3) They generally push the limits of the platform (lots of LOD)
4) They have large development teams
AAA usually means that a game needs: Lots of time, Lots of Money and Lots of People.
Fortnite is low fidelity, has been in early access since it's inception (many years at this point) and has a relatively small number of people (<100 direct, compared to >1,400 for The Division) making it, even though it's been "in development" for many years, not AAA level.
FWIW I'm not saying AAA is "good", I'm just saying that Apple isn't chasing AAA- and "fortnite" is not a good counter-argument, the better argument might be call of duty or sims, which have mac ports, but they're definitely not first platform and they actually are emulation shims using Wine.
Fortnite itself does not claim to be AAA, and it's totally fine not being AAA, it's actually better to avoid making AAA games and instead go for "AA" with low fidelity because the stakes are lower and you can take more risks in making a more interesting and engaging game.
Go to game credits, count everyone until they mention a publisher.
Industry convention: Project only: Main studio first, ancillery studios next, outsourcing parties next, then the scope broadens and includes the publisher, HR, office team etc;
But anyway, Tim even says that they shipped with only 25 people. Even if they broaden it now the scope was defined early.
If you really want fortnite to be AAA, I concede, I don't really care.
But then: please give me at least 1 other example.
>It's not really an accident, Apple strongly dislikes AAA games. If they support any kind of game development it's purely indie in nature and not high fidelity.
Except that's not true at all when it comes to iOS. Back when Epic and Apple actually got along, it was common for Epic to be part of iPhone keynotes showcasing the graphical capabilities of the new GPUs.
I think Fortnite was also regularly featured on the App Store back when it was actually available there.
Obviously they have a much larger interest in supporting gaming on iOS since every game sale brings them 30% of the cut, but there's no general AAA game hatred at Apple.
That's certainly true, and for most people making games on PS4; their openGL has a weird DX11-ish 'shim' that you can use to make your code cleaner (at the expense of a bit of performance).
It's used heavily in Snowdrop; but in order to work on Stadia they had to abandon the shim and work on Vulkan. So, even when porting to linux GL was not preferred. (I worked there during this transition).
With Valve being founded by a bunch of Microsoft employees, maybe they took a lesson from Microsoft: if you have enough money to keep trying, eventual success is likely. Microsoft did that with many products, most relevant in this context: the Xbox series.
As strategists use to say: "Wars are lost". If you can maintain yourself long enough, your enemy will eventually be out of resources to maintain themselves.
Google doesn't "keep trying", it rewrites replacement products frequently. The nearest Microsoft equivalent to Google's (lack of) strategy was Skype being replaced(? Complimented by?) "Skype for Business"
The Steam Machine was very effective at convincing MS not to lock down 3rd-party app marketplaces (as they were threatening to do before it came out). It was really more of a threat than a sincere attempt at a commercial product.
The Deck certainly is, it's sold out into 2023.
People want PC games on the go, Valve provides it in a now common form factor, at a reasonable price, with a trusted name.
At least 110 thousand units were preordered on the first day [1]. I agree being sold out is not sufficient evidence of popularity but that is not the only evidence.
A contrarian and provocative tone like this invites downvotes, I'd prefer to see this discussion remain constructive instead of throwing accusations and hyperboles that derail into a flamewar. Given the low-effort content of these replies, I would consider that trolling.
I do not completely disagree with the parent though. We did not get exact figures, and in my circles, only die-hard geeks have heard of the Steam Deck, even regular gamers who spend a lot of time on Steam didn't know what it was. Granted, I asked them just a little after the announcement.
Anyway, I am not sure valve is trying to rival Nintendo here. They probably want Steam Deck to become profitable, and I think they would welcome competitors who come up with Steam Deck alternatives, especially as SteamOS becomes available.
But the case being argued is demand for the Steam Deck which is demonstrably high.
The counter point to that was that production was low, therefore selling out doesn't mean much. However, the Steam Machines didn't sell out at all so at the very least there is more interest in the Steam Deck than Steam Machines.
Sure, it doesn't follow that the Steam Deck will necessarily be a success, but the GP's arguments seem lazy and bad faith, which is presumably why they are being downvoted.
I will get downvoted for every remark made against Steam and the dream of Linux taking over the world of AAA games (originally written for DirectX/Windows anyway).
Whether or not Steam Machines sold well or not is irrelevant. They were an intermediary step to them not being beholden to an OS vendor. In the last couple of years, they have arrived at their destination -- games run on Linux unless they have anticheat that prevents then from running on Linux.
If they overcome the anticheat hurdle and reach 100% of top games running with decent performance, there are plenty of people who have a computer that they use exclusively for video games and web apps that will gladly forego paying for Windows.
> If they overcome the anticheat hurdle and reach 100% of top games running with decent performance, there are plenty of people who have a computer that they use exclusively for video games and web apps that will gladly forego paying for Windows.
Windows is essentially free these days. You might have to pay for it once (and I got my Windows key from a grey market site for $15), but each Windows version update has been free after Windows 8.
Any gamer that buys a pre-built computer will end up with Windows by default and isn't likely to wipe it to install Linux. And even the ones that build themselves are still likely to use Windows just because that's what they're used to.
The overlap between PC gaming and Linux enthusiasm isn't as big as you think it is.
Everything you said is definitely true as of right now, but if Microsoft somehow finds a way to extract a cut from every transaction on Windows (fears of this around the release of Windows 8 is what initially spurred Valve's Linux investment), Valve actually would stand a chance in hell of convincing people who use their computer for video games and webapps nearly exclusively to switch over to Linux if games and microtransactions worked and were cheaper.
We're not really talking about the group of people that are into both PC gaming and Linux. We're talking about the group of people that are into PC gaming that would be fine with a Chromebook if they weren't playing games, and would consider Linux if games were cheaper on it.
It's also worth noting that you said 'plenty' which is absolutely true.
A rounding error for Windows install base would be a significant increase in Linux market share, and that is the sorts of levels I would anticipate with complete support and a successful launch of the Steam Deck (increase in market share by 1-2%)
They are relevant as they were released with the same kind of party "we will take Windows down!", lets see how much the people actually learned from history.
That is a laughable characterization of the business decision Valve has made, and learning from history isn't worth much when the landscape has changed so much in the last five years. Have you actually tried to run games on Steam on Linux recently? Did you try it five years ago, and have an understanding of what it was like then?
Proton and Steam on Linux _are a deterrent_, and within the last couple of years, it can be said with confidence that the strategy is viable and worth continuing. This was never about selling consoles -- Valve makes their money from their cut on sales and microtransactions.
It's true that this can still not go the way Valve wants, but it's like counting on me to do a standing double frontflip because it's not strictly impossible. The only thing in the way of near-100% compatibility is anticheat, and EAC and Battleye both plan on supporting WINE/Proton when Valve releases their hardware. EAC has actually supported WINE for a while now with a little legwork on the part of the developer, so it's not even a question of whether or not it's technically feasible.
