Here’s another look at the survey results as collected back in April 2021. After looking at cloud gaming trends, hardware choice between AMD and Nvidia, and the usage of different game stores, we will use this time to focus on how respondents assessed different predictions we published back in February 2021 about what would happen by the end of the year in the world of Linux gaming.

Let’s first jump at the results, to explain what respondents believed… or did not.

Now a few comments about the predictions themselves and the results from the survey.

1. Valve will introduce a new version of SteamOS / new hardware using Linux.

Most respondents did not believe this was going to happen. Only one out of four thought this was somewhat likely or very likely for SteamOS, and only about 18% for the hardware part. Turns out that this small minority guessed right, as it’s now very clear that Valve will achieve both by the end of the year following the introduction of the Steam Deck and its companion SteamOS 3.0.

Verdict: (LIKELY) REALIZED (both of them) as long as Valve does not screw up with the release window.

2. There will be less and less native ports because of Proton.

While we will have to have a careful look at actual numbers, this definitely feels like it is happening, and even Feral has recently tweeted that they would not port Total War TROY because of “less demand for native titles since Valve’s launch of Proton.” A small majority (57%) of respondents expected this would be the case. I am actually very surprised by the low percentage of “very likely” answers here, since it should be fairly obvious that the incentive to develop native ports is going down and Proton can do the job for you from day one.

Verdict: (LIKELY) REALIZED.

3. Steam will come to ChromeOS and lead the growth of Linux gaming

Here it’s polarized, with a small majority thinking this would not happen. Well, we do know that ChromeOS is now very likely to support Steam in an official fashion very soon (probably in 2021 while we are still waiting for a formal announcement). Yet it’s not clear if such a move would lead to a major and visible growth of the Linux market share on Steam. It might, but Chromebooks are not known for their strong hardware capabilities - so the amount of serious gaming you would be able to do on such devices will be limited to smaller, older games, or low footprint indie titles. But who knows, there could be a massive casual gaming audience waiting for a Steam client on such devices. We will have to wait for the end of the year to make a definite call on this prediction.

Note that there are also hints that Chromebooks with discrete AMD GPU are coming on the market soon. But such units will take a long time to penetrate the market so for the time being, most of ChromeOS gaming should be on devices with limited capabilities.

Verdict: UNDECIDED.

4. Proton’s compatibility will be more visible in Steam

Initially, that statement was related to having something like ProtonDB ratings (either community-driven or Valve-driven) to indicate the Proton compatibility directly in the Steam client. Most respondents in the survey expected this to occur this year (81%). Turns out Valve’s answer may look like “full compatibility of the Steam catalogue” following the Steam Deck reveal. Now, we are far from reaching that stage, at least with the current version of Proton in early August 2021 (6.3-5). Valve may be working on a different version in parallel, as well as specific EAC support for several titles. Whether or not the compatibility will be realized as promised is still a question mark. Garry Newman, dev of the Rust game, confirmed on Twitter he got EAC working through Proton. It might not be long before other games receive EAC compatibility as well.

Verdict: UNDECIDED, while it’s headed this way.

5. Nvidia will fully support Wayland in 2021

Most respondents (65%) expected that would be the case, and they were right! Nvidia has made huge steps forward to supporting Wayland in official capacity with the latest 470.x drivers. These new drivers make it possible to render directly with hardware acceleration to XWayland in a Wayland session.

Verdict: REALIZED.

6. Linux will gain market share on Steam in 2021

So, you may have seen the news recently that Linux has reached 1% in the July Steam survey despite being around 0.8~0.9% for a very long time. It’s almost symbolic as it comes following a ton of recent good news packed into a very short time. I’m not yet fully convinced this is a meaningful metric or simply a fluke from the data (the survey data is not completely reliable from one month to the next). In any case, most respondents expect this to come true (72%), and it’s just a matter of waiting a little to see if the 1% market share is a trend or simply a temporary outlier. The Steam Deck should also have an effect on this, but since it will only be shipping in limited quantities at the end of the year, its impact probably will not be felt by the end of 2021.

Verdict: UNDECIDED, need a few more months of data.

7. Epic Games Store will support Linux in 2021.

One of the most outlandish predictions if any, since Tim Sweeney has never shown much love or interest towards Linux. We do know that Epic Games has granted some money to Lutris a while back (which supports installing the Epic Games Store through WINE), but there are no other signs since then. With the advent of the Steam Deck, running SteamOS by default, there’s now more chance than ever for a proper Epic Games Store client to exist for Linux. Tim Sweeney has praised the Steam Deck for its openness, referring to the fact that it will be possible to install Windows and then access the Epic Games Store on it… so that’s not very encouraging.

If it comes, it might not be within 2021… and if Epic persists in supporting mostly Windows, well… we will probably know by the end of the year anyway what they intend to do in the future.

Verdict: UNDECIDED, and does not look very likely.

8. EAC will be supported by Proton in 2021.

Only 36% of respondents believed that this would be the case, and Valve has surprised everyone by announcing they were working directly with EAC to make it work under Proton (for all games? Or specific games? We lack details at this stage). While we are starving for more information, it looks like it’s headed in the right direction.

Verdict: (LIKELY TO BE) REALIZED.

Next Steps

I thought initially that respondents would act as the “wisdom of crowds” when it comes to assessing the possibility of predictions. Turns out that the “wisdom of crowds” is a mixed bag in this particular case:

• Most respodents did not expect SteamOS/Valve hardware
• Most respondents did not expect EAC support

They did get right:

• Nvidia supports Wayland
• Less native ports (a shy majority)
• Potentially a higher market share of Linux on Steam

It’s always easy in hindsight to change one’s mind and rationalize everything and forget about their poor understanding of the future, but the fact remains that most of us had a blindspot when it came to Valve’s actions going forward.

If you think in terms of incentives, most of the predictions listed in there actually make sense:

• Valve has no incentive to spend years of efforts to develop Proton, were it not for releasing their own hardware using Linux, to avoid the strict dependency on Windows.
• Less and less native ports for Linux: there is no incentive for such ports to exist when a zero-work porting solution like Proton works in about 60% to 70% of cases. It’s not perfect but certainly a lot better than devs having to understand how to support Linux and keep a port alive year after year.
• Google markets Linux machines as Chromebooks and could benefit from the gaming aspect to further grow its penetration on the market. Therefore having Steam supported on ChromeOS makes total sense.
• Valve has an interest in making Proton’s compatibility more obvious to end users, to avoid dissastisfaction, refunds, or unwanted support tickets. It can be ignored for Linux gamers that are less than 1% of the market, but not if they launch a new device that will sell millions.
• Nvidia has no choice but supporting Wayland: X11 is deprecated (while still maintained), and big distros now ship with Wayland by default. The Linux market is small for Nvidia, but they also have to show that they lead in the industry across all platforms they support. Which is also why they would support DLSS and ray tracing on Linux too.
• Linux gaining market share: would be a natural consequence of all of the above, as long as the competition (Windows or Mac) does not drastically change.
• Epic Games supporting Linux: that’s definitely the weakest incentive, but Epic is apparently pissed off by the proprietary aspects of several platforms (such as iOS) and having a Linux client (along with Proton support) would go a long way into safeguarding their future against a possible threat of Windows progressively turning into a less open platform.
• Finally, Valve would need to crack the EAC nut for Proton to even mean anything. There’s no point in having Proton if the highest profile games using EAC cannot run at all in multiplayer: this would push away a lot of potential users who would revert back to Windows before one can say “Deck!”.

That’s it for now! Let’s reconnect a few months down the road and see how much has changed, hopefully for the better.