Bridging the Platform Gap: Incentivizing Native Game Support on Steam

[Samuel]

Hey everyone,

I’ve noticed that many games on Steam—often built with engines like Unity, Unreal, and Godot—are available only for Windows. While Proton has made huge strides and now offers a nearly seamless experience, it still feels like there's a missed opportunity. It might be in everyone’s best interest if games were natively available on more platforms.

I propose that Steam incentivize developers to build native versions for additional platforms when they upload their games. For example, Steam could boost advertising for games that support multiple platforms or provide clear data showing developers the percentage of potential players they might be missing out on. Ultimately, expanding native support could benefit both developers, by reaching a larger audience, and Steam, by attracting more users.

What do you all think about this suggestion?

[Ettanin]

https://store.steampowered.com/hwsurvey/Steam-Hardware-Software-Survey-Welcome-to-Steam

Windows 96.10% -1.48%
OSX 1.58% +0.61%
Linux 2.33% +0.88%

There is barely any incentive for developers to target anything but Windows and maybe Linux (for better Steam Deck compatibility). The effort would outweigh the potential additional turnover.

The size of the community outside of Windows doesn't justify the costs porting and patching for Linux and MacOS would incur. There is little to no return on such investment.

[Samuel]

Originalmente postado por Ettanin:

https://store.steampowered.com/hwsurvey/Steam-Hardware-Software-Survey-Welcome-to-Steam

Windows 96.10% -1.48%
OSX 1.58% +0.61%
Linux 2.33% +0.88%

There is barely any incentive for developers to target anything but Windows and maybe Linux (for better Steam Deck compatibility). The effort would outweigh the potential additional turnover.

The size of the community outside of Windows doesn't justify the costs porting and patching for Linux and MacOS would incur. There is little to no return on such investment.

Yes, I understand that Windows remains the dominant platform. However, engines like Unity, Unreal, and Godot inherently support builds for multiple platforms with minimal changes. This isn’t a traditional “port” but rather a native build process built into the engine. Correct me if I’m wrong.

[Ettanin]

Building is one thing, testing the other.

So you have additional testing effort, possibly cross-OS interoperability if the game has multiplayer.

On MacOS it is even worse. To compile on MacOS, you have to own a Mac device. In addition, there are additional fees for the development license and a winding approval process for each patch published.

It's not like clicking one checkbox and bam it can be played on Linux and MacOS.

[HikariLight]

Still would require teams converting the game to work on those systems.
That takes time and money, something that may not be worth it due to the smaller user bases.
If they think cannot recoup the development costs for the conversion, they won't do it.

[BJWyler]

Originalmente postado por Samuel:

What do you all think about this suggestion?

I think it shows that you don't really understand how game development, and software development in general work. Along with understanding how and why Linux will never a "popular" OS until certain things happen and fall into place. It's a nice pipe dream you have, but reality speaks differently.

[Samuel]

I agree, macOS is a whole different ball game—Apple's ecosystem certainly complicates things.

But for Linux, cross-platform issues are largely managed by compatibility layers like Wine or Proton rather than by the developers themselves. While those layers are doing a commendable job, exporting a game natively for Linux would ease their burden considerably.

In most cases, it's as simple as selecting the Linux build target and building it, as outlined in Unity's documentation.

https://docs.unity3d.com/2022.2/Documentation/Manual/Buildsettings-linux.html

[Samuel]

Originalmente postado por BJWyler:

Originalmente postado por Samuel:

What do you all think about this suggestion?

I think it shows that you don't really understand how game development, and software development in general work. Along with understanding how and why Linux will never a "popular" OS until certain things happen and fall into place. It's a nice pipe dream you have, but reality speaks differently.

I've built my perspective on years of experience working with industry-standard engines like Unity and Unreal. In my time as a senior software developer, I've seen that while some modifications may require deliberate effort, they’re well within reach for many indie projects. The process might not be instantaneous, but the workload is generally manageable and can bring worthwhile improvements.

[Ettanin]

Originalmente postado por Samuel:

I agree, macOS is a whole different ball game—Apple's ecosystem certainly complicates things.

But for Linux, cross-platform issues are largely managed by compatibility layers like Wine or Proton rather than by the developers themselves. While those layers are doing a commendable job, exporting a game natively for Linux would ease their burden considerably.

In most cases, it's as simple as selecting the Linux build target and building it, as outlined in Unity's documentation.

https://docs.unity3d.com/2022.2/Documentation/Manual/Buildsettings-linux.html

You'd still have to test under Linux which itself incurs additional effort. Not everything on Linux can be done 1:1 as it can be done under Windows.

Also, you look aside the fact that there's not one version of Linux but several various flavors with their own quirks and incompatibilities. Even the UI and window management systems can differ accross various Linux distributions and in some cases even within the same Linux distribution.

[Samuel]

Originalmente postado por Ettanin:

Originalmente postado por Samuel:

I agree, macOS is a whole different ball game—Apple's ecosystem certainly complicates things.

But for Linux, cross-platform issues are largely managed by compatibility layers like Wine or Proton rather than by the developers themselves. While those layers are doing a commendable job, exporting a game natively for Linux would ease their burden considerably.

In most cases, it's as simple as selecting the Linux build target and building it, as outlined in Unity's documentation.

https://docs.unity3d.com/2022.2/Documentation/Manual/Buildsettings-linux.html

You'd still have to test under Linux which itself incurs additional costs. Not everything on Linux can be done 1:1 as it can be done under Windows.

Also, you look aside the fact that there's not one version of Linux but several various flavors with their own quirks and incompatibilities. Even the UI and window management systems can differ accross various Linux distributions and in some cases even within the same Linux distribution.