Once Valve achieves near-perfect compatibility, Microsoft can't change legendarily backward-compatible APIs for the sake of pissing in Valve's cornflakes. This doesn't mean that there will be a large Linux gaming market share now or in the future, but it does mean that Valve has an option if Windows starts to threaten their cut.
I feel like Microsoft is spending a lot of time shooting themselves in the foot right now, though.
In the days when "DOS/Win31 versus OS/2 with compatyibility" question was asked, OS/2 was the expensive choice, both to license and in terms of hardware requirements.
With the high Win11 hardware requirements, Linux is actually more accessible as a supported platform, both on the low-end (i. e. you could strip it down a lot further for a device with small storage) and for a large number of older enthusiast PCs (6th and 7th generation Core or first-gen Ryzen)
You also have an incredibly weak buzz for WIndows 11 right now. Some people are up in arms about stuff like "requires a Microsoft account on the affordable version" or "we've moved the UI in weird and confusing ways" but nobody is saying "it provides cool features A, B, and C to justify the hassle."
A well packaged Linux launch could provide a compelling message just from the perspective of "here's an OS where all your games work, but it's not also a ticking support timebomb."
TBH, I feel like the old early-2000s Linux style of "let's choose custom window managers and theme the crap out of everything" would also appeal to the gamer crowd, but I feel like that's sort of disappeared in modern desktop Linux.
Could you expand on that? I don't see the situations as being directly comparable at all. Maybe in technology terms, but the situational context is completely different as far as I can see it, especially as it pertains to Linux share of the Gaming market.
Basically Proton allows game studios to focus on Windows and DirectX, just like what OS/2 got were ported Windows applications and only a tiny minority bothered to natively target OS/2.
As the article itself mentions the Linux hardware on the survey is still at 1%.
Strongest part of the article seems to be "Call it Valve’s middle finger to Windows", which is author's comment. AFAIR, Gabe Newell didn't like MS policies with windows 8, but I never got the feeling of SteamOS release being anything close to "we will take Windows down!".
Please take this as sincere advice: avoid comments that show contempt and disdain. They're disrespectful, distract discussion from main points and are not good for you or anybody else.
It's not the first time I'm a target of such comments from you.
> Good luck with that, Steam Machine were selling like hot cakes.
I assume you write "Steam Machine were selling like hot cakes" using irony and sarcasm. At the same time "Good luck with that" is like a recommendation on something you guess will fail or merely derogatory.
Irony and sarcasm aren't conducive to healthy discussion, though. Ultimately if I'm going to show respect to you as a participant I would expect to have respect shown in return.
In terms of your actual contribution to discussion it means your points are weak and easy to counter.
I game mostly on Linux and I remember just a few years ago having anxiety while purchasing a new Windows game because I wasn't sure that I would get it working.
Since then the support has become so good I have practically no anxiety at all. If the game doesn't launch first try it's usually just a couple settings that need tweaking. Most things just work right out of the box.
Do you have any anxiety around how the games work after launch? I enjoy a lot of older games and sometimes they glitch in some cases. A VM of the original target platform seems the most reliable IME.
In particular the game Uboat stopped working after an update and I was unable to get it going. To be expected however, as the game is still in Early Access.
Another early access title that I ran into problems with after launch was Postal 4. I found a workaround is to replace the bundled direct X installer with some other silent installer that returns the expected return code.
Other issues tend to crop up around custom launchers. Usually in this instance I will search online for the command line arguments to bypass the launcher or simply launch the game directly.
The last subset of issues I seem to experience revolve around fullscreen/window mode. I find that some combination of ALT+Enter and F11 typically resolves this. If not I usually resort to running the game in a Window and undecorating it. However in some games this does not capture the mouse and can cause the window to lose focus if you click outside of the game window.
gamescope should help with that by running games in their own Xorg server as well as handling the scaling, and gaming on wayland/sway helped me a lot with getting games to behave consistently with other programs. It does have its own issues sometimes, certainly. I hope gamescope will help route around them.
Apple is not even against other operating systems on Macs. Bootcamp has been a thing for a long time (and apparently they're not against ARM Bootcamp were it possible [1]) and recently marcan mentioned how Apple added something that wouldn't be helpful for Apple, but is helpful to the Asahi Linux project.
I don’t understand the complaint. Steam is available and Apple clearly is playing ball with them, so what’s wrong? Is that the Windows API translation layer isn’t done?
Dealing with Apple isn't something Valve would want. They focus on Vulkan, Mesa and Wine. And Apple are pretty negative about Vulkan for some unknown reason (which seems to be even more than their usual NIH mentality). I.e. like when they were resisting adopting SPIR-V for the Web.
So Apple have no one to blame but themselves for it.
Starting at Win7 the telemetry and invasive advertising ramped up to the point where I don't feel I can customize the Desktop to what I want without heavy tweaking.
And if that amount of tweaking is necessary why not go Linux instead.
So got rid of my Win10 install, went with KDE Neon+Steam. And now I wonder what took me so long.
I haven't tested my entire Steam library yet but my top 10 games run perfectly, some times with some changes to the Proton version but beyond that even workshop mods are available.
I am seriously impressed and can't imagine myself going back at all.
If only Adobe would release their software in Linux, then, I could ditch MacOS too.
Been running Fedora + Steam for over a year, now. The experience has been great. It’s hard to imagine going back. I do miss Mac hardware, though. My 1.5 year old Dell XPS already has a bad key, and a faulty WiFi card. It’s complete garbage.
My next laptop will be the Framework 15” 4K option whenever they make one…
I've had plenty of Macs, including the butterfly keyboard (which did go bad on me). But. It took that butterfly keyboard longer to go bad than the Dell XPS. The Dell is honestly a steaming pile of crap.
Unpopular opinion: I miss my butterfly keyboard. I never had any of the flaky-key issues. And I loved the crisp feel of the keys.
I don’t need key travel. I like the futuristic (and unrealistic) idea of typing on glass. In some ways, the butterfly was just a step away from that, but with the slightest bit of physical feedback.
At the least, I definitely did not loathe the butterfly like mini.
Same here. The butterfly felt much better than the Magic Keyboard. I also liked the way it sounded.
What I didn't like is that it went bad after 2 years of not very heavy use (I have an external keyboard). If Apple could make a reliable butterfly keyboard, I'd love to have it.
OTOH, I appreciate the new ones having a keyboard that feel closer to the pre-Magic ones (same feel as the pre-butterfly laptops). It's OK, if not a revolution.
Just curious, but do you use the laptop’s keyboard when playing games? Even for the limited style of gaming I do I can not imagine not using a full sized wireless keyboard and mouse.
I mostly use a real mouse and good mechanical keyboard for work, and a PS3 controller for play. This is why the laptop’s keyboard failure is even less excusable.
The way that Blender is developing they could get into trouble. Might not yet fit perfectly into the toolchain of many artists, but I guess it will make ground in a professional setting. I haven't missed photoshop either, I only paint and model as a hobby though.
