#StandWithUkraine

Russian Aggression Must Stop


Headed for another dark age

2021/07/28

Tags: linux gaming

This post is complete speculation and guesswork, but I want to put these predictions out there so that I can eventually play my "told you so" card.

My prediction is that Linux gaming is headed for another dark age to the likes of the post-Loki era of the 2000s. Within the span of the next 5 or so years I predict that Linux gaming will be in a state someone would label "dead", or at least more so than now.

Since the announcement of Valve's Steam Deck and their developer-facing marketing turning its focus almost entirely on Proton, we have now seen some giants of the Linux gaming ecosystem essentially give up. Ethan Lee is thinking about retirement, Feral Interactive has stopped porting efforts citing lack of profitability due to Proton and who knows what the average indie dev's views on the topic are. It's also worth noting that these big players didn't just make ports of video games, but also created libraries, frameworks, tools and driver contributions to improve the overall ecosystem beyond just individual games. Various developers targeting native Linux with their games also creates demand on game engine and tooling companies to make their offerings compatible with Linux.

So, the current gamble with a focus on Proton has the following potential outcomes:

  1. Proton-first approach gives Steam Deck and the accompanying ecosystem a big early boost in market share, creating an incentive to target the platform natively.
  2. Steam Deck gains market share, but Proton is established as the de facto deployment target. Companies actively test their games to ensure compatibility. Native development and ecosystem tooling suffers due to being a second-class citizen.
  3. Steam Deck gains some market share, but is deemed largely compatible with Windows and as such requires no special attention from developers. Developers will develop and test their products on Windows. Linux gaming ecosystem is essentially at the mercy of Microsoft developers.
  4. Steam Deck gains some but ultimately insignificant portion of the market share. Valve's interest wanes and developer resources are moved towards more exciting projects. Proton development stagnates, native development efforts die down until the cycle begins anew.

I think ultimately the signs show that Steam Deck will end up gaining market share, but I am not really convinced that it's going to suddenly revolutionize the Linux gaming market. I also don't think Valve actually has any big ideas or plans, despite how much some people might want to believe that they have some sort of an ingenious plan in place. I believe Valve to be fickle and unfocused and their commitments will end as soon as they find something more interesting to work on, unless what they are doing right now massively pays off.

I believe people developing games on Linux will be the first to hurt in all of this. Playing games is one thing and I don't doubt that often you can do that just about as well on Proton as natively, sometimes likely even better. But if Linux is no longer a valuable target, that means maintaining Linux support in tools like Unity and Unreal Engine becomes costly bloat and will be left to stagnate until it is removed as technical debt. And if there's no need to deploy to Linux from these tools, running them on Linux also becomes a potentially unwanted feature. After all, if Proton runs Windows games so well, surely those silly people can also run the developer tooling through Proton as well? And so what if the software breaks Proton compatibility every now and then, it's not like you are working on something valuable, right? If you were, you'd use a supported platform.

I've always believed that for Linux gaming to succeed, we need developers to use it. That starts out with them installing Linux on some of their testing machines to make sure their builds work, getting them acquainted with the Linux community through interaction and learning from bug reports. Then developers see value in using Linux, which results in developers spending more time working on Linux and potentially even doing development work there. This then drives demand for developer tooling that works on the platform, which makes Linux more appealing to more and more developers. Some of the developers also end up making improvements to the ecosystem and the operating system as a whole, which continues to feed the cycle. Hell, some developers will probably also learn about the virtues of free and open source software and thus contribute either their games or tools for others to iterate upon.

Proton (or rather, Windows) becoming the main target for deployment works to halt the feedback loop. It sets Linux up as a simple consumption platform that is compatible with existing processes. It makes native deployment targets redundant and threatens the existence of native development efforts. It erodes the platform's autonomy and centralizes efforts to the likes of Microsoft and Valve.

I predict we will feel the sting of that within the next five years. I might be wrong, Icculus seems to believe that the Wine strategy will pay off in the end for example and he's far more in tune with the history and goings-on of the Linux gaming ecosystem than I am. But, if it turns out that I am right, I will absolutely rub it in everyone's faces, and at that point we can only hope that the open source side of the ecosystem has grown strong enough thanks to projects like Godot Engine so that developing games and potentially other software natively on Linux won't be a massive pain.

>> Home