Linux games are a dead-end business... NOT!

[GolfHacker]

We released Dirk Dashing for Linux 10 days ago, and we are simply stunned by the results! I thought it important to share this information, for those who might be interested.

To set the frame of reference for the following data, I should remind you that the Windows version has been out since October 6 (31 days ago).


As of today, 33% of our total Dirk Dashing sales have come from the Linux version. For the month of November so far, we've sold more copies of the Linux version than the Windows version each day. This surprised me, especially since the initial feedback on happypenguin.org came from unhappy laptop owners who couldn't play the game on their slow machines. If this trend continues, I expect the percentage of Linux sales to Windows sales to go up significantly.

Traffic to both of our web sites (mygamecompany.com and dirkdashing.com) spiked big-time on the last three days of October. On each of those days, we received about 20x the amount of daily traffic we had received on any previous day. Ever.

We've been swamped with lots of supportive e-mails from Linux users who have tried the game and enjoyed it. Almost all of them thanked us for porting the game to Linux and supporting their favorite OS. Many of them told us they had forwarded our web site address to their friends and family. We also had a lot of inquiries about Fashion Cents, and if/when we expect to make a Linux version available.

This past week also saw a spike in sales for our other games, and a number of customers entered various Linux news sites as the advertising source when they purchased the game. This also surprised me, since these are Windows-only games right now. But I found out a number of Linux users have their systems configured to dual boot, and they currently use Windows for gaming (since the gaming scene on Linux is so poor right now).


I must say, I was stunned by these all of these results.

I guess the moral of the story is that it doesn't always pay to follow the crowd. Nearly all of the advice and feedback I have read in various forums and articles is that Linux is a dead-end for companies and should be avoided like the plague. What really bugged me is the blanket statements that are made by people and the general perception of the Linux user base as nothing more than a bunch of people with radical views who refuse to ever pay for software - it also bugged me that such ideas are never challenged and are just blindly accepted by everyone.

Dirk Dashing for Linux was an experiment to test these statements for myself, and find out what the Linux market was really like. What I am learning is that the Linux user base is actually very diverse, and there are a lot of people who use Linux simply because they don't like Windows and want an alternative - at the end of the day, they don't care about the ideals of the FSF or the GPL, and they are not interested in open source politics. They just want something safe and reliable that they can use. And they are very hungry for commercial-quality games!

While Linux may not be a viable platform for every kind of application, I think it is certainly viable for games. And I am so glad we tried a Linux version of one of our games - this has turned out to be a huge shot in the arm for our business!

Finding information on the web about developing games for Linux is tough. It isn't documented very well. The information is scattered across a myriad of web sites, and there is a lot of conflicting information. I'm thinking about writing a series of articles in order to collect everything I've learned - what development tools and SDKs are available, ways to handle the problem of multiple Linux distros that are all a little different from each other (how to build your game, how to install it, etc to make it as distro-independent as possible), Linux-specific considerations to take into account, etc. Would anyone be interested in such articles?

[radishan]

It would be fantastic if you put something like that together.

It's great to hear of your success. Well done.

[mot]

I would be definitely interested in the packaging/installation process, how to make a distro-independent packages, how to create dependencies for them (ie. require Java 1.4 already installed).

[soniCron]

I don't want to speak for him, but I do know he's using Bitrock InstallBuilder (http://www.bitrock.com/products_installbuilder_overview.html).

[Sharpfish]

This is interesting (As all avenues for sales are), and it's good to hear about your success. I also believe it's good for linux users to have games available if they want to buy.

I am under the impression though that only core or games that appeal to traditional gamers would interest most linux users, is that a correct perception?

A strictly casual game may not find a happy home on a linux system but I would be interested in demographic info on linux users/installers to support that.

And to the original poster, did you also have MAC versions or just windows/linux, if so how did the mac sales compare?

[tolworthy]

How does the Linux crowd feel about emulators (or "not emulators" like WINE)? My game is made with the (Windows only) SLUDGE adventure game engine. But SLUDGE games tend to work fine under WINE. I would like to promote my game to the Linux world, but would they scoff and be disgusted that I couldn't offer them a "real" Linux game?

[Bad Sector]

As a Linux user, i would like to second that "Linux users are hungry for quality games". Most Linux games are either AAA games ported from Windows, usually requiring some configuration that a novice user could not perform, or badly written, unfinished opensource games.

There are exceptions, of course, but they're hard to find and very few in numbers.

Personally, if i was about to buy a game, i would buy the Linux version if it was available. And i would think it less to buy a game or not, if it had a Linux version... i like it when game developers support Linux, so i could buy a Linux game that i wouldn't buy if it was available only for Windows (okay, the game matters more, but i'm talking about the "buy-not buy" edge here).


Now, Aveyond for Linux? Please?

[Tertsi]

I'm definitely interested to know some of your development and marketing Linux information/experiences. With the upcoming nearly automatic Linux port for Mitorah Games Engine powered games I think there will start to be some more Linux 2D indie games.