I've also had a really suprising experience with my linux gaming setup: Playing old retro windows game has become easier, because wine can run a suprising amount of things. I recently itched to play Civ2 and after some wrangling with the setup file, I can now run Civ2 entirely transparently on my linux box. Same for a bunch of other Win95-era games - no need to fiddle with compat settings on windows and hoping, a lot of them just work in wine. That's pretty amazing.
The only thing I've found so far that just doesn't work is Tiberium Insurrection, because it does some weird things I don't understand. But with the way Win11 is going, I rather take that kinda problem.
Interestingly, I have windows software that were working with wine previously but new versions don't seem to work for some reason. I used the new bottles thing. Compared to a copy of among us which played seamlessly.
Unfortunately you can't purchase a copy of Photoshop on any platform. You can only buy a subscription, and what's more starting Photoshop starts 5 or 6 other programs that slow down your system and does who knows what. Fuck Adobe, just another company that doesn't care about users.
How is the performance of games? What I understand is that proton is a layer/adapter to run windows games and should create a FPS hit. Or do you get the same performance of running on widows?
Modern games can be run on Linux three ways: with native Linux builds, with Proton/dxvk/d3dvk or SteamPlay, or with VFIO.
The Linux native builds tend to be a version behind or have poorer support than the Windows ones. It's based on the devs, of course, but the take away is that it's historically been a crapshoot.
SteamPlay/Proton provides native or better performance over Windows for what I play. I've been using it so long that I actually don't know which games have native Linux versions anymore. It 99% "just works".
VFIO is virtualization where hardware is passed through to the guest OS. I use KVM+qemu on Linux and attach PCI devices like GPU and storage to the VMs. This is a native Windows game build on a Windows VM on a Linux host. It performs within a tight few percent of bare metal Windows, and the host is fully functional while in use.
Linux gaming options range from direct and inflexible, to feature parity with native performance and compatibility, to complex but guaranteed native support. I've relied on the first and last less and less over the years as Proton has improved.
> VFIO is virtualization where hardware is passed through to the guest OS. I use KVM+qemu on Linux and attach PCI devices like GPU and storage to the VMs. This is a native Windows game build on a Windows VM on a Linux host. It performs within a tight few percent of bare metal Windows, and the host is fully functional while in use.
How about displays, though? It doesn't seem like you could have a proper integrated desktop between host and VM when using VFIO.
Lookingglass, as another user mentioned, peers directly into the framebuffer for an effectively latency-free display experience on the host.
You can also connect the graphics device that's passed through to an external display.
As an example, my SO and I can both play a multiplayer game from the same computer this way. I can use a VM with lookingglass on my Linux host while she uses a monitor connected to the GPU her VM has attached, along with a passed-through hub with mouse and keyboard.
It helps that Threadripper has two NUMA nodes, but it's possible on any system with good IOMMU grouping.
> As an example, my SO and I can both play a multiplayer game from the same computer this way. I can use a VM with lookingglass on my Linux host while she uses a monitor connected to the GPU her VM has attached, along with a passed-through hub with mouse and keyboard.
You could also run regular multi-seat Linux with such a setup.
It should also be possible to do it without multiple GPUs, but that's unfortunately more involved (if anyone has a simple way to do this in wayland land, I'm all ears!).
> What I understand is that proton is a layer/adapter to run windows games and should create a FPS hit
Your understanding is wrong. It's largely a reimplementation of Windows APIs. Those implementations can perform better or worse than the real thing depending on quality of implementation and the underlying OS.
Some games will perform better on Linux. Some will perform worse. Some are about equal.
Well, it can be a layer as well. If a game makes DirectX calls, and those calls are translated on the fly to corresponding Vulkan (or in some cases perhaps OpenGL) calls, that adds a layer. That is, unless the calls are AOT or JIT transpiled or something similar -- I don't know if they are.
System calls are also something that would require an additional layer, as the required Linux system calls need to be made in addition to intercepting the Windows system calls in Wine code. Some other APIs in Wine may also be implemented on top of corresponding Linux APIs rather than just being standalone replacements for Windows code.
The majority of code in Wine is probably just reimplementations of Windows APIs, which could be thought of as being parallel to the Windows implementations rather than layers between a Windows API and a Linux one, so you're correct there.
I'd generally expect there to be a performance hit on average, but as you said, it doesn't always happen, and it's not as large as one might expect if thinking about an emulator or something similar.
most games the performance hit is going to be not noticeable when to some, mildly annoying to others, and totally unacceptable to o a few. Like if you were trying to be competitive in a modern AAA fps, prolly won’t be ok. But big non twitch 3d games are still very good!
'twitch' in games typically means latency-sensitive, like a competitive FPS where being 10ms faster to shoot than the other guy means you win the round. Many games can handle 100ms of latency no problem, so running at 50fps instead of 60fps is no big deal. On consoles those games typically run at 30fps to begin with, which is at minimum 16ms more latency than playing at 60fps on a linux or windows PC.
Pretty miniscule. Putting things into perspective, it's still fully possible to hit 120 FPS on my GTX 1050ti in Overwatch. A more direct comparison can better elucidate the tradeoffs though: https://youtu.be/voXc1nCD4IA
I've got a few hundred games and the only one I noticed was noticeably worse on Linux is Star Citizen, and that's a game still in alpha so the optimisation in general is awful anyway.
IME it's not that great a hit in most games, maybe 5fps. Admittedly I have super high end hardware so 5fps isn't as large a hit someone might take on low-mid end hardware so YMMV.
A 5fps hit has no meaning. 6fps->1fps is much different than 1005fps->1000fps please use mspf (milliseconds per frame) or at least a percentage of the fps lost.
I wonder if Wine can be used to run Windows versions of adobe software on linux. The adobe software is the same thing preventing me from switching to using a linux distro full time.
actually, proton is two things: wine+dxvk. dxvk is a windows (!) program that translates directx to vulkan. between the two of them theyre all thats needed to run games.
proton is in addition funded and tested thoroughly by a critical audience, linux gamers ;)
I've had no luck with Photoshop in Wine or VB seamless mode (PS's custom window decorators are probably the reason for the latter). A full-desktop Windows VM is still good enough with guest extensions though, so I can easily resize the whole Windows desktop with just a maximised PS window inside.
The recent Linux Challenge [0] on Linus Tech Tips has helped raise Linux's profile for gaming. Linus Sebastian's enthusiasm for the upcoming Steam Deck has really raised the channel's appetite for Linux Content and as such is exposing it much more to its audience.
We also have Anthony Young to thank for laying the groundwork for Linux Content on LTT, and in turn we have Linux content on LTT to thank for a greater awareness, acceptance and adoption of Linux among gamers (helped, of course, by the advancements being done by Valve with Proton, Steam Play, the Steam Deck, SteamOS 3.0, Steam Input, DXVK and all the rest).
My daily driver is Ubuntu 20.04, and I stopped dual booting with Windows back in December 2019, which up to that point had been the way I did all of my gaming.
I haven't looked back since.
Sure, there have been some games that haven't worked. Some of those just needed some tweaking, and for some I just waited a bit until Proton caught up. For the rest that still don't work, I am content with just shelving until such time that they work.
