646
this post was submitted on 13 Dec 2023
646 points (97.8% liked)
Linux Gaming
15906 readers
51 users here now
Gaming on the GNU/Linux operating system.
Recommended news sources:
Related chat:
Related Communities:
Please be nice to other members. Anyone not being nice will be banned. Keep it fun, respectful and just be awesome to each other.
founded 4 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
I think, people here look at it from the wrong side.
The code changes required for Linux support aren't the issue.
But if they support Linux, they have to support Linux. This is not some student's first indie game, but instead a massive game with up to 290 million monthly active users. That's 3.7% of the whole world's population! (And it's also more than the number of total Linux users.)
So supporting Linux means they need to test on at least all currently maintained versions of maybe the top 20 or so distros on all sorts of hardware configurations. That would increase their testing costs by around a factor of 20.
They also need to support customers if they have problems. Considering the variability of Linux configurations, chances are high that this comparatively small segment of players will consume an aproportional amount of difficult support requests.
And lastly, if the Linux version of the game has some serious bugs on some setup, it might likely be that all these Linux users think the game is shit and start talking badly about it.
So it's just a simple cost calculation: Does Linux support increase or decrease the total profit?
And if the variables change, the calculation changes with it. Exactly as Sweeny said in his post. People like Sweeny don't care about ideals or about which OS they prefer. They only care about money.
And the revelation that a CEO likes money and dislikes risk isn't exactly hard to figure out.
I'm not saying that it's good, but top capitalists tend to be capitalists.
And in the end, I'm pretty sure someone who has all the business figures and frequently has to defend those in front of the shareholders probably knows much better what makes business sense than any of us. Someone like him goes where the money flows.
It absolutely does not mean that.
Pick a steam deck, support a steam deck, 3 major releases. If the SD runs on enterprise Linux that's a 10 year support window.
That's a perfectly viable plan - much like "releasing on x box" - and with an understandable market clearly delineated. Everything else can be "hey try, but don't call us" and we'd all still try.
This is a really good idea--they officially support the steam deck, and that means it's unofficially supported on other Linux distros. The community gets what it wants without a huge extra load on Epic.
Honestly, I'd just test on Steam Deck (performance, recent libs) and Debian (desktop experience, older libs) and that's it.
They also need to fix any exploits they find, which means they probably need Linux devs.
You don't have to support all distros anymore. Just take whatever windows build and test it with Proton.
Sure, but things work differently under the hood on Linux vs Windows, so they still need to validate every build. That means QA resources every release (and they release often), as well as development efforts to patch any Linux-specific exploits.
If supporting Linux doesn't bring in more money than other dev efforts, it's not worth it.
Proton with what? Stable or experimental? DXVK or Wine3D? X11 or Wayland? Nvidia closed source or open source?
That's just what I came up with. There are probably a few more of these questions. Even Proton alone is not an easy target.
Especially if you want some low-level anticheat. And you know, if they have one platform that is easier to cheat, cheaters will all use that platform.
I don't know about you, but playing with tons of cheaters doesn't seem like a lot of fun to me.
You are making it sound way more difficult than it actually is. If some lone indie developer can manage it, a huge corporation with billions of dollars could do it without a thought.
Lone indie devs don't have to care about giving support to players, testing or cheaters.
So sure, if you completely ignore the difficult/expensive parts, the rest is super easy.
Yeah that's just Linux evangelists here, it's not really worth discussing anything with them lol.
Pick any Proton release you like. Support X11, Wayland will then probably work anyway unless people have nvidia gpus. Support DXVK and force people to use it.
If you support one way to run Proton you're already most of the way there. Besides that majority of games just work fine without any work whatsoever.
They should actually target Wayland. Since Wayland will be what supports HDR, VRR, and is what the Steam Deck and most distros use.
Seems like you didn't read my first comment that you replied to before.
But still, your view is totally fine for a little indie studio, but that doesn't work for a game with >200 mio players.
Lol, nvidia...
Most games that work on Steam Deck aren't technically Linux-compatible and therefore have no "Linux support" needed. Proton has come very very far, and most games are running the Windows exe through Steam using Proton.
In fact, I've played several games that do have native Linux support, and they still play better using the Windows version through Proton. On my Steam Deck, and on my shitty non-gaming laptop.
So no, they don't have to support anything new.
Exactly. Making the game WINE-compatible is not the same thing as committing to support. In reality, the only thing stopping WINE from working is Epic Anti-Cheat and the absurd thing about this is that Epic already gave EAC a WINE-compatibility mode -- they're just actively choosing not to turn it on.
