This will be an exercise in careful reading.
I'm a little hesitant to come out boldly for this opinion... but when I weigh the pros and cons, it makes sense. In summary, it's my "Tools vs. Toys" philosophy. I think games, as in video games for computing systems, should not be held up to the same standards of FOSS freedom as other kinds of software.
Tools are different. Tools are compilers, editors, web browsers, servers, drivers, platforms... in all of these cases, I have seen that consistently the Free/ Open Source development model not only makes maximum liberty possible for everyone, but produces a higher quality product.
But I see that it's not necessarily so with video games. In games, the results seem to stack up at random. I know some Free/ Open Source games that are better than many commercial/ proprietary games, and I know some commercial/ proprietary games that are better than many FOSS games. Some win by engineering, while losing on media quality, and vice versa.
The Linux blog world has been buzzing a little about the game market on Linux. Unfortunately, the current examples cited in that link just happen to be products that the Cult of Helios is shilling for, so, ugh! Can't touch that with a ten-foot pole. As it is, simply because they happen to be focused on games right now, I can expect the flamin' hot-pants idiot brigade to descend on my comments section like a plague of locusts. (And H. goes from never mentioning games once in all his years to suddenly talking about nothing on his blog but these games for six months? The money's going in his pocket, bet your fur.) The Cult of Helios craps on everything. But oh well.
Notwithstanding those examples, the big picture is that yes, Virginia, there is a market for Linux games. And I think encouraging that market is a good thing to do. As I rediscover with both excellent FOSS games (like Neverball), and excellent proprietary games (like Quake Live), there is no reason that Linux cannot be as good a gaming platform as any other system.
But I just think that making all games for Linux have to be licensed as FOSS isn't that important. I will carefully lay out my reasons:
Games are not mission-critical. The world will not come to an end if you can't patch a game. It's not like you have a printer that won't print and you can't fix it because the license is closed - the example which famously prompted RMS to create GNU.
Games are more about the artwork than the code anyway. If you're talking pure engine over visual, audio, and interface appeal, heck, Nethack beats everything. But the Diablo series is very similar in playing style to Nethack, but whoops butt on the sound, video, and interface department.
Artwork is very seldom free. We saw this with idSoftware's previous releases of FPS games as open source. The Doom source code is now open, but you still have to pay for the levels packs, because the art, sound, and design is a separate matter from the code. Code, while being an art form, is also an engineering matter, and is so easier collaborated in the open method. Given enough eyeballs, all bugs are shallow; but artwork, on the other hand, usually suffers in quality when it has too many hands involved. Ever seen a film where the script was written by a committee?
Art takes more time than code. I, myself, have offered a lot of artwork to the public for free. But I also do commercial work. I'll be the first to tell you, I put a lot more effort into graphic design when I'm paid for it. Code, on the other hand, doesn't take the same proportion of time. If I work for two weeks concentrating on just one project, I can design a major software application that possibly may be world-class and land me a spot in history, or I can draw a half-believable interior room that will get a few votes on DeviantArt, but will still not be as popular as a funny stick-figure comic.
Games are less satisfying to design than applications. By that, I mean that when you make an application that does something, you are giving the world a tool. I would far rather have been the author of something like sed than I would the greatest five video games made today, because the author of sed can know that generations of users will go on using his tool to make life easier for themselves, whereas games will have a brief following of fame that decreases as the game becomes outdated, eventually being forgotten by subsequent generations (not counting rare nostalgia/ abandonware nuts like me).
Game enjoyment doesn't seem to be impacted much by the source code being open or closed. As long as it installs and plays when you want it, do you care?
Now, I told you that this would be an exercise in careful reading. If you've made it this far without posting your flaming rebuttal, you know that the following misinterpretations are what I did NOT say:
- ...that all FOSS games are bad.
- ...that all proprietary games are good.
- ...that all games should be closed.
- ...that even all parts of games should be closed. Physics engines, for instance, might fare better as FOSS, so mixed licenses might do well.
- ...that FOSS games can't make money.
- ...that DRM is justified.
- ...that proprietary software companies like Microsoft and Adobe are justified.
- ...that proprietary licenses for video games don't also have problems.
- ...that a robust, profitable market in FOSS games is impossible.