I have been consistently surprised at what games actually work through Proton. For instance I played Titanfall 2 & Battlefield 1 without any issues this year. On Linux.
I probably spend around 200$ each year on video games, and this initiative by Valve will be the sole reason I will continue to do so (and not having to run a Windows instance in my home).
What if you want to install games from other stores, like Epic, Twitch, or GoG? Despite never buying any games there, I have a bunch that they gave out for free.
I don't have Epic or Twitch accounts, but Gog works excellently -- there are several alternative clients (e.g. lutris) that let you install Windows games into a proton prefix through GOG with a nice gui. Plenty of GOG games are actually linux-native -- and they "just work" quite well, amazingly including lib / .so constraints -- they have a standard script that does quite a lot to fix this. If you don't want a nice gui, there are a ton of CLI downloaders that "just work" and save you clicking buttons a lot.
GOG Galaxy for linux is "in progress" and has been for ~7 years. Alas despite being the most-requested item by far in their feature-request tracker, I think it's unlikely to happen soon. Galaxy itself works well inside a proton / wine environment, and will auto-update Windows games.
There's a program called Lutris [0] that will handle GoG, Heroic will handle Epic [1]. Not sure about Twitch games, though, as I'm unfamiliar with their offering. If you mean Itch.io then Lutris handles that too.
Most of those games are redeemed through other clients like Origin or Epic. There are some games only available through the amazon gaming app, but they're mostly small indie titles.
Epic games come to Steam when they leave Epic Access, I had no idea Twitch could install games, and the games they'd support are likely the kind of trash I'm glad is being filtered out from other stores, and GoG is DRM free and therefore easily downloaded. Their client sucks anyway.
If it's not on Steam, you'll have to use Lutris. It's a slightly better experience than having all those stores on Windows, but not as simple as Steam.
> Linux machines make up only a bit above 1% of all Steam users.
to my knowledge, this number has remained remarkably consistent over the past 5-ish years, even as Proton/Wine become more and more reliable. it seems that library compatibility isn’t that much of a determinant for which OS a typical user chooses to game on.
but if Valve’s console effort is successful, then that number would necessarily rise. and i can see a linux console replacing something like my Switch. at the end of the day, console gamers don’t really care about the underlying OS — just the hardware and the game library (and miscellaneous HTPC features, etc). if such things can be abstracted from the actual console, that gives a huge leg up to incumbents. might pave the way to some Cambrian explosion of console development differentiated only by hardware.
> it seems that library compatibility isn’t that much of a determinant for which OS a typical user chooses to game on.
I would argue that linux would need to be installed by default for that to change dramatically. Most people don’t even know linux is an option on desktop. Not that many people bother to change their search engine to something other than google, and that doesn’t require much effort.
Switching to linux is an investment and takes time to use - let alone understand it.
I can’t wait to see what follows after the steam deck is released. Hopefully with all the publicity companies will finally start taking linux seriously.
> I would argue that linux would need to be installed by default for that to change dramatically.
I disagree. There are good reasons people choose not to use Linux and it is annoying that the community continues to tell itself variations of this kind of thing so it doesn't have to face that fact. Even starting out a potential user has to navigate a ludicrous number of distros, all with slightly different ways of doing things and various levels of inadequate and out of date documentation, and as soon as they ask for help they'll likely be told to try a different distro regardless of which one they pick.
Even having Linux preinstalled won't change this because, as we've seen, every single manufacturer wants to build their own distro instead of throwing their effort behind an existing one.
> There are good reasons people choose not to use Linux
This can be true at the same time as "linux would need to be installed by default for that to change dramatically", and in fact the latter almost certainly is. There's strong scientific evidence for the claim that default values matter (e.g. with organ donation rates[1]), and no reason to believe that that wouldn't apply to Linux, as well.
> it is annoying that the community continues to tell itself variations of this kind of thing
Barely true, definitely not relevant. The majority of Linux users will happily tell you the flaws that it has.
> Even starting out a potential user has to navigate a ludicrous number of distros
False - you're conflating "available options" with "number the user has to care about". A quick Google search "which linux distro should i use" will have Ubuntu topping most of the lists, which is the one users should start with. Not a problem.
> as soon as they ask for help they'll likely be told to try a different distro
As someone who has asked hundreds of Linux questions from a half-dozen different sources and never once gotten this reply, this is definitely false.
> every single manufacturer wants to build their own distro
That's not the case for Android or Windows - why would it be the case for Linux?
You seem to have had a very bad experience with Linux and the Linux community. I'm sorry about that, but you have to realize that it's very non-representative with the average user experience.
> This can be true at the same time as "linux would need to be installed by default for that to change dramatically", and in fact the latter almost certainly is. There's strong scientific evidence for the claim that default values matter (e.g. with organ donation rates[1]), and no reason to believe that that wouldn't apply to Linux, as well.
Of course changing the default to Linux would increase numbers overnight, that's not what I'm arguing. What I'm arguing is that it is possible to increase the numbers pretty dramatically just by making a good product and that Linux Desktop isn't actually that great. If you want more people using Linux Desktop, the best way to accomplish that is to make it better.
> The majority of Linux users will happily tell you the flaws that it has.
Maybe, but by far the loudest members are the evangelists. That's what the outside sees, and that's who you invariably end up interacting with when you ask questions.
> False - you're conflating "available options" with "number the user has to care about". A quick Google search "which linux distro should i use" will have Ubuntu topping most of the lists, which is the one users should start with. Not a problem.
https://distrochooser.de/ is the first link I got for "which distro should i use". It's like one of those "which DBZ character am I" tests. https://linuxnewbieguide.org/overview-of-chapters/chapter-3-... is the second. But hey, let's say it is Ubuntu since that isn't unreasonable. 20 minutes later that Ubuntu user will encounter their first Arch user while asking a question about why Network Manager is doing something silly, who will naturally advise them to switch to Arch. I've been using Linux Desktop off and on for 20 years now and this hasn't really changed much.
> As someone who has asked hundreds of Linux questions from a half-dozen different sources and never once gotten this reply, this is definitely false.
Just more 'Works for me! (tm)", another common thing in literally any discussion about Linux. By saying that this is false just because you have not personally experienced it you are dismissing anyone who has had a different experience.
> That's not the case for Android or Windows - why would it be the case for Linux?
Actually it kind of is the case for Android, but since they have a unified package repo it isn't as noticeable. Regardless, my assessment is based on notable "Linux installed by default" manufacturers like System76 and Puerism using in house distributions instead of just Ubuntu.
> Of course changing the default to Linux would increase numbers overnight, that's not what I'm arguing.
I now understand what you mean by this, but it took me reading your comment three times to get it. You quoted "I would argue that linux would need to be installed by default for that to change dramatically." and then wrote right afterward "I disagree", which did make it sound like that was the thing you were disagreeing with.
I think that one of the implicit assumptions of "linux would need to be installed by default for that to change dramatically" was "without any other significant changes happening". I think that this is a reasonable assumption, because you can always imagine some extreme scenario e.g. Microsoft announces that it's going to hand over all user data to some large government, and millions of people stop using Windows overnight, or maybe they start charging $500/year for Windows 11 Home Edition - both highly unlikely, but also both causing Linux to gain marketshare without becoming a default. So, the assumption "barring major event" is a sound one.
