Why make games for Linux if they don't sell? | Hacker News Reader

Why make games for Linux if they don't sell?

theregister.com

62 pointspjmlp71 comments

I don't expect to play any big AAA titles but I also remember only having Nethack and tux racer. So I am over the moon being able to play these current titles on Linux.

Caves of Qud, Dead Cells, Hollow Knight, Rimworld, Oxygen Not Included, Stellaris, Civ, Streets of Rogue, Pillars of Eternity and many more. Those are just my favorites so I hope others will list theirs.

Hey don't forget XBill, Linux game of the year 1994

Rings a faint bell. But one game I loved that I didn't mention was Konquest. It came with KDE on RedHat 6 or 7.

It was a turn based space strategy game. I absolutely loved it and spent a large part of my adult life looking for a modern alternative.

Galcon 2 seems to be a real-time alternative ; not sure if it would match your expectations.

There was also a tank game, and a star trek game where you configure your ship and fight with it (and if you configured max cloaking, you regained energy while cloaked). Good times.

tank game = warzone 2100?

Do you know that wojak meme where a crying guy is wearing a happy mask? That's how I felt playing tux racer in the mid 2000s

Nethack on the other hand...

Porting games to Linux is definitely not something you do for the money. Linux sales on Steam hover in the 1%-2% area of total sales, less than even MacOS.

On the plus side, if you use an engine like Unity, making a binary for Linux is no longer much of a problem (you don't have to wrangle xplatform libraries etc). On the minus side, actually testing and debugging things to make sure things work for customers who use a variety of distros and hardware is hard, especially since Steam tends to only officially support a single (usually outdated) version of Ubuntu at a time.

Larger teams will have the resources for this. But where it really hits hard, is the small indie team made up of Windows devs (as game devs tend to be), when the port is not completely turnkey (like, say, porting a Unity game to MacOS) and requires debugging in a completely foreign environment, when you know you might not get sales to pay for this work, and you're already under deadlines to fix a million other things.

>> Porting games to Linux is definitely not something you do for the money.

Part of the article suggested that building Linux first is easier, because porting to Windows should mostly just work.

And Mac ! If you're doing a Mac port you may as well do a Linux one too.

I regularly go through my Steam wishlist and remove any game which doesn't support Linux. I am probably the only one, but it is something I do.

Some do play with Proton though.

Check out the db https://www.protondb.com/ I have some games playing with proton and there are no issues at all.

You don't consider if they run using Proton?

Ultimately I want to support devs who support Linux, but if it works on Proton doesn't it also get the flag on steam for support of SteamOS/Linux. I am only going by the Steam flag.

It doesn't get marked for every game that is playable. They have a list of games that are officially supported by Proton but there are so many more that work well that you can play if you enable the option in the Steam settings

No, proton compatibility isn't really signified by the flag as it's not official support. That being said it seems Valve doesn't do much to help promote Proton compatibility either, as the only good resource I know to check compatibility is protondb.com (which you should take a look at by the way).

Same here. I don't know why but I do. I know windows software will still basically work after about 10 years, especially now. But I still feel scarred from earlier steam sales where devs just dropped the ball and let their games rot to be only compatible on Windows 7 or jesus even XP.

There's very little money for the most part in updating old games. The same is true with a lot of old books, among other things.

Stadia changes the math a bit. If you’re already targeting Stadia, which uses a Linux variant under the hood, the time to make it widely available might be worth it.

Might be... but (I assume; I don't have any specific knowledge) that Stadia runs everything on a single known software stack, and quite possibly a single hardware stack. This would make it an easier platform to target than the wider range of general consumer hardware running Linux.

Personally, I run Linux as my daily driver and I hope game developers see the success of Steam's Linux support and decide that targeting Linux is viable and worth doing. But I wouldn't assume that building for Stadia necessarily has a big overlap with building for consumer Linux.

Valve seem to have been addressing that with their runtimes.

See https://github.com/ValveSoftware/steam-runtime

Consider that Stadia is not the only cloud gaming platform. Also, even if the overlap between Stadia and consumer Linux is small, it's still probably much bigger than the overlap between Windows and consumer Linux

When making a game for Stadia you can target a very small subset of hardware and test on those on one single linux distribution/software stack (at the moment all stadia servers are identical). When you make a game that anyone can download on their machine the amount of different hardware + os/driver combinations grows a lot and requires a lot more effort for a quality product.

