The second part in my investigation into how well developers are doing on Linux with selling their games via the Steam store.
I hope you all enjoyed the first part; it was a good experience talking to developers and seeing their numbers, so I decided to reach out to more developers for your reading entertainment!
They are listed from best to worse performing titles, so be sure to show your love for the lower ones.
Harvest: Massive Encouter
For the last 12 months
77.1% Windows
5.4% Mac OSX
17.5% Linux
If we count the whole lifetime though;
96.2% Windows
2.3% Mac OSX
1.5% Linux
Just bear in mind it was released in 2008, the Linux version only got released at the end of 2012, so that 1.5% figure is actually really, really good.
Jens Bergensten, Oxeye Game StudioUnless Steam for Linux really takes off in the future, we're not sure it's the effort for us to offer our games on multiple platforms. It's very time consuming for such a small team as Oxeye has.
Kairo
75% Windows
16.2% Mac OSX
8.8% Linux
This is after Linux had a delayed released.
Door Kickers
89.0% Windows
8.0% Mac OSX
3.0% Linux (Ubuntu distributions account for 57% of the Linux users)
Mihai Gosa, KillHouse GamesThe stats were different in our first month since the Early Access release, where Linux was around 6%, it seems things have changed.
We also released a free demo for almost two months. That was before getting on Steam, but Linux downloads totalled at 5%.
The initial development time for porting on Linux was about 2 work days and another 3 days added in the following months for updates/fixes.
All the code written for the Linux port is contained within a single file of 1320 lines.
Most problems we had on Linux were compatibility issues, which were solved by using an older Ubuntu version for compiling the game. We also never managed to do proper alt-tab when the game is running full-screen, which seems to bring major anger to our users :(
Mac OSX initially took longer at about 3 days, mostly due to the programming IDE (Xcode) and the time it took to go buy a Mac machine, but didn’t have to do any fixes on it afterwards.
Since they’re both POSIX systems, parts of the code are common.
Even though it depends on factors like what technology/engine you’re using (we’re using a home-brewed engine) and how you architectured your code in the first place, I can’t imagine why more developers don’t do it.
Bottom line is Linux+Mac brought us 11% of the sales for a single week’s work, so yeah, it was definitely worth it.
Trine 2
Units sold:
2.1% Mac OSX
1.9% Linux
Revenue:
4.9% Mac OSX
4.2% Linux
Since 01.01.2013
Awesomenauts
91.46% Windows
6.70% Mac OSX
1.84% Linux
Sales % per platform for the last 4 months.
Maia
The Windows figure is over inflated since many people redeeming Kickstarter keys were forced to use Windows for several weeks after Steam launch. Looking at the player figures or sales this month, the actual number is roughly 8% with Mac at 5% or so.
97.5% Windows
1.4% Linux
1.1% Mac OSX
Simon Roth, Maia DeveloperLinux was one of my target platforms from conception. I'm far more keen on supporting it than OSX, which has been nothing but trouble in comparison.
Since I use standard, Linux friendly libraries such as SDL, it's really a no-brainer to support. Using different compilers on multiple platforms is a good way to weed out complex memory bugs so even if it wasn't profitable to support, it would be worthwhile to me. Interestingly, due to the new GCC features and the low system overhead on Linux, Maia Linux often outperforms the other platforms in testing. It's also far easier to support currently, as Linux users are generally more technically proficient with their systems.
I primarily develop and test on Mint, as I prefer the "Windows" style GUI rather than Ubuntu's horrid new interface. I think Linux adoption was hit hard by Canonical totally losing the plot since around their 8.10 release.
For future games I will definitely support Linux, and am considering a controller support for the Steam box. I'd really like to see ATI put out some better drivers as the NVIDIA ones are currently a mile ahead.