...especially given that the quality improvement necessary to see a major Linux adoption uptick would be...probably a once-in-a-century computing event, given how much better Linux has become (and worse Windows has become) over the past decade with no significant change in Linux desktop use.
That's because you altered my query from "which linux distro should i use" to "which distro should i use". Actually using the query I cited comes up with [1], which lists "Ubuntu" as the very first item, along with that same value being cited in the Google infobox in the search results. Expanding "distro" to "distribution" (which a new Linux convert is likely to do) also gives similar results, so your example isn't representative.
> 20 minutes later that Ubuntu user will encounter their first Arch user while asking a question about why Network Manager is doing something silly, who will naturally advise them to switch to Arch
This is something that you're repeatedly cited, but I've never had happen to me, nor even seen happen to someone else in an online conversation, after going through thousands of questions...
> Just more 'Works for me! (tm)", another common thing in literally any discussion about Linux.
These situations aren't remotely comparable. "Just works for me!" takes a single data point and extrapolates it to invalidate others' experiences. I'm taking thousands of datapoints (things that I've read: forum posts, Stack Exchange threads, IRC chat sessions) and generalizing that to say that the experience that you're describing is unlikely - which is 100% valid, because unlike hardware issues, every web search is distinct. If "switch to <other distro>" is common advice, then the probability of me encountering it once as I search the internet approaches 1.0 very quickly.
So, yes, if I've trawled through thousands of various kinds of internet posts about Linux and never encountered someone saying that a question-asker should switch distros, then it is not "likely" at all.
> You quoted "I would argue that linux would need to be installed by default for that to change dramatically." and then wrote right afterward "I disagree", which did make it sound like that was the thing you were disagreeing with.
..because it is? The statement asserts that being installed by default is a requirement of increasing Linux adoption 'dramatically', whereas I assert it is possible without installing it by default everywhere.
> I think that one of the implicit assumptions of "linux would need to be installed by default for that to change dramatically" was "without any other significant changes happening"
I don't see how it could reasonably be read that way: "Absent any changes except the one I propose, nothing will change unless my proposed change occurs" is a tautology. Not to mention that that interpretation disregards the context of the rest of the post, which very much makes it sound as though parent believes there are no significant barriers to adoption and the only reason people don't choose Linux is that people by and large don't really choose at all.
> There are good reasons people choose not to use Linux and it is annoying that the community continues to tell itself variations of this kind of thing so it doesn't have to face that fact.
This is true, but I think it's actually irrelevant to the point GP is making. Even if Linux had absolutely no downside for any user in this community, and every person who would want to use it switched to it, Windows is so ubiquitous and does a good enough job that I doubt the given 1% number would raise more than a few percentage points.
At 1%, any extra percentage points represents at least a 2x increase. That's not dramatic? I would think a 10%+ increase would be easily achievable between Glorious PC Gaming Master Race, People What Have Entirely Too Serious Hobbies, and People Who Use Computers For Work if said people actually wanted the Linux Desktop experience.
At any rate, I can't back up my hypothetical quantification of "if Linux fixed some of its most well known issues" gains, it's just a gut feeling from 20 years of interacting with personal computers and the people who use them.
I agree. Windows is "free" for most gamers too. If you're building your own PC you need to buy a license, but the majority of steam users are definitely on pre-builts or laptops.
Dual-booting needs to become easier and less bug-prone. People need to be able to try Linux before committing their main system. One year ago I tried setting up dual-booting on my system with Ubuntu and it irreparably trashed the Windows boot process; and I needed to do a full nuke & pave. At this time I don't perceive installing Linux to be a good idea unless you're buying a dedicated Linux workstation from System76 or similar.
Looks like it has been flat for 3 years and then suddenly grown a bit in the last 6 months. Don't fall for the linear trend line on a graph that clearly isn't linear.
Ive been gaming on Fedora for the past seven years. I think it has limited what I buy and play, so at tgis point I dont know what I'm missing, but it looks like at least 75% of steam's top 100 games work fine on linux [1].
It would be really nice to have it as an alternative since games bind a lot of users to Windows, which becomes a more and more intrusive OS without providing too many benefits to the user.
Valve already had fears about Windows 10, luckily their store and Windows platform wasn't successful in the end. But their intend is still pretty clear in my opinion. In the long run this will also benefit developers since they would have a dependable system to base their games on.
> Valve already had fears about Windows 10, luckily their store and Windows platform wasn't successful in the end.
A few years ago I'd have agreed with you, however the xbox game pass has been hugely successful for Microsoft to push their store and platform. I think a lot of PC gamers would struggle to switch to linux because of that alone. Looking at steam stats alone is selection bias and doesn't tell the full picture.
If you've ever tried using the Xbox Game Pass for PC, I think them offering it in itself will move users away to other stores and platforms. Not only are download speeds abysmal, but it frequently takes forever to even start a small update. Oh, and no option in the GUI to backup your games. You can only change your home computer 3 times. Reinstall your computer a few times and you will no longer be able to use offline play for a full year and you'll have to wait days to reinstall your games.
I am a subscriber and I will admit I hate so many things about the app, I genuinely don't know how a company the size of MS can make the UX mistakes it does.
Anecdotally though, I haven't seen the issue you mentioned with the download speeds. It always maxes my 150Mbit connection (probably pitiful for some parts of the world admittedly) so I've never felt a need to backup games because I can re-download them whenever. I'd be interested in hearing about this 'changing home computer' limit too, I've got my game pass running on 3 computers right now so I'm curious if I've hit the limit.
As I mentioned in another post, I am a daily linux user and want to see it become a wonderful gaming platform. That said, even with the UX issues of game pass, it's hard to argue on the price point. For a none techy who doesn't care about OS choice or brand loyality, I think it's going to win in the long run.
True, game pass is a really good deal if you want to save money and you don't care about ownership as much (also a problem with steam, but to a lesser degree). But they still struggle on PC. They just recently rebranded it to pc game pass instead of xbox game pass and they want to improve the horrible app.
I think Steam will at some point perhaps react to that, but many PC gamers are still quite vary of MS and I don't think it has affected their business too much. Not only because their older sins like GFWL.
Well, without the hard numbers on subscribers it's going to be hard to quantify one way or another if microsoft are 'struggling on pc' or not. Needless to say though, their catalogue of games on offer for a small monthly fee is absolutely insane. From various twitch streamers/youtubers in the gaming industry people I've only heard praise for its value for money for the customer (anecdotally of course). Clearly microsoft are throwing money at the market right now for growth, how long it lasts I don't know.
Don't get me wrong, I have been a linux user for 16 years and would love to use it for my gaming needs, I think we just have to be realistic here. MS are pumping money into the gaming industry, through the game pass and buying studios. They are not going to let linux become the dominant gaming platform without a fight.
I’m guessing what they meant is the Windows version of the Xbox Game Pass, very recently renamed to “PC Game Pass”.
