Is Linux support still needed nowadays, or is Steam OS good enough?
Is Linux support still needed nowadays, or is Steam OS good enough?
Discover topics like pc gaming, linux, video games, indie game, and the magic of the internet at Imgur, a community powered entertainment destination. Lift your spirits with funny jokes, trending memes, entertaining gifs, inspiring stories, viral videos, and so much more from users like unchartedsec...
Hey there!
I'm a solo dev working on a sci-fi grand strategy game (I didn't manage to find if self-promo is allowed so I'll keep the name for myself).
I was updating my planning and started to think: since my game will be published on Steam, it will be playable on Linux using Compatibility Mode even if I don't specifically target Linux itself. I myself play on an Ubuntu and this allows me to play almost every Windows game (old ones are more capricious, but recent ones are ok).
So I'm wondering, is there really an advantage to have native Linux support nowadays? As a solo dev, the thing I lack the most is time. 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 or add features instead for example.
On a more general note, what do you other Linux players expect from a Linux game?
Steam Deck and Linux desktop user here.
I wouldn't prioritize a native Linux binary if I were you. For some of my games that have a native Linux version, I still run the Proton compatibility version instead because it runs better. Proton is insanely well-optimized and adapted to Linux desktops, your own code will have a hard time competing with that.
Just make sure to pick an engine / tech stack that works well with Proton and find one or more testers who run SteamOS or another Linux distribution.
As a personal anecdote, recently I installed a co-op videogame on my Linux Steam machine and I couldn't get past the main menu, I wasted quite a bit of my own and my friend's time before realizing it was a bug in the Linux build. After reinstalling the Windows version through Proton everything worked flawlessly.
Please don't publish a Linux build unless you plan to test and maintain it.
I see this with things on Itch.io a lot. I think there's some engine that has an option to export a Linux distribution of the game, so they ship that without testing.
Yup. Usually not as well optimized either, if you do get it to work.
Yeah I agree, it's honestly impressive how well it works. As soon as I have the possibility I'll take a Steam Deck and try to make my game compatible with it, I'm sure if I manage to make it work on it that should eliminate 90% of Proton compatibility-related bugs if I don't make a native build.
I'm still seeking for answers from people for whom it doesn't work well or who have bad experiences with it, because I'm sure there are cases where it doesn't work
I had some bad experiences with wine/proton, especially on older hardware. There are sometimes weird bugs that are very hard to reproduce. For example: I couldnt play starfield for about a year because most textures were missing. No idea what was causing it. After a year it suddenly worked. Generally if you check Protondb for any game you will find that a few people have crazy bugs, even if everyone else says that it runs perfectly. I think I dont have a single game in my library that has a "Platinum" compatibility rating.
I think it boils down to how would you prefer to support linux players: By fixing bugs in your port, or chasing after bugs in wine/proton. The negatives of native ports that other people have raised are because many (larger) developers make the initial linux port, and then forget about it. Civilisation 6 for example has a good linux port. Yay! But they stopped updating it years ago. Many bugs were never fixed, and linux players dont even have access to the newest dlc! Compare it to running Civ 5 in wine. Things mostly work, exept mod support. Most mods just dont work. Who knows why.
I think native ports may have become easier nowadays too, because of the steam runtime. Steam can run linux games in their flatpak-like runtime, so ideally you only have to support that.
If you are planning on open sourcing your game (maybe some time after the commercial release) then a linux port would be very appeciated for packaging reasons. Other than that, your preference.
Half of the games with a native build I've tried didn't even launch. With Proton they all work perfectly.
I prefer a working Proton build over a untested broken native build. If you can't afford the time to test the Linux builds, go ahead and just rely on Proton.
Testing on Steam Deck so it can be verified probably does make sense from a marketing standpoint though. Mostly checking that controls work, graphic settings are not too demanding by default and font scaling works properly.
I think it's what I'll do indeed. And for the Steam Deck I intend to make the game fully compatible with it as I've read a lot of time how it enabled people with otherwise few opportunities to play to get a bit of "me time". Being a dad I empathize a lot with that struggle haha
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.
I would focus on getting steam deck verified instead, if possible.
One billion times this. That's a checkmark that increases your marketshare by a lot. Desktop Linux users (not all, but some) look to that as a good indicator on whether it'll run for them because it's even harder to make it run well on the steam deck. Kind of a "two birds one stone" thing
I was going to do it anyway, but your answers gave me even more confidence in how important it is! Thank you :)
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.
I use Bazzite, which is literally designed for gaming (pretty much steamOS), and more often than not, the Windows version with proton works better than the native Linux runtime.
It's not the OS, it's the game.
Agreed. Targeting handhelds with good controller support and also no DRM (or at the least, Proton friendly) is more important to me as a consumer than a native Linux version.
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'm using Ubuntu myself to develop actually so I'm kind of doing that all the time. The problem is, my machine isn't everyone's machine, my drivers aren't anyone's drivers, and so on.
In the end I think that I'll try to give Linux builds to testers to see if they report a lot of bugs or not and decide at that moment if it's too much work or not.
Sounds like a good strategy. I don't know all the details, but from the way I understand it, your Linux build would be run by Steam in a Linux container, Steam Runtime. So build and test against that if you can. If I have it right, maybe the Steam Runtime container will give you more confidence if you are able to go with a Linux build.
edit: If you sell your game through Steam.
Linux native games are great as long as the engine is open source. If it's closed source, it will eventually stop running on up-to-date distros some time after the developer stops updating it. For closed source games, it would be better to develop for excellent wine/proton compatibility so the game will continue to be playable long after it stops getting updates.
