Slashdot Mirror
Why Gaming Sucks On Linux
lseltzer writes "Efforts have been made to improve the situation, but things have actually gotten worse for gaming on Linux rather than better. If you're a gamer you're just plain better off running Windows and dual-booting (or VMing) between the two operating systems than hoping your games will run in Cedega or some such product." From the article: "So where does all of this leave Linux gamers? One word: Windows. Yep, you read that right. If you're a gamer, do yourself a favor and just buy a copy of Windows and set up a dual-boot system. Why bother to torture yourself with the headaches presented by Linux gaming? Why should you continually not have the games you want to play? Why settle for half-assed solutions that might or might not run the games you crave so desperately?"
Dual booting is okay for games, but how about when you want to play music in the background? And all of your music is on your ext3, Reiser, or XFS file system? How is the read/write support from windows back to linux? I think there have been advances on read/writing NTFS from linux, but how about the other way around?
It really drives me mad when slashdot refuses to post articles about the last 3 games we released, despite at least 30 or 40 people (that I know of) sending in messages about it, and then go criticise the state of Linux games. If they did their bit maybe our company would be in a better position to get the licenses for more games.
Tux Games. Your complete source for native Linux games.
Maybe open source guys should get off their behind and create a decent competitor to directx. Open GL is an ok solution, you have write extensions to do anything fancy and there's directx which has everything there for you.
Have you ever been to a turkish prison?
I used to also get frustrated that I couldn't play
any of the windows games on my Linux box and that
none (or nearly none) of the game makers would port
their games to Linux. It took me a long time to realize
what the problem was:
Simply put, the game makers can't keep up with the pace
of Linux. Let me try to explain what I mean. Take a look
for example at how many Linux distros there are out there:
something like 300 now or so based on distrowatch.org.
And some of those distros run on different hardware platforms
than just x86.
On top of that, not every Linux distro has the same versions
of the same libraries at the same time with the same patches
and compilation options.
There is no way that the proprietary software makers can
keep up. That's why Oracle only supports RedHat Linux and
only for certain versions. This is why Adobe's effort to
provide 1 flash plugin (version 9) for _all_ Linux distros
by doing some funky static compiling is at best misguided
and at worst massively brain damaged and stupid because it
won't work for any platform except for x86.
It's no wonder that the game makers prefer writing to windows:
It's not just that windows has 90% market share -- what kind of
profit oriented businessman would say: I could make 10% more
profit with a little bit more work --- but screw that, 90% is
good enough at most! And that's assuming they _could_ get massive
penetration when realistically all they could expect is at most
10% of the windows population.
The genius of Microsoft Windows is _not_ that it's ubiquitous.
The genius of Microsoft Windows is that it's stably obsolete ---
it's glacial in its pace of development. Five freakin' years of
the _same_ libraries! My God, that's hog heaven for proprietary
software makers!
Contrast that with the Linux world, where 6 months is a lifetime.
If you're software is older than 6 months, it's already obsolete
and might as well be dead. The only kind of software that could
survive that kind of testosterone laced Darwinian death-to-the-weak
Jungle is open sourced software, and only in the case that the
Linux distros maintain the downstreams versions of the software
package to guarantee it will work. So, it's no surprise to see
that the only (popular) games that run on Linux natively are
open source games: Tuxracer, frozen bubbles, gcompris, bzflag,
cube, nethack and all its variants, blob wars, etc.
The only counterexamples that come to my mind are:
Enemy Territory. Are there any others?
--Johnny doesn't use proprietary software if he can help it.
OpenGL != DirectX.
OpenGL is close to what Direct3d offers, but Direct3d is offering a LOT more than OpenGL currently can. Even the latest additions to OpenGL do not bring it to what Direct3d 9.0, let alone 10.0 offer
Now if you want to include the fact that the DirectX suite has a tonne more functionality than just Direct3D, like DirectSound, DirectPlay, DirectInput, DirectDraw, there is no way in heck that OpenGL can compete with it. DirectX is exactly why developers have made the move to it (including Carmack), you don't have to worry about all those things like sound drivers and engines, network stack plug-ins and the like.
Compare Direct3d to OpenGL all you want, but before attempting to FUD, at least know your facts.
Until Linux offers a similar all in one API for game programming, and until Linux users actually go out and pay for the software (read about Carmack's tracking of Linux SKU's for Quake sales), then there is no way in hell any developer interested in making money is going to focus the time and resource to make a portable game, espeically for the Linux platform.
And you know what? They're probably wasting their time. The people who are interested in indie games are going to be gaming hobbiests who are gonna have a closet full of consoles and a Windows machine. There's the techie/computer nerd crossover, but even the majority of that crowd doesn't use linux.
Gaming on Linux sucks because users need a reason to switch to Linux first to make it a viable platform. OSX is a success and it still sucks for gaming, why should Linux be any better? "Because it's free and open source and you should support open source" isn't really an answer...
Wise men say, "Forgiveness is divine, but never pay full price for late pizza."
Older doesn't mean good either.
There's the chance you'll want to play with your friends. There's the chance not all of your friends run Linux, and they may have played and grown tired of the older (though not necessarily good) game you just bought.
We see this type of comment all the time. A few problems:
1. Drivers. You have to include drivers for all the hardware that your game needs (graphics, sound, network, input devices), including future devices in these categories. I've heard some interesting ideas for ways around this. They're interesting, but lead to a complicated and lengthy install process. You need to get your OS to boot on everyone's crazy hardware configurations.
2. Networking and patching. If you're making a network enabled game you're asking people to put their computers on a network (likely even the Internet) with an OS that can't be updated without spinning another DVD. There will be security holes found in the OS. Lots of people running the exact same unpatched OS version will be playing these games. Just because the OS installation on the CD/DVD can't be modified, it leaves open an attack vector to data on the hard drive. Game manufacturers like the ability to easily patch their games for bugs, anti-hacking techniques, and other random things. Argue all you want that they should get it right the first time, they wouldn't want to give that up.
3. Rebooting. People don't like rebooting. It takes a long time. They have to disconnect from IM programs, they have to turn off their music players, etc. They lose their software stack and configuration info (think configurable input devices that require userland apps for configuration). Configuration that would be shared between multiple games must be redone for every game you buy. It's more difficult for people to minimize the game and post the video of their latest frag to their website.
4. Licensing. What OS would game developers use for this? Windows or some similar variant. The driver support and developer tools are there for Windows, and most PCs sold today have undergone QA on Windows. Paying the licensing costs would drive up the price of the game.
And what is gained? The day-to-day experience (that is, the experience outside of patching and installation) for Linux/Mac users would be the same as now: a reboot into an OS they use mostly for gaming; in fact, since a real Windows installation would be more useful than the game OS the experience would really be slightly worse. The day-to-day experience for Windows users would be much worse: two reboots to go from regular use to game and back. Reboots between different games.