Microsoft has been moving gaming on windows and xbox as a united front quite a bit in recent years, they seem to have come to the conclusion that keeping things console exclusive isn’t a priority any more (still exclusive to environments they control, of course).
I bought Doom Eternal and Overload (spiritual successor of the 'Descent' franchise) over the cyber-Monday, and still haven't spoiled my uptime by booting Windows play them. With Steam's Proton there's been almost no problems or even obvious performance loss. Smooth 4k 60hz experience for both on Ubuntu 20.04 w/ Geforce 1070. No tricky optimizations or drivers.
The only issues were some audio crackle in Doom before I implemented the forum wisdom to set PulseAudio to 48khz, and, the annoyance that 'tab' stops working in game if I alt-tab out and back.
One of my favourite things about the great work done on Proton is that some Windows exclusive 90s era games of my childhood run better on Linux than they do on modern versions of Windows.
Live in the pod, eat the bugs, load the anticheat kernel module.
Or don't play the game at all.
Viability of Linux as a game platform means that sooner or later, most of the shitty things about gaming on Windows (root-your-box anticheat, root-your-box DRM, a million shitty storefronts) will come to Linux because they're table stakes for the publishers' business model and without them the games simply won't get ported.
Anti-cheat is really only a problem for the big multiplayer games.
But it’s a huge problem. Cheats are big business for the top multiplayer games. My gaming friends and I simply stop playing for a while when new cheats come out and run rampant because losing to cheaters every other round gets old fast. I understand why developers have to do something about it.
If a Linux version was released and deemed easier to cheat with, scores of cheaters would be installing Ubuntu that night just to start cheating. This is why I don’t see Linux versions of AAA multiplayer games coming any time soon.
Very good on summarizing the problem. No, it's not sales, it's because some people are outright horrible and will do anything in their power to get those imaginary trophies. If someone has a decent solution on this that doesn't rely on client-side DRM, I'll (and many, many others) would gladly hear your solution, as long as it's actually workable and not "ban them" (everyone tried that but it's easy to evade) or "actually, give your personal info" (I think everyone will agree this is worse than most DRM solutions).
> I think everyone will agree this is worse than most DRM solutions
You think wrong. A social credit score system is odious for real life, because the only way not to play is to die. On the other hand if you ruin your gaming reputation as a cheater and nobody wants to play with you, well that's on you. Playing CoD or whatever isn't a human right.
It's also ethically acceptable because it would be purely opt-in. Cheaters and others who don't want to provide proof of identity are welcome to play with other players who don't care about cheating.
Rooting my computer is a far greater violation of my autonomy than asking me to prove who I am.
I'd like to see games integrate a "trust score" mechanism where known configurations like immutable distros e.g. Fedora Silverblue, cloud gaming platforms and consoles don't have to run a client-side anti-cheat at all because they already are using known, unlikely to be hacked configurations. This way competitive gamers can opt for more restrictive platforms to play games from, but then don't have to install a rootkit.
> and without them the games simply won't get ported.
The actual effort required to port games these days is relatively small. The major engines like Unity and Unreal Engine make it a breeze to release games across all currently used major platforms.
I don't think you're understanding the argument. Without the anticheat and DRM rootkits the AAA studios won't produce a Linux version because they believe their business models depend on those components.
And this is what makes me crazy about Epic: Unreal is their engine, they've said in the past that they "love Linux", and yet they don't port their games to Linux.
Sounds like they don't love spending multiple person-years on ports + maintenance to expand their userbase by 1-3%. At their scale of operation I can understand it, even if I personally ship Linux game ports. Once they sell a Linux user a copy of <game> for $60 they have to keep supporting it.
Who cares about them running code as root. Modern day malware can do everything from userspace: autorun on startup, key logging, token / credential stealing, cryptomining, software installation, pop up ads, remote desktop / screenshots, recording mic / camera, getting your ip, DOSing a sever, acting as a proxy, replacing sudo with one that steals your password, popping up fake password prompts, etc.
Currently, I use Geforce Now [1] to play games that have unsupported anti-cheat solutions. When I started, they didn't support Linux and you had to fake a ChromeOS user-agent, but nowadays you can simply use it with Chrome/Chromium.
How are the wait times? A year ago I tried the service when I discovered that Cyberpunk overheats my PC, and even with the paid tier, I had to wait in lines for 20-30 mins.
I'm on the founders tier, play occasionally and have only had to wait once, while they had some maintenance in some DC a few weeks ago. I've had 2 complete crashes of the session though, which sucks because save files are synced when the session ends, so i just lost all my progress.
Valve (with the help of kernel and Wine contributors) have been making great progress working with the anti-cheat vendors. BattlEye [1] and Easy Anti-Cheat [2] are already working on Proton.
Each game's developer has to opt in to enable it. There's a community-maintained site tracking the status of each game. [3]
The latest Proton experimental version really improved over this(talking about battleye anticheat).
However the game vendor still has to make a minor change in most cases.Usually nothing starts if the anticheat doesn't approve it, so in most cases right now games that use anti-cheat on steam don't work because of "normal reasons"(i.e not using some special graphics flags, use vulkan/etc) not because the anticheat itself.
90% of my games could run fine on Arch+KDE but there are just some very weird bits that got frustrating, especially with multiplayer games or games where the player community utilize a lot of external utilities (network log parsers for MMOs are one example). This was as of last month.
Also I’m the sort of gamer that has a lot of friends I play games with. We often will play and talk about even a single player game at the same time, kinda like a book club. This is actually kinda common, talking to other gamers. You kinda play what your friends play. It can be FOMO, but more often it’s just really cool to have that shared experience.
So if something drops on EGS and I want to play along with my friends, then I need Windows installed. At the end of the day, if I have to keep a windows install around for even one game, I’m just going to use windows. The end.
Fair enough, many of us on the other hand are repulsed by the Windows telemetry, poor customisation, resource usage, etc so I just keep a second drive to boot into Windows those few hours a week where I want to play a Windows exclusive.
Call me uninformed, but I actually have this article to thank for introducing me to Proton - I have switched to Linux a few months ago, had installed Steam and played a few native Linux games, but so far I was completely unaware that most Windows Steam games can actually run on Linux too. You can blame Steam's very conservative default setting for this - only games that they have tested and certified themselves are allowed to run by default, to try other games you have to select "Enable Steam Play for all other titles" in the Steam settings (a dialog I had never seen before). But now I can happily (re)play most of the games I have amassed in my library over the years (and partly never played) - on Linux!
Yes! I had this same experience. I was in linux + steam for a while before I realized I had to opt into proton-ifying windows games. They should enable this by default and keep providing good UX about it being in play in case you want to disable it.
I recently wanted to play some rounds of de_dust (don't judge) and while the game runs its performance are way worse than when running under windows. I know, small GPU, etc. but still.
If you're getting worse performance on Linux than Windows, it's probably because you don't have hardware acceleration working. You might be using the Mesa drivers (https://mesa3d.org/), which are awesome because they make it possible to run 3D software even if your hardware doesn't support it, but are generally much slower than the official drivers from your GPU vendor (they might even be using software emulation!)