Rust
The developers where kind enough to show us that image when I requested it, so that makes:
95.88% Windows
3.57% Mac OSX
0.55% Linux
Again this number may seem small, but I doubt the guys at Facepunch Studios think it's not worth it (For those who don't know, Facepunch Studios created Garrys Mod) . When you look at it that 5,621 * £14.99 selling price is around £84,258 give or take, that's without taking taxes and Steam's own share into account, but that's still a lot of money, probably more than some indie games will ever make.
Project Zomboid
99.942% Windows
0.05% Mac OSX
0.008% Linux
% of total revenue made since PZ appeared on Steam.
Looks like it's time to show Project Zomboid some sales love don't you?
Chris Simpson (known as Lemmy) from Project Zomboid's The Indie Stone had this to say:
Chris Simpson, The Indie StoneI remember though coming up with a figure of approx 3-5% linux back in the day we offered direct downloads, though obviously this is a much lower sample of people. In truth its not really easy for us to tell definitively, but for the large part if the cross platform support isn't too heavy, then its beside the point.
On the whole the low numbers don't bother us as we have a lot of pride supporting all three platforms, but more crucially we firmly believe in time those numbers are going to swing dramatically in Linux's direction with SteamOS, even if it takes a few years. We worked on a game using XNA, which is MS proprietary technology, and while MonoGame exists now, we were stung badly by MS's reckless abandon of it. And with Windows 8 and MacOS both taking worrying steps toward forced application signing and funnelling more and more through their own stores, it's not about making money but securing the future freedom of our game not to by tied into a closed system governed by a company who could just screw you on a whim. That's why Linux is and should be of prime importance to developers, particularly those doing alpha funding business models where they have to rely on income from a single game for years.
OH I should add, and this is of prime importance further to our numbers. We don't feel we've done a good job, even an acceptable job, at supporting Linux thus far, due to the problems we've had getting familiar with it as developers. As such we've probably lost a huge chunk of sales from Linux and to a lesser extent mac due to our poor support of them. We have a demo of the game we urge people to try before buying, and due to the nature of Linux and our inexperience we've yet to hit on a way to make our game 'just work' and this may have turned a good % of people on Linux off buying our game. we're improving this and hope to come up with a way to make PZ more accessible to Linux users without manual set up and such, so the numbers may look a lot more flattering to Linux at that stage.
Again a big, big, thank you to all the developers who could spare time; we know they are always stupidly busy and not always doing fun things. Paperwork takes a lot of time to do!
What do you all think of Part 2? Are you surprised or do you think it's still about right for where Linux is in the market right now?
Some you may have missed, popular articles from the last month:
The comments on this article are closed.
I usually pay for my games from my windows desktop, but tend to buy linux games that run on my notebook. Can we see Usage stats ? how many people actually play on the various OS's ?
0 Likes
I don't think people understand. Microsoft has fought tooth and nail to convince developers their platform should be the only platform to develop for. They have backed it up by pushing kick-backs (Money) to developers to not only develop solely for Windows but to also publicly crucify Open Source.
In using Windows, you become part of the problem. You inadvertently tell developers you want Windows only Software and Games. By buying and playing Games on Windows, you contribute to Microsoft's success on it's war with Linux, Open Source and Multi-platform.
DirectX, Basic, XNA, crippled OpenGL, back-channel funding, SCO; when is it going to sink in people? When it's too late?.
In using Windows, you become part of the problem. You inadvertently tell developers you want Windows only Software and Games. By buying and playing Games on Windows, you contribute to Microsoft's success on it's war with Linux, Open Source and Multi-platform.
DirectX, Basic, XNA, crippled OpenGL, back-channel funding, SCO; when is it going to sink in people? When it's too late?.
0 Likes
Quoting: Mike FrettIn using Windows, you become part of the problem. You inadvertently tell developers you want Windows only Software and Games. By buying and playing Games on Windows, you contribute to Microsoft's success on it's war with Linux, Open Source and Multi-platform.