The success of Stadia means more developers who are comfortable with Vulkan and linux as a platform. That should give us better ports and even more Vulkan first titles.

That could lead to better driver support (at least from Nvidia :| ) on 'nix, and so on and so forth.

And is it successful?

Success of Stadia? It's barely Presence of Stadia at this stage.

Meant is more hypothetically, not as if it is a current fact.

When someone makes a game for Stadia they don't really care if it works on linux or Vulkan in general. All they care is that the game works on the custom hardware and drivers Google/AMD provides for Stadia. That is the only platform they will test that version on. (This is very similar to how console development happens. You have a fixed set of hardware so you just develop for that without a care for anything else really)

Sure the distance is smaller then starting with the Windows or Xbox/Playstation versions but there is still a lot of work to get it to work on any random pc someone builds.

"All they care is that the game works on the custom hardware and drivers Google/AMD provides for Stadia."

I understand, and agree. Exposure to Vulkan, Linux, and the related optimizations is helpful and I personally considering it a bonus and assume it will have downstream benefits (maybe im delusional) for gaming.

They are already comfortable with it thanks to Android NDK APIs, and you don't see those games coming to GNU/Linux.

If a game is on Stadia it can already be played on Linux via Stadia and a web browser. It can also be played on macOS. I thought that's where you were going with that.

They are saying that if a game runs on Stadia then the vendor must have already done the work to port it to a Linux-based platform, since Stadia is a Linux-based platform.

Android is a Linux-based platform. The hard part of dealing with Linux is that the userland-is-a-free-for-all model means there are effectively hundreds of Linux-based platforms out there to deal with.

That is true, but even the most niche Linux platform is still closer to the mean of Linux platforms than Windows is.

Nobody actually uses Stadia, though?

Because if you don't then you can be sure that they will never sell.

And Linux is the only game platform without garden walls.

Windows still lets you download and install software last I checked.

Does MS allow one to sell games for both XBox and Windows?

For Xbox you must use the Xbox Store. Why are you changing the goalpost though? They only mentioned Windows.

Which allows you to just plainly sell games from your own website or use one of the many stores.

Oh side note, you can change your Xbox into a "devkit" and install basically any homebrew app. So technically, you can bypass its walled garden.

One addition, you can sell on all stores at the same time.

Yes. But you don't need "allowance" on Windows.

You'd need one for the Xbox store or the MSFT store. But so far, you can sell something both through the MSFT store, with Xbox branding or not, and outside of the MSFT store.

For now. My guess is that one of the reasons Valve has been working so hard on the Linux gaming front is the fear of this changing.

Yes, Valve steered towards Linux right at the time Microsoft made some noise about privileging its store, when they were bent on mimicking Apple in all things. MS seem to have come to their senses later on, but by then Valve had built too much stuff to throw it all away.

This should really be a case study: don’t piss off your customers and developers before you have all your ducks in a row. Something that clearly others have not learnt, e.g. RedHat with their CentOS fumble a few weeks ago.

You're saying they may take an Apple approach and attempt to further scare people from downloading trusted .exe's and such?

I think Microsoft has been having a bit of a change in attitude just because they're more focused on making money in software. They don't care where you buy it so much as they get a chunk of it somewhere. But what do I know.

Windows 10X original design, which got delayed due to COVID, uses sandboxing for whole userspace applications, regardless if they are UWP or Win32 based.

https://www.youtube.com/watch?v=ztrmrIlgbIc

Not everything. Some things, Windows Defender will erroneously and (somewhat) silently remove, causing inexplicable errors or crashes while hiding its actual errors in a notifications center. It gets really annoying sometimes after doing updates of actually desired software.

If the software is new or obscure then it'll be blocked by Defender unless the user knows the override ritual. Code signing may help.

Feels silly comparison...if someone can't dont that you can rule out them running Linux.

Sorry if I was unclear. The code signing requirement to bypass the ritual means Windows is more of a walled garden than Linux.

> And Linux is the only game platform without garden walls.

