Check what actual developers say about it. Controlled or not, they simply make a Linux port. Amount of testing is limited for sure, but still. Same port can be used for running on regular Linux, granted they test more.

Porting costs are going into enabling Vulkan, after they were stuck with DX12. Going forward they might be smarter and will have Vulkan support from the beginning.
It doesn't matter what they are interested in it for. Install base is what matters and Valve estimates demand in millions for it. I bet it will be way higher than Stadia.

Next marketing step for Valve is to nudge the likes of CDPR to release native games for it.

No, they don't. Did you not read what I said? I didn't say that CDPR didn't invest in Stadia support. I said: Desktop Linux is not the same as Stadia. Graphics and audio are not the only issues (although they do consume the bulk of the processing time for translation layers). Any developer who says that developing to the Stadia API is the same as developing for Linux is certain to make a broken Linux port. If you think testing is the only thing that needs to be done, you're clearly never written any cross-platform software.

Are you a member of the CDPR management team? Do you know why they chose to support Stadia? I seriously doubt it was due to the vast number of users. Marketing/PR buzz (lack of desktop Linux support does not generate any significant social media response, even if that lack of support comes after explicit/broken promises), financial guarantees/incentives from Google, potential for unlimited income due to being a subscription service (admittedly I'm not sure if the games being played get a cut from the subscription fee, though, since you have to pay up front for games as well), and other reasons are far more likely.

And, as multiple game developers have already said, proton support for their games is sufficient (sometimes implicitly, such as Larian saying they made changes to BG3 to make it more compatible with proton). The number of desktop Linux installations is completely irrelevant. Steam Deck users are even less relevant, since all most of them will end up caring about is whether or not games run sufficiently well on their machine (which proton ensures it will, to the most part). Who cares if it's broken on other audio hardware, other video hardware, other controllers, other software in general, or doesn't follow XDG guidelines? Yeah, a very small subset does care. That's not the numbers you're preaching, though.

I'll tell you what: I'll agree that a Linux port is on the way once I see Galaxy for Linux. I don't want it, and won't use it, but it (or at least a subset of its features) is required by at least the gog ports for CDPR games. It's also something that was actually promised, explicitly, and in writing. All of your other speculation is just pointless wishful thinking.

I trust actual developers who work on this more. And they contradict what you said directly. Unless it's a game using Stadia specific functionality (which CP2077 is not), it's simply a Linux port. That's about it. I'm not even interested in debating this, it's a waste of time.

Sufficient or not, Proton will never be as performant as native release. Developers of the size of CDPR not caring about performance? I guess possible, but it's pretty poor taste.

I think what he meant is that a Linux port will most likely work on Stadia, but a Statia port (i.e. optimized and only tested on) will probably be a broken Linux port. It even happens on Windows when some developers only optimize and test their game for NVidia (or AMD) card and it end up being totally broken on anything else.

The "controlled" part is what makes all the differences, it doesn't matter if the code is 90% the same, having a controlled environment means you can test, debug, optimize for a specific hardware, for a specific OS and totally ignore the rest. And it gives you very little support.

That can make the difference between a viable platform and a non-viable one.

No, the only thing that matter is Linux gaming market share, it doesn't matter if Valve sells hundreds billions of Steam decks, if only a fraction of them are used for Linux gaming it's not going to do anything.


To go back to proton the ironic thing in all this is that Proton was originally created by Valve to "protect" themselves against Microsoft in case they decided to lock down Windows.... except Microsoft doesn't give a damn, they only care about Azure and cloud.

But on the other side, thanks to Proton it makes it a lot easier and cheaper for Google or other game streaming companies to offer games as they can do it on cheaper Linux infrastructure.

So by trying to counter Microsoft, Valve ended up helping streaming which is probably a much bigger thread against them than Microsoft ever will.

I wouldn't be surprised if in a, hopefully, distant future all new games will be running exclusively on Linux... except they will be only available via streaming.

So, uh.
How's that Stadia workin' out, anyway?

Here's the thing. Unless you can convince every telecom to switch to fibre optics for every last kilometer and square meter, and deliver at those speeds, "Cloud Gaming" isn't going to work.

Especially on most people's crappy Wifi setups. Plus, cloud gaming ignores the history of gaming as a general rule of thumb; so if you want to play something retro like Akalabeth, who the heck is going to offer an instance for you?

Don't get me wrong I absolutely hate Stadia and game streaming in general and nothing makes more happy than a streaming service failing.

However I do think that it is the biggest threat toward gaming, there is a reason why despite OnLive and other similar services total failure, new services keeps popping up and most of the big players, either in gaming or cloud infrastructure, starts proposing their own streaming services.

Luckily for now between the stupid pricing scheme and the lacking Internet infrastructure in some country limit its growth, but I am very afraid that it would last that long and that soon it will be "good enough" for a lot of peoples.

I’m always baffled when I see Linux users rejecting Microsoft’s walled garden to enclose themselves willingly in Valve’s golden cage… Valve does not care *at all* for native Linux gaming. All they want is to be independent from Microsoft, Linux support is only a side-effect.

If you follow their marketing, all is about Proton™ (a.k.a. rebranded WINE). They advise game developers to test their games with Proton, not to work on native builds.