[Pallav Nawani]

Mitorah Games Engine is going to be open source right? I will certainly try it!

As for game sales on Linux, it depends on the type of game you have. If you make a Breakout Clone or a Bejweled Clone, linux users are not going to buy it. Riotball got ...15 Linux sales in total. Now I am not claiming that it was a world class, commercial quality game, but for sure it was way better than any other linux breakout game. And well, almost all the feedback I got was positive.

On the other hand, linux users are happy to buy RPG/RTS/FPS etc. Basically linuxers are hard-core gamers, and that's the sort of games they are going to buy.

[Tertsi]

Mitorah Games Engine is going to be open source right?

Yes, there will be both open source and static library versions available.

[GolfHacker]

I don't want to speak for him, but I do know he's using Bitrock InstallBuilder (http://www.bitrock.com/products_installbuilder_overview.html).

That is true - we are using Bitrock InstallBuilder. That was the first step for us for creating a user-friendly installer. (There are some other options, but this was by far the best, in our opinion). However, we discovered that Bitrock does not do everything yet. For example, it doesn't automatically add shortcuts to the system menu - we had to do this ourselves. This is something I would capture in the how-to article.

[GolfHacker]

I am under the impression though that only core or games that appeal to traditional gamers would interest most linux users, is that a correct perception?

A strictly casual game may not find a happy home on a linux system but I would be interested in demographic info on linux users/installers to support that.

And to the original poster, did you also have MAC versions or just windows/linux, if so how did the mac sales compare?

The perception doesn't match my experience so far. Some of my customers are moms and dads with kids, and they are looking for games for their kids that will run on Linux. A couple of my customers are elderly (one gentleman actually bought my game for his wife who enjoyed it). I'm not sure about the others - I know about these customers because they wrote us and told us.

Because the Linux user base is expanding to encompass more and more Windows refugees, I personally think there is a growing market for lots of different kinds of games, casual games included. I know if some of my favorite casual games were available for Linux, I would snatch them up!

As for the Mac version of Dirk Dashing, we're still working on that port.

[GolfHacker]

How does the Linux crowd feel about emulators (or "not emulators" like WINE)? My game is made with the (Windows only) SLUDGE adventure game engine. But SLUDGE games tend to work fine under WINE. I would like to promote my game to the Linux world, but would they scoff and be disgusted that I couldn't offer them a "real" Linux game?

Two of my customers who wrote to us mentioned emulators. Both complained that WineX and Cedega won't run their kids' favorite Windows games, and that these projects haven't shown much interest in working toward those goals. Their opinion was that WineX and Cedega only focus on the mainstream/hardcore games, and many of these are violent. They want educational programs and clean family-friendly games for their kids that will run on Linux. I think they might be satisfied if WineX and Cedega supported their kids' games, but I personally think they would be happier with games that ran natively on Linux. After all, Cedega isn't free - it is a subscription. And native games are likely to run better anyway.

[GolfHacker]

On the other hand, linux users are happy to buy RPG/RTS/FPS etc. Basically linuxers are hard-core gamers, and that's the sort of games they are going to buy.

I disagree with this completely. This doesn't match my experience so far at all. And I think this is the very sort of mindset many developers have that needs to be challenged, because it just isn't true. Maybe it was true in the past, but not anymore.

Market conditions are changing. There are a lot of unhappy Windows users that are looking for alternatives, and I personally know a lot of people who are not looking forward to Vista. Some of my existing Windows customers have already moved to Mac, others to Linux. And they are asking me if/when I plan to support other platforms. And these are not what you would call "hard-core gamers".

And as I said, the level of interest in Linux versions of our other games (which some would probably classify as casual games) was not something I originally expected.

I think there is a growing market here that indies could potentially do very well in.

[tolworthy]

Two of my customers who wrote to us mentioned emulators. Both complained that WineX and Cedega won't run their kids' favorite Windows games That's interesting. Thanks.

[zoombapup]

So we are basing a discussion about the relative possibilities of linux sales based on 10 days worth of information?

I think statistically, we can pretty much discount that kind of information unless we are talking LOTS of sales.

If people were rushing to buy the linux version over the windows one and we were talking SIGNIFICANT sales, then maybe I would trust it.

But you didnt mention how many sales. So it could be that 5 bought the linux version and 1 bought the windows and oooh.. amazing %age of linux buyers!

Only.. it doesnt really work like that.

Let us know some figures in 6 months time? That might be more interesting.

[GolfHacker]

Tolworthy,

If SLUDGE games run fine under WINE, then you might be ok. You will definitely want to advertise that up front. Just be aware that WINE isn't necessarily easy to configure and use for the average user, particularly one who has just migrated from Windows and knows next-to-nothing about Linux or WINE.