You mean Intel HD 520? Haven't tested that one myself, but Intel usually works out of the box as long as your kernel version is recent enough. It's possible you have the wrong driver loaded though.
To check what driver you're using, install a package called hwinfo (`sudo apt install hwinfo`) then run the following: `hwinfo --gfxcard` and look for the line called "Driver". On my Thinkpad with intel HD4600 graphics, the driver is called "i915", but yours might be slightly different. If it says something else, it's probably a software rasterizer. This wiki page has a list of mesa drivers with descriptions: https://wiki.archlinux.org/title/OpenGL#Installation
You could also use glxinfo, like `glxinfo | grep "OpenGL vendor"`. On my Thinkpad, the output is "Intel Open Source Technology Center"
Here's the output. But how do I know the game is actually using these drivers ? Also, when I tried running with proton I couldn't join any servers (insecure mode) but I suppose it's a VAC thing.
That looks like you have the right driver. It might be a source engine/porting issue then. There's not much that can be done about that, except run it under proton. However, as you said, that probably won't let you play on some servers.
Have you tried running the game's benchmark with and without Proton? Disable vsync and set all the graphics settings to the same level, then see if one performs better than the other. That would at least confirm that the issue is due to a shitty port.
> But how do I know the game is actually using these drivers ?
I'm not expert on the Linux graphics stack, but if you see that in hwinfo, then it probably means the game is using it unless you have some funky switching setup (like Nvidia Optimus). The safest way to check be to read the game's logs, since it might print out that information (similar to the output of glxinfo). There are debugging tools like Renderdoc you can use, but that might get you VAC banned.
Launching Steam through a terminal will show you the standard output of the Steam client, but idk if it also shows you the output of launched games. There's probably a way to do it.
I'll try flatpack during the holidays but it's unlikely I'll fiddle more with it. Until I decide to finally get a musheen with a decent GPU I'll stick to windows for the little bit of gaming I do.
Imagine, after months of work and sleepless nights the day has come, you have to submit your thesis. But during the final corrections your Windows decides to shutdown and update the system for an uncertain duration.
It wasn't my thesis, but I was part of the people that helped to solved the resulting mess (the update took 4 hours and the deadline was 2PM).
That day I learned to never rely on Microsoft ever again.
Kernel updates are not forced like is the case on windows. And yes, linux allows you to update the kernel without rebooting: it's called kernel live patching.
Linux has no mandatory automatic updates in the first place,* so this situation doesn't ever need to take place.
To answer your question, it is possible to update the Linux kernel without rebooting, but you may run into some issues with kernel modules that would be solved with a reboot.
* Edit: After writing this comment, I remembered that Snaps on Linux have automatic updates that can only be deferred up to 90 days. However, Snaps don't handle Linux kernel updates. I don't recommend using Snaps.
> Edit: After writing this comment, I remembered that Snaps on Linux have automatic updates that can only be deferred up to 90 days. However, Snaps don't handle Linux kernel updates. I don't recommend using Snaps.
I didn't even know that. It's one more item on the list of reasons to not use Snaps.
The situation improved after the May 2019 update, but the updates will still be automatically installed after you pause them every 7 days for a maximum of 5 times (35 days total):
Automatically installed, yet no automatic reboots. Maybe it’s because I’ve never used the Home edition, but I’ve never had Windows force a reboot on me. Sometimes my laptop won’t wake up out of sleep mode and it reboots, but that’s about every 2 months.
The last time I was really angry about Windows was with Window Me. The entire TCP stack took a dive which I couldn’t recover from. Even replaced the NIC. Replaced that with Windows 2000 and haven’t had a serious issue since.
Windows 10 pro also did that. Not sure if it still does, because after it happened to me once I waded into the policy editor and disabled all that shite. However, there was at least one forced update that happened to me after that. No idea why.
But windows is also weird. People always talk about ads or having candy crush forcefully reinstalled after updates. I strongly suspect these things vary by region, because those two at least I never noticed.
Staged Windows updates are a pretty common source of behaviors that do require reboots.
Everything from HID devices not responding, task manager and utilities like Settings not opening, device drivers left in an unstable state, and more can happen. To their credit, it does usually fail in a way that allows users to close applications before they're compelled to restart.
I’ve had Windows spontaneously restart on me while I’m doing stuff, to apply an update. If it was somehow my fault, I have no idea what caused it, and so it’s Microsoft’s fault after all since it wasn’t clear.
My computer once decided that 1 A.M. in the night is a great time to automagically boot up and do updates. I had to physically turn off power, so it wasn't able to do that. Even disabling that feature in the bios didn't work for some reason. I still don't know how the machine was able to do it. Fast boot wasn't even enabled, because it caused other problems when it was.
It stopped happening someday, but it's still haunting me. It doesn't really build trust.
On Linux, one can instruct the BIOS to turn on the computer at a specified time after suspend or even power off using rtcwake [1]. I don't know many things about Windows but I guess they are doing something like this.
Yeah, that's what i tried to turn off in the BIOS. It's actually one of the earlier UEFI enabled ones from MSI. I guess it's a bit bugged. It also tends to freeze when i have it open for too long. There are no updates anymore and i don't get a decent CPU for it, so i'll probably upgrade it soon-ish. Just not sure if i'll get MSI again or maybe try something else this time. Asus seems neat.
Since Windows 10. It is possible to disable it, but Microsoft fought very hard to prevent people from doing so by putting up a ton of road blocks. It was a loudly decried 'feature'.
I do not use it on the desktop so I have no idea if that works, I would guess that binary graphics drivers are unpatchable with live patch. I've patched the kernel on servers like that for a long time now, I recommended it warmly no problems so far!
It works the same on desktop as servers. I don't run anything that requires binary graphics drivers though (got far, far away from trying to use Nvidia on Linux many years ago), so I can't speak to that.
There is a program called needrestart which you can run after updates to see what needs restarting to load new copies of libraries etc. Generally speaking you only need to actually reboot for kernel updates.
When you update, you update everything - OS and apps all in one go unless you manually install stuff yourself outside of the package manager. Updates take from seconds to a few minutes unless you are running Gentoo in which case its from minutes to days or even weeks whilst your compiler crunches its way through vast seas of source code and you fix the various issues along the way 8)
You can automate the whole thing or not - up to you.
Grandparent was a facetious question intended to deflate the great-grandparent, which suggested that Windows' annoying need to restart for every damn thing was comparable somehow to Linux's update requirements.
Which as parent points out are functionally nonexistent, except for a rare kernel update (and you don't have to restart if your glasses are thick enough to live patch).
I often suggest Manjaro, which is smart enough to do all the updating for you, but is also Arch enough to let you do it all manually.
Maybe this is new in Windows 11, benefit of the doubt and all, but this isn’t a thing on any other versions of Windows. Forced reboots aren’t a thing. Misclicks, however, are.
Windows 10 will definitely update and reboot without the user taking any action.
One thing to remember is that many streamers are using dual-PC setups and the bigger streamers may only run these systems for streaming -- not for personal use.