While I mostly agree with what you say, I do think that you are underestimating the peer pressure to use Windows that people encounter every day. You buy a computer that comes pre-installed with Windows, you go to school where you are required to use Windows, you graduate and get a job where you are required to use Windows... Blaming people for something that they believe they have no control over, is not going to solve anything. Presenting them with a compelling alternative will.
0 Likes
Admin can you please research "How Steam counts percentage of people using Linux machine"?
0 Likes
PZ had been on Desura for a long time since Alpha and became very well known, I bought it then and I guess also many other people did, Steam sales might be not that relevant.
Some developers saying it took just a few days for them to port a game and fixing bugs means they're doing things right in the technical side, platform agnostic = future-proof. They also think is worthy not only because they earn money but because they gain freedom, any company not concerned about this will loose big opportunities. I'd say those are bright minds that will lead the way for others.
Some developers saying it took just a few days for them to port a game and fixing bugs means they're doing things right in the technical side, platform agnostic = future-proof. They also think is worthy not only because they earn money but because they gain freedom, any company not concerned about this will loose big opportunities. I'd say those are bright minds that will lead the way for others.
0 Likes
Quoting: DrMcCoyHow exactly are they counting the platform of Steam users?
Initial download? Time the user spents running Steam on a platform? What if someone downloaded using Wine first? What if they didn't download the game at all?
This is actually an excellent follow up question to this interview series. Do the developers count platforms on their own, does Valve count it for them, how does this whole counting business work?
0 Likes
Quoting: FutureSutureI find it odd how Oxeye Game Studio is considering ditching Linux when Linux has been selling three times as many units as Mac OS over the past year and has almost caught up to Mac OS in overall sales as well. Might as well ditch Mac OS too, no?
one would have to guess they went into multi platform THE wrong way, or at least they learned nothing from it. if game porting will cost more than your sales say you can expect from marginal platforms, there is no point in port. it is better to simply focus on next project, this time with having pre-process where you focus on making decision what to use and why. one thing is almost sure, if your game sold to 5% of windows population, you can expect 5% from other populations as well (in translation, if your game sucks too much to pay for it for most people on windows, it sucks the same for most people on other platforms too)
develop for windows and then port is simply not how you start the project and based on their comment this is exactly what they did. and to make it worse, their game is not selling in millions in the beginning, which makes it really dubious on what they were expecting.
select tools that work everywhere, then start windows development is how one lowers cost and maximizes profit. at the point where your windows release works, you'll be day or two away from having it run everywhere if you choose right path. and those 2 days will cover profit from all marginal platforms which will probably make it most profitable 2 days in whole time frame.
0 Likes
Quoting: aaaaaaaaone would have to guess they went into multi platform THE wrong way, or at least they learned nothing from it.
I feel bad for developers that fall for the propaganda of how they should develop and what tools they have to use. You're right that if you don't build with multiple platforms in mind the porting isn't worth it.
I know a guy who has been trying to complete a game for the past 10 years (multiple games have been attempted). He makes all the wrong decisions about the tools to use. He also won't touch Linux "until" it gets native DirectX support. He seems to think that Microsoft is in the process of open sourcing all that code and that Linux is going to adopt it.
So the real problem stems from the knowledge and research that a developer does before they start development. There are so many tools out there that you don't have to know anything to start making a game. Microsoft shifted the tools in use back when they released DX9 and the Khronos group failed to advance OpenGL fast enough. That stagnation period from 2.1 to the 3.x series is what pushed OpenGL to the back burner of development. Just look at games created prior to DX9, most either used OpenGL exclusively or had the option to switch between OpenGL or D3D. The saving grace is that OpenGL is now advancing faster than DX, it's not tied to an OS, and mobile devices are requiring developers to learn OpenGL (at least if they don't use an abstraction tool Unity3D).
0 Likes
Quoting: migiziQuoting: aaaaaaaaone would have to guess they went into multi platform THE wrong way, or at least they learned nothing from it.I feel bad for developers that fall for the propaganda of how they should develop and what tools they have to use. You're right that if you don't build with multiple platforms in mind the porting isn't worth it.