The easier you can make it for a user to download, install, and run your game, the better your sales are likely to be. Conversely, the harder it is, the more likely the user will give up in frustration and cost you a potential sale. If you are committed to using SLUDGE, then at a minimum, you will want to provide step-by-step instructions for your Linux users for how to install and configure your game under WINE so they can run it. If possible, provide them with a script that automates the setup and configuration (or better still, incorporate the script into your installer so the user never has to deal with this setup and configuration at all).

[GolfHacker]

So we are basing a discussion about the relative possibilities of linux sales based on 10 days worth of information? ...

Let us know some figures in 6 months time? That might be more interesting.

Point taken. I'll consider that.

However, the initial results of the Linux release were so surprisingly to me, I wanted to share my findings for the sake of starting a discussion and getting people to start thinking about it. If the market conditions for Linux games really are changing, then I thought this is something people should be aware of, and something they might want to start watching. But what do I know? I'm just a little startup that writes games part-time. Feel free to ignore this thread. Or conduct your own experiment.

[michalczyk]

That's great to hear golfhacker! I believe Linux is far bigger than most win users think. Just see how many Linux mags there are today. Not too long ago there used to be only one. All those linuxers must be really hungry for good games since there are so few of them around. A nice vertical market for indies to fill in.

[amaranth]

This is great news GolfHacker! I for one am going to this this avenue a shot. :D

[arcadetown]

Similar to Mac versus PC for developers. Developers can get direct exposure on Linux sites so they'll get a higher % of direct sales from it. However in the larger market it would probably provide a tiny % of sales (http://forums.indiegamer.com/showpost.php?p=113417&postcount=8) and questionable if worth the effort. I suspect it would even be lower % than Mac as our internal stats show Linux users are just 0.2% compared to Mac being approx 1 - 2%.

[Ryan Clark]

Just wanted to add our experiences to the thread, though we've also posted them elsewhere.

We get about 2 to 3% of our sales from our Linux versions. Not a huge result, but definitely worth the extra week or so of effort to get the Linux versions running. I also suspect that we get the occasional "sympathy sale" from people who use Windows for gaming but Linux for everything else, and buy our Windows version in part because they want to support developers who make Linux games.

As Brian said, Linux sites are fairly easy to get exposure on... it's like a mini version of the Mac market. This will get you an easy spike of Linux sales at the start, but things always taper off after that and Windows sales will tend to predominate.

[Pallav Nawani]

I disagree with this completely. This doesn't match my experience so far at all. And I think this is the very sort of mindset many developers have that needs to be challenged, because it just isn't true. Maybe it was true in the past, but not anymore.
Well, it matches my experience, and I've been selling RiotBall for 2 Years now...
Congratulations to you if your game is selling well on Linux. I will note though, that your game is the type of game that linuxers like (a action platformer), and does not disprove my earlier statement.

[Bad Sector]

So, it's a nice thing that i make my games in Linux and then port them to Windows :-).

Although doing a proper Linux release is more than simply compiling a Linux executable.

[GBGames]

So, it's a nice thing that i make my games in Linux and then port them to Windows :-).

Although doing a proper Linux release is more than simply compiling a Linux executable.

I'd actually like to see such information collected somewhere. I'm currently researching what is needed, and as said above, the information isn't all in one place or consistent. gpwiki would be one such place, I think.

[Bad Sector]

Well, first of all you'll need to read and understand the linux Filesystem Hierarchy Standard (http://www.pathname.com/fhs/). This says what goes where, etc. But to make it easier, for commercial games you probably want to place them in /opt/your_game. However the correct thing to do is to break the game's files into "binary", "crossplatform data", "platform specific data", "configuration", etc and put them in the correct directories (usually inside /usr/local, because the convention goes that anything in there is managed by the user and everything else is managed by the distro's package system).

Then, you'll need to figure out how to create an installation system. Creating .run files (self-extracting and installing, based on shell scripting) is explained here (http://linux.org.mt/article/selfextract). This script should extract itself into a subdit /tmp where an installation program gets compiled (you can use Gtk2 and Glade to build the installation program, it doesn't have to be anything fancy) and executed. This will handle the game's installation process (that is, copying the files to the correct locations :-).

Another issue you want to take care of, is shortcut creation. You need to detect which windowmanager or desktop environment the user has and create shortcuts for this. You should create shortcuts for at least KDE and GNOME, but WindowMaker and FVWM (or whatever is the name) are quite popular too. Do not create desktop shortcuts, no matter how tempting it may be. Both KDE and GNOME use the ".desktop" subdirectory for the user's desktop and both use different shortcut formats, so creating a KDE shortcut won't work in GNOME and vice versa. Many distros install both desktops by default and this is the configuration you should expect in a newbie's system.

Also note that Debian, especially, has a "menu system" which is a simple tree structure for programs with a "games" entry. You should put a shortcut there if the installer is ran from Debian because it makes the shortcut in almost every menu system, windowmanager and desktop environment used.

[GBGames]

Ok, let's put it on the wiki! Otherwise, it is just another forum to search through to find the information. B-)