What Tim's really saying is this:
I've tried running fortnite on Linux . It installed fine started to play and then I get booted out because of the anti cheat . I believe the game would run fine if the anti cheat supported Linux .
That's the fun part, the anti cheat does support Linux. Apex Legends also uses easyanticheat and it's compatible with the Steam Deck, with a Gold rating on protondb.
Yeah. Fortnite's implementation of EAC literally sabotages itself on Linux mainly because Tim Sweeney doesn't want it to run on Linux
That's even more annoying because the game will run . It just boots you out before you touch the ground. I was thinking about playing it again lately but I can't build with controller like I can with a keyboard and mouse.
Just release it as an unsupported "beta". The Linux users will figure out any issues and give higher quality bug reports.
The only thing stopping Fortnite from running on Linux is the anticheat. The anticheat it uses it made by Epic, and has a specific option for WINE compatibility.
If I remember correctly it actually uses two separate anti-cheat, and the second one not made by Epic doesn't have Linux or Wine support.
But it's still a weak excuse that they could just make a Linux version without that redundant second anti-cheat.
Unless they changed it, the other one is Battle eye which also has Linux support
I'm going to do a hard disagree here - they don't have to support Linux, just add compatibility in terms of anti-cheat for Linux. Proton is likely good enough to run the game itself but the anti-cheat sees Linux and just craps itself.
They don't even have to provide support - League of Legends runs on Linux if you install the game using community scripts and custom proton, and while the client runs poorly nobody spams the Riot Games support about how the "Linux version" client doesn't work the well because people understand that it's a community effort. Riot themselves have only made a statement saying how they'll try not to break the game for Linux users, and that's pretty much it.
League of Legends is a massively popular game as well, yet Riot barely has to do anything to maintain it on Linux, let community fix issues that come up, let community provide support as it's their tools.
And while I do understand that porting an anti-cheat to be more friendly to another operating system isn't an easy task (such as for Rust, where they tried to make the anti-cheat compatible with Linux but it introduced other issues so it got shelved), I think you're vastly overstating the amount of areas a company has to cover for a game to be playable on Linux.
There's a difference though.
If the game doesn't work for (some or all) Linux users, that's not a big problem from Epic's POV. They'll lose a couple users that wouldn't have been able to play the game without Linux support anyway.
But if the Anticheat faills on Linux, that is a completely different story. Then cheaters would all dual boot over to Linux to cheat all they want. That's now a problem for the whole game's user base and consequently for the publisher as well.
Something as low-level as an Anticheat would have to be rewritten almost from scratch to work on Linux and this one really needs to be tested with every possible permutation of installed relevant software. Because if one combination is found where it doesn't work, you can be sure that the day after every cheater will be running this config.
(Just to check, do you have a background in game development and/or low-level Windows/Linux programming? I got all of that and I can tell you, nothing that looks easy from the outside is actually easy. I think you are vastly underestimating how much work goes into something until it "just works as expected")
Speaking as a former game cheater...
Cheaters are going to cheat. Booting into Linux isn't going to change that.
Anti-cheats just keep the filthy casuals from cheating. A broken anti-cheat on Linux would be fixed pretty quickly.
Sure, but that's dev resources they need to spend on a small market, and they'd suggest need to hire Linux devs or pull from other projects. It's quite likely the math just doesn't add up given the likelihood for profit for other uses of those resources.
I doubt Epic would lose money in it, but they probably wouldn't make as much as other options.
EAC has a check box for Proton compatibility. Battleeye is linux native. All they have to do is check a box, and test to see if they can break it. If they let it out in the works and there's some influx of cheaters, they can check the box again. Halo Infinite, Apex Legends, Smite, Battlebit etc etc were all capable of checking the box and testing.
I suspect Sweenys hesitation over support is caused by a lack of control.
Upgrading EAC in an unreal engine game is trivial, it's basically baked into the engine. They update EAC all the time.
Dude, steam ships with a bunch of libraries enabling cross distro support. It ain't that complicated https://gitlab.steamos.cloud/steamrt/steam-runtime-tools/-/blob/main/docs/container-runtime.md
To be honest.... Yes it's that complicated. I've read that, Apparently valve had to spent massive ressource to figure out the load order of librairies and what to include for the steam runtime.
