Linux has evolved to a pretty decent Desktop platform these last years. It features a fast, stable and productive work environment for various purposes. Linux (and especially Ubuntu) is doing his job as a everyday desktop system very well - maybe even better than Windows in some cases. Since when do you get a whole office suite shipped with your OS; for free. Ah yes, you bought your machine at where you get both your OS and your office suite installed on your computer. But maybe your preinstalled Windows doesn't even have a photo album manager, or an audio editing suite, or a versatile graphics program, or.... Well, you get my point!
So, Linux seems to be more or less comparable to Windows when it comes user-friendliness and software range. However, there are alot of areas in which Linux should improve. One of them is a very important one in modern computing. And that's Gaming.
Gaming seems to be the blind spot of Linux. In my opinion, Linux's popularity is so low, because it lacks incentives as a gaming system. And lots of people use their computer for gaming, thus it appears that there is a huge potential in gaming. But the actual problem is the lack of funny and well-made games for Linux.
But why? Why are there so few Linux games around? Linux has more or less stable support for the latest graphics cards, it doesn't lack any kind of OpenGL functionality, it doesn't lack speed or stability, it has possibilities when it goes to realization and development of specific parts in a game engine which also run on various other platforms - why is everyone avoiding Linux in game development? Why are there so few Open Source games that can be compared with commercial titles anyway? Free Software should be a model that applies to almost every situation, shouldn't it...?
The actual thing is a bit different, though. The reason why commercial games are rarely ported to Linux is clear: It's not profitable. Way too few costumers. Simply creates unnecessary extra costs and work, which can be avoided. The only way to change that is to raise the popularity of Linux itself or by deliver an extraordinary working environment. And probably won't change alot in the following years.
The other thing is free (as in freedom) games. A very strange phenomenon. They are developed using the same model that is very successful almost everywhere in the IT world - but they still can't compete with commercial titles.
The main reason behind this is that most FOSS projects can't count on experienced and talented artists. Just in case that you don't know what artists do: They draw conceptual sketches which illustrate the style and the atmosphere of the game; they make and animate 3d models; they make textures; they are responsible for the visual part of the game and the realization of main gameplay concepts (such as leveldesign). This part gives the game an atmosphere (along with other important elements such as storyline etc.)That's the problem of most FOSS games: They don't have a particular style, they lack atmosphere, detail and realism (or "rational" surrealism).
So, where will we get those people with superpowers called artists? (I don't want to say that artists are the key to everything, but you can't ignore the importance of specialized people within such a project. There's a very wide spectrum of arts and skills involved in gamedevelopment: Programming, Design, Drawing/Sketching, Digital Arts, Music and SFX, Writing) That's pretty simple: By providing an advanced, artist-friendly and productive work environment. An engine or framework where also someone with a limited knowledge of programming can work from. A nice toolchain, a user-friendly frontend. That's actually it. If there is potential, people will be interested, people will use that software. That's something elementary - a programmer thing. However, something like that still doesn't exist in the Open Source world.
3D engines like OGRE, Irrlich or Crystal Space (which is more or less a game engine, too) need lots of programming (or have a very, very unfriendly scripting interface) and don't feature toolchains that actually deserve that label. The Quake 3 engine (and all its forks) is technically good, but its frontend for artists is outdated and horrible. Sauerbraten features easy leveldesign and shaders, but fails when it goes to more complex stuff.
Not only the engine toolchain is important, but also all available programs for creating art, such as GIMP, Krita, Blender, K3D etc.
This issue must change if Linux and Open Source game development shall improve.
PS. I'd personally suggest to create a central knowledge base or community for Linux/Open Source gamedevelopment and art.
-Andreas "Sindwiller" Ratchev