What do you consider a garden wall? Anyone can create and publish a Windows binary without permission from anyone. It might even be possible to do this from Linux, using 100% GNU tools, though I'm not entirely sure about that.

Such binaries will create a warning if they aren't signed, but from a user's perspective that isn't materially different from requiring a "chmod +x".

Over the last several years, I've stopped buying games that aren't on Linux. It's just too much of a hassle to be switching around when there's plenty of Linux-supported fun to be had.

I think there's a certain chicken egg aspect to this problem as well. Devs don't target Linux because people don't game on Linux and people don't game on Linux because there's no games. (Obviously this isn't the whole story there are other reasons people don't use Linux and i admit this loop is probably relatively minor)

No the real problem is that almost nobody use Linux desktop computers, and the ones that do don’t want to pay for games. It would be crazy for a game developer, who has to spend a lot of $ on art etc. to focus on Linux.

These two options can work well:

1. The Radiohead model: download and pay anything from 0$ to X$ as you wish. IIRC with the Radiohead release X == 100$.

2. Sell Merchandise. If your game features a character like Lara or Mario, this could go very well.

There used to be a lot more of this in the early days of indy development. Postcardware for instance, were you would "pay" for the software by sending the author a nice postcard.

Feels like an admission that there is no formal market.

Radiohead was not an "indie" until they were already huge from several traditional label-distributed albums. They're more of an Activision.

They don't sell? Eh? Says who? The article offers no backup for this claim. It's just plain wrong.

Have we already forgotten how when HumbleBundle first launched, Linux accounted for 20% of its sales? (then worth only a 1% "marketshare"). The other 20% being mac (4% marketshare) and 60% windows.

Heh, dunno, seems quite a few folks claim minimal sales for linux. I have noticed the finding linux support for humble bundle games is MUCH less than it used to be.

Steam seems to have pretty hard numbers and the linux number are tiny.

A number of Indie game developers have published their sales numbers. All show close to no sales on Linux.

How are stadia games packaged and shipped? Would it be possible for anyone to build a linux box to match the stadia hardware/drivers, and thereby make them run a stadia game?

If that's the case, then perhaps that's the way to go for anyone who wants to build a Linux AAA-gaming box.

200-300 people is enough? Surprised it's that low even for a lone dev

Lone dev in low-cost country (Romania) and seemingly churning out a title every few months. Say 5 titles per year at 30 euros with 200 customers is 30.000 EUR per year, or 2500 per month - a very respectable salary even in countries like Italy or France. Throw in a couple of freelance contracts and you’re doing pretty well.

The only limit, really, is that it works only as long as there are very few competitors.

In patreon bux, that could keep you going for sure. I think people often underestimate how much of a game changer it's been for indie creators.

I really don't understand why anyone would waste the energy to target Linux beyond making sure that your Windows build also works on Wine.

Did you read the article?

Yes. Fine, if you want to do it as a labor of love, but on an economic basis, it's moronic.

Even when that's done it never seems to be officially supported. Probably because too much is out of their hands to make any guarantees.

You simply cannot offer support for Linux users because everything above the kernel is optional. Even if you limit yourself to a few distributions they have a bad habit of changing fundamental userland components every 2 years. Hell, Ubuntu nearly dropped 32bit lib support and consequently their status as the only Linux officially supported by Steam.

I believe this argument is slightly weak. I understand that the ecosystem is very diverse but even on Windows most new games only officially support Windows 10. If you're on 7 for any reason then it's your problem if the game doesn't work. Similarly if you're willing to go down the Linux route developing and deploying only on Ubuntu, for instance, can be regarded as a "single platform". It's my prerogative to use whatever I want on my Linux box but then it's my problem to make sure I have all the required components available for my bespoke setup. I haven't used Ubuntu in years but give me a "Ubuntu executable" and I'll find a way to make it work on my desktop. And if you don't want to limit yourself on Ubuntu Valve is now rolling out their "steam runtime containers" which are reproducible environments for developers to target.

There are still issues with the drivers, I'll give you that. But at least AMD/Mesa is moving fast to adopt features. NVIDIA is going to be NVIDIA but at least they are not skimping on implementing OpenGL and Vulkan extensions.

Reply on news.ycombinator.com