Granted, all they made is open source iirc. But it was a massive pita
Yes, their first attempt used load order overrides and search patch patching. Now, it uses linux containers to ship an isolated environment. Think of it as more similar to docker (or LXC/LXD). That said, I haven't used it myself to so cannot comment on how difficult it is to use. Most people here are advocating for them permitting proton use without necessarily supporting it officially though. Which can easily be done by changing an option in EAC.
Did you read the second line of my post?
The code changes aren't the issue.
Did you read my comment? They ship with libraries to unify distribution across distros
I said: Code changes are easy, all the other things in regards to supporting playing on Linux (anticheat, support requests, testing, ...) is hard.
You said: But code changes are easy because steam has libraries to unify distribution.
Do you see the problem here?
What are you going to tell me next? That code changes are easy?
It's more that epic have a competing store and Tim doesn't want to do anything that might help steam gain more traction.
If it can be made to run via Steam, then they only need to support it as far as getting it installed in Steam. Either Proton or native, it can be made an invisible issue from the user perspective. They have made a choice not to do so.
EA/Respawn somehow haven't had a problem with doing that with Apex legends.
Apparently, their cost calculation is different. Also, Fortnite has about 50x active users compared to Apex Legends. That also changes a lot.
Sweeny said it doesn't make business sense for them and if it will make sense in the future, they will support Linux.
I'm pretty sure that someone who does know their business figures and frequently has to justify them to shareholders has a better overview about what makes business sense for them than anyone of us.
Every time someone makes the business argument all I can think of Microsoft flopping with Windows Phone despite all their money. Google failing with Stadia and losing opportunity they had with hangouts to imessage. LG bowing out of smartphones. Blackberry and Nokia too late to enter smartphones despite prior dominance. Epic was so late into trying their hand at digital distribution until 2018 when doing it earlier over the past decade would have made entry easier.
Companies just because they have money doesn't mean they know what they are doing. And sometimes even less than random people.
Well, if half a million people are guessing on a choice of two options, some are going to get it right. But that's not due to the insight of the people, but due to numbers.
These examples really don't apply here.
Also, there are some other factors in play that you didn't consider.
Smartphones exploded between 2007 and 2010. It went from nothing to almost everything in just a few years, and those who got lucky and where ready at the right time managed to take the new market. Windows Mobile proves that it's not enough to be super early. You need the right timing in both directions.
There is no indication that Linux will have >50% market share among gamers within the next 3 years. Yes, it nudged Linux over the 3% mark but at that rate it's going to take a long while. Also, contrary to smartphones vs feature phones, the steam deck is an additional gaming PC for on the go. It doesn't replace desktop gaming.
Also, when it comes to mobile gaming, the Steam Deck is a distant fourth between Android, iOS and the Switch.
And even if you limit the scope to x86 mobile gaming, they are by far not the only competitor. There are lots of others, many of them using Windows, who do the same.
And the biggest edge the Steam Deck is it's value, because Steam subsidizes the Deck with their Store sales. Most people don't care whether it runs Linux or not.
Does it have to support every permutation or just a standard one like SteamOS?
It feels like none of the replies to you actually read your comment. I appreciate you taking the time to offer up possible explanations with examples. Thank you!
Except he's completely wrong because containers work specifically to solve the problem of deploying across different configurations. Valve already figured this out a decade ago with the steam runtime. That's why I can run a relatively obscure OS like Bazzite and nearly my entire library of AAA just works like it would on any other distro. You can run a container across hundreds of thousands of different configurations, it doesn't matter.
Yeah, pretty much all answers are "You are wrong, the code change is easy".
Kinda sad that people don't make it even to the first line.
lol why are you simping for them? they made a choice not to do this. they could easily do it with their manpower if they didn't, you know, keep laying people off in order to maximize profits. You're also overinflating how difficult it is to make games cross-platform compatible with the tools available today.
It sucks a lot when people are so deep in their petty trench fights over brands that they think there is only "Me for this, You for that, You simp".
I don't care about Epic and neither do I care for Steam. I buy my games where I get them the cheapest: Key resellers. And I don't care on which online store the cheapest price lands.
If I was still developing games, I'd deploy them on both or on the one who pays me the most for an exclusivity deal.
With that out of the way: I am only explaining simple backgrounds to people interested to listen.
But sadly so many people fight over an online shop as if it was politics.
Do you fight like that for your favourite online retailer? Or your favourite supermarket chain?
What Steam and Epic do is business. They are no charities. They do stuff that makes them money. So any sane user should see it as a business transaction and buy where the price is best for what you get.
LOL I have no skin in this game. Your comment is pure projection and I think that you have demonstrated precisely what I was arguing.