Games: The Boundary of Open Development?
by: Sam Williams

To help determine the natural boundary between Open Source and closed source game development, BeOpen consulted with two of the leading members of the Linux gaming community.

Scott Draeker is the president and CEO of Loki Entertainment Software, a software company that specializes in porting popular PC-based games over to the Linux platform. Although Loki has donated many parts of its software infrastructure to the Open Source community, its games remain proprietary, mainly because of licensing restrictions but also because of what Draeker sees as the limitations of the Open Source development model.

Jorrit Tyberghein, meanwhile, is a volunteer project leader for Crystal Space, an Open Source, LPGL-licensed 3D engine. Like Draeker, Tyberghein sees the Open Source development model flourishing in some areas and struggling in others.

BeOpen: Computer games represent an interesting frontier for Open Source development. Unlike other desktop applications, games tend to be hybrid organisms -- half software program, half artistic work. How do you balance the strengths of the Open Source development model with the entertainment demands of the medium?

Draeker: A computer game is a creative work which is expressed in software. Some kinds of games require software as a medium, while others are played in mediums such as game boards or cards. It's important to keep this distinction in mind when discussing computer games. While much of the development work amounts to software engineering, it really goes to being able to successfully present the game you are creating to the user. The software is not, nor will it ever be, the game itself.

When we talk about Open Source we're generally talking about software. Since a game is not software, it is a little misleading to talk about an Open Source game. It's a little like talking about an Open Source Star Wars.

Many people have distributed games in a medium of Open Source software. Most of these games are fairly simple and resemble the shareware and freeware games available for other operating systems. Commercial games, on the other hand, generally take a team of full time programmers, artists, producers and support staff several years and millions of dollars to complete. It's difficult to imagine how Open Source developers could match the speed, quality and quantity that the commercial gaming industry gives us each year. It would be like trying to film Star Wars with your friends on weekends.

What does make sense is to use as many Open Source tools and libraries as you possibly can. A game developer using OpenGL, OpenAL and SDL could spend more time concentrating on the game aspect of the project and less time writing software that already exists. Open Source solutions reduce costs and development time for both commercial and shareware games. They also promote open standards which improve the user experience and the reliability of your product.

Tyberghein: I think that the Open Source development model is at least useful for the coding part of a game: the 3D engine, physics, AI logic, even game logic. I'm not sure about the art that is used in a game -- i.e. textures, models, sounds, music, etc. I don't think the current Open Source licenses can easily be used for those and I can imagine that an artist might be a little reluctant to just release to the for public. Also there is a fundamental difference between Open Source and Open Art. The main reason to release something Open Source is so that you can create a developer community that will develop and improve upon the project. I don't see that easily happening with art. Every so called "improvement" on an artwork will most certainly not be seen that way by the original artist.

What I think will happen more and more in the future is that game companies will either use available Open Source tools -- like 3D engines -- or else create their own toolkits that are released as Open Source. These are then combined with copyrighted artwork and maybe closed source intimate game logic.

This does not mean that I think open art is impossible. I just think that it is less beneficial compared with Open Source.

BeOpen: Where do you see the Open Source development model flourishing in the commercial game "space?" Will it be limited to infrastructure technologies and tools or are there enough economic incentives to drive 100 percent Open Source game development?

Draeker: Game companies make money selling games. How do you make money when everyone can play the game for free? If there's an answer to that question which increases the financial return to game companies, they'll jump on it. Unfortunately, I'm not certain that model exists. Are you going to show people enough ads in the middle of the game to make it profitable?

Open Source infrastructure does make sense. The product is after all the game and not the software. If Open Source can reduce the cost of creating that software, then the game will cost less to produce and be more profitable -- or at least lose less money!

Another option might be to Open Source all the software for the game, but require that the data files (art, sounds, maps, movies, etc) be purchased. This approach is similar to what companies like id Software have done for years, releasing enough source code to allow users to modify the game and create their own add-ons, but still requiring users to purchase the full game in order to use any of those mods. This is great for users, but keep in mind that the initial development of the product was done in the normal, closed-source cathedral manner.

Finally, Open Source is an excellent way to give your commercial game immortality. There's very little commercial incentive to support a game even just a few months after its release. Your fans, however, will keep a game alive indefinitely if given the chance. Just look at all the effort put into emulators, ROMs, clones and the like. Let them, and you'll be rewarded with a diligent, enthusiastic and free, marketing and development team who will keep the franchise alive.

Tyberghein: I can see 100 percent Open Source commercial games happening in the (near) future. What you pay for when you buy that game is the levels that have been designed, and the artwork. You don't pay for the game code itself.

The same reasons that are valid for using Open Source for infrastructure technologies are also valid for games IMHO.

BeOpen: What sort of a relationship do you see evolving between commercial game companies and the Open Source development community? What efficiencies can Open Source development offer to coax commercial game developers away from proprietary software and proprietary hardware platforms?

Tyberghein: There are two kinds of relationships between commercial game companies and the Open Source development community. The first relationship is commercial game companies releasing their engine code as Open Source. This has already happened with Quake and some parts of the Unreal engine. The second relationship is commercial game companies making a new game using a toolkit -- i.e. 3D engine -- already found in the Open Source community.

The first relationship has already happened a few times and I expect to see more of this in the near future. As far as I know only Loki software currently uses the second strategy. They use SDL (for example) to port commercial games to Linux. In the future I also see the second strategy used more -- commercial games using Crystal Space. for example.

Draeker: id Software has already developed a great relationship with the "mod" community, and they've done this by supporting mod coders' efforts with the release of specifications and source code. The mod community thanks id Software by using those tools to add new content to id's games in the form of maps, skins, mods etc. There isn't much of a Linux mod community yet. We've been working to address that by releasing Linux versions of the necessary tools.

Coaxing developers away from proprietary tools and platforms isn't a matter of evangelism. These people know what their options are, and will generally go with the solution that offers the best return on investment. For now, writing games for Windows and the various consoles makes sense because the market is huge. Many Windows development tools are more mature and feature rich than Open Source alternatives. Loki is dedicated to addressing these discrepancies, but we can't do it alone.

Building the Linux game market is also the responsibility of Linux users. They have to stop buying Windows games. If you buy Windows games, the market will reward you with more Windows games.

BeOpen: How will the growing prevalence of Linux as a gaming platform assist the development of Open Source games?

Draeker: In two ways: Through Loki's efforts and others, the necessary multimedia infrastructure is being created and released as Open Source. That infrastructure is necessary to support both commercial and non-commercial games.

Secondly, a growing Linux gaming market means more users. More users mean more potential customers. That's important even for a game you give away for free. After all, who wants to go to all the effort to create a game if no one will ever play it?

Tyberghein: I'm not sure that Linux as such actually makes Open Source easier. The Crystal Space project runs both on Windows and Linux -- and many other OS'es -- and it is Open Source on all those platforms. However Linux is the obvious choice for Open Source development mostly because historically most Unix and Linux software has always been Open Source. Open Source is much more common on Linux than on Windows.

Sam Williams is a BeOpen.com staff writer who spends his time gaining broad perspectives on the people behind Open Source trends.