Good point indeed. Thanks :)
Closed-source games should be packaged including their dependencies (e.g. like an AppImage or whatever) rather than relying on OS-provided shared libraries.
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 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.
*Lin-lin
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
Thank you for the link, I didn't know the inner workings of it :)
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.
some native titles are absolute ass and it leaves room for people to even disable wine support because you already have native (even though native sucks ass)
Example? CS2.
Recently installed Total Warhammer 3 which I didn't notice had a native Linux client. Multiplayer didn't work. Textures never loaded in. Loading times ass.
Switched to Proton and everything worked flawlessly.
Many times I have enabled compatibility mode in Steam to run the Windows version, which actually works.
If I have a problem I don't even mention I use Wine/Proton. Many games and software devs just tell you that only Windows is supported and move on.
Well, that's what I'm doing, I'm not writing for Windows specifically, and I'm using an engine that works well with any OS. But there is a difference between making the build and supporting it, it's an actual commitment. Like others said here it's a bad experience to try to use a native build and have it being bad or worse than a Win build that goes through Proton.
In the end I think I'll make builds and try to find testers for every OS, and try to estimate the viability of having a specific build using their feedback.
I honestly mostly shop for "SteamDeck verified".
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.
I've heard that Steam provides some guidelines to get your games to work well with Proton. I'm not able to find them but maybe they're behind a developer portal or something. If it works with Proton there's no reason to aim for a native Linux binary since your time will be better spent elsewhere.
Yes that seems to be the consensus in the end. Thanks for your answer :)
For the docs I've found this for Proton common issues: https://partner.steamgames.com/doc/steamdeck/proton?l=english And this for Steam Deck compatibility: https://partner.steamgames.com/doc/steamdeck/recommendations I'm sure there are other resources, but if someone finds the thread later and needs a starting point, here they are
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.
I personally am fine with Proton, the native Linux build usually doesn't work as well as the Windows one because devs don't usually care because of smaller market share (or don't have the time as is often the case with indie studios).
What's the name of the game? I'll add it to my wishlist if it sounds like something I'd enjoy.
Also, if I may ask, do a GOG release, I do 90% of my gaming there.
Proton it will probably be then :) The game is Uncharted Sectors, I don't have a Steam page public yet, but you can subscribe to the mailing list if you are interested, subscribers get access to the beta (https://uncharted-sectors.com/)
I'm not that familiar with GOG, I'll take a look at it. Does it have a similar Linux compatibility layer? Because if not I do need to make a native build then
Will you only publish it on steam? It would be great to have games available on other platforms like GoG too.
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.
Well, I'm using Godot which can target all three big OS, though clicking on the export button and actually supporting a platform is not the same.
In the end I think I'll try to make builds for all OSes and have testers for each platform, and estimate the viability of the builds from the amount of bugs they give me back.
In the end I think I'll try to make builds for all OSes and have testers for each platform, and estimate the viability of the builds from the amount of bugs they give me back.
Don't forget to consider the possibility that getting more bug reports from Linux users doesn't necessarily mean the Linux version is more buggy, but instead that Linux users may be better at finding bugs or more willing to report them.
This is a good take... You can tell when newer (usually indie) games were built on OS-agnostic engines; you can't really tell the difference between the Linux runtime and proton.
Otherwise, proton is often better.
Native is always better, but Proton (what's used by steam OS to run windows games) has worked for me for every title I've tried (even some small indie titles)... Sometimes graphics settings can be inconsistent, but I'd say it's a pretty safe bet if you're using an established engine it will just work with Proton
That is my experience too, I guess some Linux builds are just poorly optimized in comparison to the Windows ones.
Valve's engineers did a wonderful job with Proton honestly
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.
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.
Wat? SteamOS is Linux....
Yes my bad haha I had trouble to find another way to write that title which wasn't too long but would mention Steam
I think just calling it Proton would work, most people here would understand that I'd think.
I think some anticheats make a distinction.
You mean so that they can allow SteamOS users but block anyone using any other Linux distro?
I've never heard of that but so far any game with anti-cheat that works on the deck works on Linux
Get us some Linux exclusives and we can stick it up towards Windows amirite
Haha problem is I have a kid to feed, and I fear that cutting myself from 90% of my potential player might be a bad play if I want to be able to continue feeding him
Should Vale release a general Steam OS and folk move away from Windows en masse it's probably going to be worth it. Proton does a great job but a tested and supported Linux build will be better. Some solos do it, but I think it is more based in ideology rather than necessity.
Your time and effort as a solo is going to be limited though, testing and supporting a Linux build is going to be work, and if you want to keep things closed source you are not going to be able to leverage the community in the way open software does.
There might be steps you can take to be more compatible with Proton and/or Steam Deck, which probably would be a good idea, and give you much more effect for your time spend.
Nowadays I view Windows as a gaming layer that has been bolted on the great Borg mothership that is Linux in the form of Wine or Proton in route to the goal of Total World Domination.
A bit like Embrace, Extend, Extinguish but in reverse, if you will. So it's fine.
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.
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
I think you should just write it for Linux. Since Windows has that Linux Windows Thing (I don't remember what it's called), it'll probably work there as well as long as you mark it as a dependency.
if a bug arises from using proton, it would cause the same issue for users that are forced to use proton.
just click the button that builds for linux. build for as many platforms as you can. if you cant fix all of the bugs due to time, then you'll become a better programmer for your next game and not make thise same mistakes <3
Well, it's not a quick game made in 3 months and then I go on to the next. I fully intend to support it and give my players the best experience I can. So it's not as simple as "click the export button", I have to actually commit to support the builds.
I think the only big benefits that come from having a native Linux port is performance, and compatability (in case running through wine / proton didn't work.) Other than that I can't think of anything else.