It's very possible the streamer missed any update nags because they are mainly interfaced with the gaming PC.
And depending on their setup all the user input might be going to the gaming PC and the streaming PC might think it's currently inactive but with a pesky OBS process left running.
This is a thing on Windows 10, I was watching a video with my sister on the PC running Windows 10 and all of the sudden, it just rebooted into updating Windows.
One thing is that Windows Home does not have an option to postpone or delay updates, this is only for Windows Pro editions. When there is an update, it will start doing it automatically with a timer and it sometimes does not pop up on top of everything you're doing.
Not only for gaming. I have been using Blender for over 15 years on Linux and 3D hardware support is getting better each year. Maybe some people don't like proprietary drivers, but I have been running Nvidia drivers just fine with a RTX card (on Windows you will have proprietary drivers anyway).
And as a Steam user I can only agree with the article. Games are running just fine.
I haven't tried much in the way of very recent games. It is nice to be able to play games I've bought on steam over the last nearly 20 years on underpowered laptops I put linux on. Makes me look back through my library and focus on the older ones instead of more recent or competitive games I'd play on my primary gaming pc.
I switched to Pop OS on my gaming rig and have been enjoying lots and lots of native and compatibility mode games for the past several months. However, FPS-focused anti-cheat isn't yet available (Battleye, EAC), so I'm unable to play my favorite title, Escape from Tarkov.
One big pain point is driver compatibility. Nvidia dropped 32-bit OpenGL, which is required for steam. CUDA packages in Ubuntu sometimes don’t pull in these dependencies, creating “dependency hell” for users.
Been several years since I even considered bringing a windows partition back to play some new game. It's pretty rare that I run into a game I want to play that isn't well-reviewed on protonDB. Albeit, I am only commenting because I went to launch a game just now and am sitting here waiting on the "Processing vulkan shaders" pre-launch step (that only processes shaders once afaik).
Really great to see such renaissance of Linux gaming where progress in Wine/Proton, Vulkan, dxvk and vkd3d gradually unlocked so many games for Linux users despite publishers refusing to address that market. That knight move by Valve which sidesteps that catch 22 blocker was really neat.
Hopefully with growth of the market, more publishers will start releasing native games as well.
As a casual gamer it has everything I need. Steam is "good enough" to satiate that gaming bug that I get every couple of months for something new. I'm still not sure if the top 10% of gamers would find it sufficient though.
Linux support has definitely gotten much better due to Proton, but the "74 of the top 100 most played games on Steam are listed as Gold or Platinum" ignores the fact that 6 out of the top 10 most played have a status of Borked
I don't want anyone to port things to Linux! Sure, when it works it's great, but it's only a matter of time before the executable .sh informs me that it expects a deprecated version of libfucku, or some new kernel thing broke it, or etc.
Instead, target the subset of the win32 API that Proton already kicks ass at. The game has thus been "brought to Linux", and will actually remain playable despite the Linux landscape inevitably changing underfoot.
> before the executable .sh informs me that it expects a deprecated version of
If that's the case, the developers aren't really competent. You can always dynamically link the version you want and pretty much any distro has packages covering a good couple releases back because not all software updates at the same time. This is very normal in Linux distros.
And, even then, the game publisher can (and many do) provide all those libraries together with their title. If you pack a couple gigabytes of art, you can have a couple megabytes of custom binaries.
Porting anything to Linux, let alone doing it the Right Way, is kind of a big ask for a lot of development teams, let alone the publisher who's surely seen the (admittedly improving!) graphs for Linux use among gamers. We can dynamically link things, but I don't want to be up to the elbow in some game's ass for compatibility's sake when a better and more durable option exists, plus less-technical friends literally can't do these kinds of fixes. And this doesn't even address any anticheat the game might require!
I applaud your ideological purity, but when something like Proton exists, wishing for competent devs and motivated end-users -- while dismissing anyone who doesn't meet your standards for either -- seems self-defeating to me. The referenced incompetent devs don't even really need to think in Linux terms or do much extra work -- they just have to watch their elbows and tuck in their shirts while writing the win32 calls they had to write anyway. If you were a dev team, BigCo or indie, shitty or competent, which would you rather do: target two OSes, or one?
> We can dynamically link things, but I don't want to be up to the elbow in some game's ass for compatibility's sake
Most likely it will be the developer. Also, making that effort, to build the app in a resilient way, will give lot of insights about possible bugs in other environments.
> And this doesn't even address any anticheat the game might require!
This is probably why one has to give up control of the game runtime environment. Publishers don’t want gamers looking under the hood, and for good reason.
> I applaud your ideological purity
Quite the contrary. I’m very pragmatic. I get Proton is an easy way to make Windows games work on Linux. My point is that the Windows API is not that amazing and that building games to be portable from the ground up is a better approach from a quality standpoint, even though it may not be the best in financial terms.
I’m very sure the “borked” games and the reason they “bork”, if discovered, will enlighten their developers in interesting ways.
> If that's the case, the developers aren't really competent.
Having worked on delivering portable Linux binaries that had to access OpenGL/Vulkan and audio - yeah, no, it's actually difficult and requires tons of Linux experience. Unless you just want to target some arbitrary subset of 'popular distros', but that's not actually portable.
Shipping a Win32 .exe _is_ the easiest way to get portable Linux games. It will work on _every single distro that has wine_, including weird ones like NixOS or Void or Ubuntu 14. It's an unfortunate fact, but the Linux userland ecosystem is extremely hostile to binaries built outside of a distro's ecosystem.
Send me a tarball of something that draws a Vulkan cube, plays music and responds to user inputs - and I'll eat my hat if it runs on every Linux machine I have in my house.
> Unless you just want to target some arbitrary subset of 'popular distros'
I looked into their website for a breakdown of users per distro, but I’m almost willing to bet that covering SteamOS, Debian Testing, and Fedora Workstation will get you almost everyone indirectly.
I don't think so. Since consoles more and more resemble PC hardware it probably will become easier in the future for any platform.
On console you can develop some things differently, a prominent example is physics and player input depending on frame rate and the renderer. That is perfectly fine since you have a completely controlled environment on consoles. It can solve some problems that are otherwise much more complicated to implement.
But on PC where everyone has different hardware this will cause problems. But lately, also with the variants that console manufacturers release, development for consoles and PCs has become more similar, so porting should be easier as a result.
Also, a Windows/Linux game needs to have a windowed mode, and work at weird resolutions, so you need a GUI library like GTK or Qt. Console games just run fullscreen.
Cheering for this dark horse quite a bit; especially given that Windows 7/10 was essentially KDE, and apparently 11 is basically GNOME, can we go head and put the Microsoft Windows Desktop the burial it deserves?
How dedicated they are to this effort is impressive - they first failed big with the Steam Machines when support was very half baked, but now they'll try again with the Deck just as compatibility is coming along much better.
I guess Linux gamers can largely thank Microsoft for this one!
As a Mac user I'm a bit disappointed they stopped their efforts on that front, since the new processors are quite capable.