Please let me know who those developers are, so I can take that into consideration if I see "Linux support". not that plenty of native ports aren't broken anyway, in some way or another, but no need to ask for it. I have read developer commentary as well, by people porting games from Linux to Stadia (whom I trust more than Windows devs making statements about Linux). Note that I said "ported". It's not a zero effort operation. It may be easy (especially if you actually know what you're doing), but it's not, as you repeatedly state, free, not even "but for some extra testing and resulting tweaks".

While you're at it, ask them how opening windows, getting display resolution and size, picking the right monitor, displaying full-screen, picking the right GPU, getting input from keyboard and mouse, and getting input from controllers (not to mention vibration and LEDs) works the same on desktop Linux and a system with no X, only one GPU and virtual monitor, and no Linux input subsystem. I'd say the same about shared audio and multi-card support with ALSA, which so many games get wrong anyway, but google states pulsaudio is their audio interface, exclusively (which makes me wonder if such games would even work on my pa-less system, where I only use apulse). If the answer is "using a portability library" (of which Google only advertises support for unity3d, although apparently some have more recently ported SDL as well), then what stopped them from using that portability library to port to Linux to begin with? Why do those reasons no longer apply? What forces other developers to use that compatibility library instead of just coding to the "native" stadia API? You can make a portable game that works on both, but if you just port the game to Stadia, it won' t just work on desktop Linux, unless you add the portability shims at the same time. And that's not even getting into the functionality provided by Steam/Google/Galaxy for server-side functionality like public time tracking and achievements. Testing might help with optimization for other hardware, but there are still other things that need to be implemented, and implemented well (these things seem trivial to me in my own projects, but enough games bungle the basics, e.g. games that crash if my controller is so much as plugged in, that I must be wrong).

Some difference, not all the difference. The bulk of the cost goes into implementing Vulkan path itself. Testing and the rest is multitude times easier than that.

It's still some work of course, but once Vulkan path is there, the effort to test and make sure it works on more hardware is way more affordable than the effort of implementing the Vulkan renderer in the first place. So if you bring any arguments about cost, controlled environment is the least of your concerns.

Even Stadia developers themselves point out, that they develop it as a general Linux version in mind, so I doubt controlled environment plays a big role for it besides just saving them time on testing because they care only about one Vulkan driver.

That's exactly what market share is - install base. Nothing else matters. That's why it's impossible to ignore Windows gaming. Because MS achieved near monopoly of pre-installed OSes on the desktop.

So if Valve will have millions of devices with per-installed Linux and especially targeted for gaming, that's a huge thing.

Here is one example: https://news.ycombinator.com/item?id=25370070

I bet they have some fork of SDL or some plugin for it, that handles Stadia-style remote input and such. So it abstracts it for them same as it abstracts it for the desktop use case.

Google could be more cooperative and publish it or even develop it in the open. But it's Google. Their approach to collaborating with the rest of the Linux ecosystem is always half baked.

Valve dodged the question about it. Basically, they want to least resistance approach. I.e. attract more developers with Proton. If that works, I assume they can start showing how native versions improve performance and let developers decide whether it's worth the cost to improve quality.

That doesn't sound like a cage since they aren't stopping anyone from selling their Linux games everywhere. I.e. if some developers make it for Steam Deck, they can as well sell it on GOG too, like already happens with Linux games that come out DRM-free.

Something like Stadia on the other hand had almost no reverse effect so far. I.e. we didn't see CP2077 for Linux coming out.

We already have more than millions of devices running Linux and able to play games. Namely Android phones.
I am not under the impression it helped gaming on Linux overall…

To be fair I do not see any excitement about this new gaming console outside of a couple self-centred communities hyping themselves, like GOG forums, GamingOnLinux, or linux_gaming subreddit. Both the wider Linux and gaming communities do not seem to care about it at all.

---

Developers can not chose *not* to sell their game on Steam if they want to see a meaningful amount of sales.
Players who bought their games from Valve will have to keep Steam around forever or lose access to their games.

It is a cage for both developers and players. The only people getting a benefit from Steam existence are Valve themselves.

Android created a huge rift by using a whole incompatible libc and graphics stack. So I don't consider Android as part of desktop Linux ecosystem. At least it's good that in contrast Valve work with upstream, including Mesa, Wayland and so on.
I guess we'd need to wait and see the numbers after it comes out. But so far it's clear Valve did put some weight behind marketing for it. So if it pays off - developers will have more interest.

I mean that has been like that without Steam Deck too. I.e. there are tons of Steam only games unfortunately. So adding Steam Deck doesn't change that. But what it does change is more incentives to make Linux games. If some of them will end up on GOG - I think it's still a plus.

I think this is the root of our disagreement ;)

In my opinion no "positive" side-effect can justify feeding a de facto monopoly, so I am never going to support Valve/Steam, like I am not supporting Google or Microsoft. If it means a slower trickle of Linux games on GOG (I don’t think it will), then so be it.

Anyway, it seems a lot of players running Linux see Steam as good for Linux but they are biased by being Steam users themselves. It is true that Valve did a lot of things to allow playing Steam games on non-Microsoft systems, but if you are not using Steam at all you get almost no benefit from their existence, while the drawbacks are obvious (DRM everywhere, store exclusivities, "2nd class citizen" treatment, etc.).

isnt linux a dead platform? I can hardly see anybody using it
you are right , android is no way linux , especially as win11 supports android apps better than linux desktops