is there really an advantage to have native Linux support nowadays?

Yes, performance and consistency.

As a dev, you honestly are asking the wrong question. You shouldn't pride yourself on the corners you're cutting; you should be using them as an opportunity to expand your knowledge and skillset.

You're adding to the pile of developer trash every time you rely on abstraction layers to do your job for you. You're reinforcing an atmosphere where windows developers can get locked-in as much as they want, and it's up to the Linux community to pick up their (your) slack.

In all honesty, I don't have high hopes for your game and you really should consider what I'm saying before you release more trash into the world.

I really appreciate when a dev puts the extra effort in to make a Linux port, but I can understand when a solo dev such as you doesn't have much time to spend on porting. To be honest, I'm just happy if the developers at the very least test the game on Linux using Proton and WINE to make sure it's working well and correctly. That way if making a native port really is that much of a hassle, I expect them to at least test it on Linux. I think that most large teams should make a (good) native port, though.

I cannot go into that site because they don't comply with the gdpr right to not allow any legitimate interest cookies with one button, hoping I will click agree instead of going through hundreds of vendors and disabling these illegal trackers

Honestly, I wouldn't bother.

Linux is a small market and people who buy exclusively native linux games is even smaller. If your game works fine with Proton that's great.

You're a solo dev and should pick your battles. I would focus on getting steam deck verified instead, if possible.

If you yourself use/are familiar with Linux and willing to actually test and polish your Linux version to the same standard as your Windows version, then a native Linux version is always appreciated.

However these days, it's probably not necessary and a lazy afterthought Linux version is like a bad console port, and because we DO have the option to run the Windows version, it's probably worse than no Linux version at all.

So it really depends on your personal feelings towards Linux, and nobody's going to judge you for not providing a native version you can't personally test and support. That's why we have Proton.

Native Linux support is always preferable if the developer can handle it or has the budget to do so. However, as long as it works well via Proton, that's the important part. I don't follow him anymore due to several reasons but Gardiner Bryant has a video about this, it's a pretty solid watch: https://www.youtube.com/watch?v=uScsmjvdwyo

Honestly, I prefer you go with Proton. Some of the native games I have are quite a few versions behind their windows counterparts. AND most likely it will run better anyways. Better for you, better for gamers. Win-win.

I honestly mostly shop for "SteamDeck verified".

I only purchase native titles, because native support means support.

Ideally you should be writing your code to be as portable as possible, in an engine that offers top-notch cross-platform support.

Wat? SteamOS is Linux....

Please publish to GoG as well. I want to support sites that are DRM free. Also I think it's smart to give native Linux support to show the industry that Linux is important rather than rely on compatibility tools that may or may not work well for any given game or setup.

Will you only publish it on steam? It would be great to have games available on other platforms like GoG too.

I find that native clients are almost always a worse experience than running through Proton. Cursors not hiding or not appearing, inability to adjust certain graphics/display settings, and, rarely, worse performance than through Proton (including not actually managing to init).

That said, Proton has its own irritating quirks. Try turning off mouse acceleration in Skyrim on Linux for instance. Have to use protontricks to find the appid, then you have to nav to it in any of the locations that it could be found (depending on flatpak, native, etc.), and then you have to sift through the prefix to find the bloody INI that has the setting to toggle.

But as an indie dev, you should do whatever makes less work for you so you can do more with your time. If supporting native Linux will increase your workload and detract from work on the game itself, then simply don't support it. Just make sure that your Windows builds work via Proton.

EDIT: Here's a specific example from today. For reference, I normally play everything through Proton so I haven't intentionally loaded a native game for like 6 months. But today I loaded up Yooka-Laylee not realizing it had a native version. Get in, alt tab to screencast on Discord aaand... the game nixed my cursor entirely. I mean, it likely just locked it to the game window, but that still means that I couldn't interact with my PC outside the game window. So quickly downloaded the Proton version and it worked just fine after that.

Its about practicality.

Even big studios cant handle supporting a native linux version and having it run well while supporting windows versions.. which usually ends up with people running the windows version via proton over the native version (Looking at you, Cities Skylines, Truck Simulator games, and more)

So I would say take the easy route and develop for windows with checks to make sure it runs okay in proton, probably a lot less headache and effort than trying to create a seperate native linux codebase.

Sigh... this is exactly the sort of developer decision people were worried about when Proton first came out, and why some Linux users were opposed to it. Obviously in retrospect the benefits of Proton have been worth it, but those folks wouldn't be wrong to say "I told you so."

My take on it is that if you're not developing Xbox-first (which would imply being forced into DirectX etc.) you should pick libraries that give you cross-platform compatibility "for free" (I know it's not that simple, but you know what I mean) to begin with. It has to be part of the plan from the beginning (and integrated into your build/test pipeline, etc.); writing the whole game for Windows first and then trying to add Linux support after the fact is a fool's errand.

Get us some Linux exclusives and we can stick it up towards Windows amirite

Id love a native Linux binary, but if you aren't going to maintain it long term after release so something like a glibc update doesn't break it then proton compatibility is probably better. What tends to happen with native games is they get released and sit basically unsupported until they just stop working.

Yes to both. I want games that work well on Linux, so I'm happy with whichever you choose to test. If you provide both, I have options.

Speaking purely for myself: native clients are ALWAYS preferred and appreciated... however as an indie, I would totally understand not wanting to commit to building one.

As a middle ground you could just accept bug reports from Linux users and do what you can to fix those issues. Aim for a platinum rating on Proton DB

eta: Also, yes, as others have pointed out, this assumes the native client is maintained

Yes. It's still needed. Mostly because not everyone using Linux is using a controller or other peripheral input device that just works natively (look at non-steam os handhelds and Bazzite for instance). On smaller screened devices, menus and keyboard specific options can be problematic regardless of other compatibility features.

Choose whatever is best for you.

That being said, as a Linux user I always appreciate a native Linux version of a game that runs well and is updated promptly.

As far as I know, there are game engines that make it easier to publish on many platforms, but I'm no expert.

At this point, Linux native is almost more like reinventing the wheel. The people working on the compatibility layers have done so much work to make a windows game feel native that you can almost think of them as cross compilers in a weird gross way

The days/weeks/months it would take me to add it and fix all the probable bugs it entails could be used to improve the game itself

Making a game multi-platform is improving the game. Massively.

The high level of skill required to write multi-platform games is why most studios don't bother: Windows developers are a dime a dozen but skilled multi-platform developers are rare. Have you got what it takes? Do you have the cojones to step up your programming game? Or are you happy to wallow in the slop with the rest of the Windows game developers?

Also, remember that Linux isn't the only POSIX OS. If you do your porting right, you get to support a shed load of other OSes for very little.