I know a guy who has been trying to complete a game for the past 10 years (multiple games have been attempted). He makes all the wrong decisions about the tools to use. He also won't touch Linux "until" it gets native DirectX support. He seems to think that Microsoft is in the process of open sourcing all that code and that Linux is going to adopt it.
So the real problem stems from the knowledge and research that a developer does before they start development. There are so many tools out there that you don't have to know anything to start making a game. Microsoft shifted the tools in use back when they released DX9 and the Khronos group failed to advance OpenGL fast enough. That stagnation period from 2.1 to the 3.x series is what pushed OpenGL to the back burner of development. Just look at games created prior to DX9, most either used OpenGL exclusively or had the option to switch between OpenGL or D3D. The saving grace is that OpenGL is now advancing faster than DX, it's not tied to an OS, and mobile devices are requiring developers to learn OpenGL (at least if they don't use an abstraction tool Unity3D).
i don't develop games (yet, but who knows since i code for fun), i develop cross platform commercial applications. and what i learned in this process are 3 steps. but, this is how i do coding last 10 years or so.
1. be careful on language&tools. it will make a lot of unnecessary work if you use different one for each platform.
2. check libraries&services you depend on. either all have 1st class support everywhere or call it a bust. any kind of WIP is out of question
3. coding structure will either make you or break you. if you parse file path for example... never, ever do that in code. create universal wrapper method which resides in place where you know its platform unsafe and do it there while keeping your code as clean as possible. it will be a life saver if your main source code doesn't contain 1 single #if SOME_BS_OS. you simply restrict those for platform unsafe and when you check that later, you see the differences are minimal and requires almost no work if you worked correctly
being that my main platform is Linux, where i don't even own Windows or Mac (>90% of my customers do) i code/build/test Linux only. i basically test my apps on the fly at customer and so far it only posed problems 2 times with 5min fix. i can say this approach works damn well. my applications usually require way more dependencies than some game should. it also makes them far more complex to be cross platform than any game.
and just reading your comment about your friend, i can tell you he is not in as bad position as one would think. when porting something shows up as too much since you're neck deep in one provider, don't force it. either don't port or DON'T PORT AND GO WITH OCCAM'S RAZOR. while the 2 seem the same, they are not. 1 is abandoning port completely, other one is realizing futility and look from other perspective... "hey, if porting is too much, why not adapt it to wine since wine is adapting to windows?" and in this moment you get same small amount of bugs to fix as in normal clean cp application. it won't work optimal, yes... but, it will work good enough in most cases
0 Likes
Quoting: migiziQuoting: aaaaaaaaone would have to guess they went into multi platform THE wrong way, or at least they learned nothing from it.I feel bad for developers that fall for the propaganda of how they should develop and what tools they have to use. You're right that if you don't build with multiple platforms in mind the porting isn't worth it.
I know a guy who has been trying to complete a game for the past 10 years (multiple games have been attempted). He makes all the wrong decisions about the tools to use. He also won't touch Linux "until" it gets native DirectX support. He seems to think that Microsoft is in the process of open sourcing all that code and that Linux is going to adopt it.
So the real problem stems from the knowledge and research that a developer does before they start development. There are so many tools out there that you don't have to know anything to start making a game. Microsoft shifted the tools in use back when they released DX9 and the Khronos group failed to advance OpenGL fast enough. That stagnation period from 2.1 to the 3.x series is what pushed OpenGL to the back burner of development. Just look at games created prior to DX9, most either used OpenGL exclusively or had the option to switch between OpenGL or D3D. The saving grace is that OpenGL is now advancing faster than DX, it's not tied to an OS, and mobile devices are requiring developers to learn OpenGL (at least if they don't use an abstraction tool Unity3D).
give your friend this vid :p
View video on youtube.com
0 Likes
Patreon
PayPal
See more from me