There are definitely nuances and challenges with Linux, especially given its diverse distributions and the additional testing overhead. My suggestion isn’t to force developers into native Linux builds, but rather to offer incentives that make it a more attractive option. For many indie projects, for example, building for Linux can be relatively straightforward (as Unity’s documentation suggests), so providing incentives would help cover the extra effort needed to ensure compatibility across the various Linux environments, rather than punishing developers with additional obligations.

[Ettanin]

Originalmente postado por Samuel:

Originalmente postado por Ettanin:

You'd still have to test under Linux which itself incurs additional costs. Not everything on Linux can be done 1:1 as it can be done under Windows.

Also, you look aside the fact that there's not one version of Linux but several various flavors with their own quirks and incompatibilities. Even the UI and window management systems can differ accross various Linux distributions and in some cases even within the same Linux distribution.

There are definitely nuances and challenges with Linux, especially given its diverse distributions and the additional testing overhead. My suggestion isn’t to force developers into native Linux builds, but rather to offer incentives that make it a more attractive option. For many indie projects, for example, building for Linux can be relatively straightforward (as Unity’s documentation suggests), so providing incentives would help cover the extra effort needed to ensure compatibility across the various Linux environments, rather than punishing developers with additional obligations.

The return doesn't justify the effort. And someone has to pay for that effort. There's nothing Valve would gain out of it nor the developer itself.

The incentive for Proton was compatibility with Steam Deck which in turn resulted in more sales for the Steam Deck. Valve will not pay out of their pockets to pay devs for Linux builds if compatibility layers suffice.

[Samuel]

Originalmente postado por Ettanin:

Originalmente postado por Samuel:

There are definitely nuances and challenges with Linux, especially given its diverse distributions and the additional testing overhead. My suggestion isn’t to force developers into native Linux builds, but rather to offer incentives that make it a more attractive option. For many indie projects, for example, building for Linux can be relatively straightforward (as Unity’s documentation suggests), so providing incentives would help cover the extra effort needed to ensure compatibility across the various Linux environments, rather than punishing developers with additional obligations.

The return doesn't justify the effort. And someone has to pay for that effort. There's nothing Valve would gain out of it nor the developer itself.

The incentive for Proton was compatibility with Steam Deck which in turn resulted in more sales for the Steam Deck. Valve will not pay out of their pockets to pay devs for Linux builds if compatibility layers suffice.

There's definitely an argument to be made that easing the load on Proton could be in Valve's best interest.

Valve’s Proton repository on GitHub shows a substantial number of open issues, which indicates that maintaining compatibility across various games and Linux distributions is an ongoing challenge.

https://github.com/ValveSoftware/Proton/issues

Incentivizing developers to provide native Linux builds—for those games where porting is relatively straightforward—could reduce the compatibility overhead that Proton currently bears. While this approach might not be feasible for every game, even a modest reduction in compatibility issues (on the order of 1–2%) would free up resources, leading to a more stable Linux gaming experience and indirectly benefiting Valve by improving platforms like Steam Deck.

[Ettanin]

It is up to the development studios of the games to target Linux, not Steam.

Additional exposure doesn't pay the employer to guarantee warm homes and food on the table for the developers they employ. Someone has to pay for the addtional effort required to make it work under Linux natively. The studios won't do that from their bottom of their hearts alone and Valve has no return on the investment if they pay the devs to do that because the potential buyers are too few to make ends meet. It's financially not viable.

If you are a game developer as you claim to be, why don't you lead by example then?

[Samuel]

Originalmente postado por Ettanin:

It is up to the development studios of the games to target Linux, not Steam.

Additional exposure doesn't pay the employer to guarantee warm homes and food on the table for the developers they employ. Someone has to pay for the addtional effort required to make it work under Linux natively. The studios won't do that from their bottom of their hearts alone and Valve has no return on the investment if they pay the devs to do that because the potential buyers are too few to make ends meet. It's financially not viable.

If you are a game developer as you claim to be, why don't you lead by example then?

I never claimed to be a professional game developer. My experience with Unity, Unreal, and Godot comes purely from hobby projects and school assignments, while my main work is in medical software development.

That said, you’re absolutely right—given the current market conditions, there isn’t enough money in Linux gaming to incentivize developers to invest in native builds. Studios are unlikely to bear the extra cost for what remains a niche market, even if a native build would ease the load on Proton. For now, relying on compatibility layers like Proton is the more economically viable option, though it would be great if the market expanded enough in a few years to change that equation.

Thanks for the constructive discussion tho Ettanin :)

[Start_Running]

Originalmente postado por Samuel:

Hey everyone,

I’ve noticed that many games on Steam—often built with engines like Unity, Unreal, and Godot—are available only for Windows. While Proton has made huge strides and now offers a nearly seamless experience, it still feels like there's a missed opportunity. It might be in everyone’s best interest if games were natively available on more platforms.

That would be nice. Problem is.. Money.

Originalmente postado por Samuel:

I propose that Steam incentivize developers to build native versions for additional platforms when they upload their games. For example, Steam could boost advertising for games that support multiple platforms or provide clear data showing developers the percentage of potential players they might be missing out on. Ultimately, expanding native support could benefit both developers, by reaching a larger audience, and Steam, by attracting more users.

What do you all think about this suggestion?

Developers can already see the percentage of gamers they're missing out on. Combined between linux and Mac...you've got at most generous estimate 5%.

A potential 5% increase in sales isn't worth a certain 50% increase in development and upkeep costs.

[Ettanin]

Originalmente postado por Start_Running:

A potential 5% increase in sales isn't worth a certain 50% increase in development and upkeep costs.

I wouldn't go so far as to say the increase is 50%, more like 15-20% which primarily comes from developing and testing OS specific binaries and dependencies under specific popular OS distros (game assets such as textures are platform independent), but it's still significantly higher than the potential increase in turnover.