Related link: http://www.thepetitionsite.com/takeaction/190498263

Serious gamers are an interesting bunch. Some of my college buddies were a
representative bunch — leaving the cases off their computer because they
swapped hardware so often, reinstalling Windows every couple of months to
improve the stability threatened by installing and uninstalling games and
DirectX components so often. Releasing dedicated servers really taps in to
that do-it-yourself mindset.

There’s a similar spirit to be found in many Linux users. Witness the
popularity of install-or-compile-it-yourself distributions. There’s something
very satisfying about knowing how everything fits together, having put it all
together yourself. (Of course, the same rugged individualism can lead to
myriad identical projects with big goals, no code, and no future.)

It’s no surprise that game companies are aware of this mentality in their
most dedicated customers. Running your own game server is appealing. You get
to set the rules. You choose the maps and scenarios. You (finally) have a
good network connection to the server.

Of course, this also means that you’re helping keep the game alive.
Instead of the company paying for massive server hardware, upkeep, and
bandwidth, you’re responsible. The joy of the hack only goes so far —
system administration is either boring or exciting, and exciting system
administration is bad. It’s not hard to see why server administrators are
starting to say, "Wait, I’m doing this for free. Is it really worth
it?"

That’s why the href="http://www.thepetitionsite.com/takeaction/190498263">Valve
Software Linux Server Boycott Petition is interesting.

While it’s nice that proprietary game developers acknowledge that Linux is
a viable platform for their game servers, it’s unfortunate that it’s not seen
as a viable platform for game clients. If the base game logic compiles and
runs successfully on Linux already, how much more work is it to port the
graphics and sound?

I’m somewhat sympathetic to the response “Quite a bit, actually.”
As Chris DiBona
unequivocally puts it, “The games industry is all [fouled] up.” When nearly
every retrospective admits that the project spent months in deathmarch mode
trying to get a game out on time, you have to wonder when someone’ll figure
out that something’s deadly wrong. Still, there are several reasons of
varying legitimacy why more game publishers don’t want to publish Linux
clients.

  • Low market size (though if you don’t publish, no one can possibly
    buy).
  • Portability concerns (though a well-designed game will minimize
    platform-specific code and good libraries such as href="http://www.libsdl.org/">SDL can take care of much of the rest).
  • Lack of time (though this is more a feature of a broken development
    process).
  • Lack of knowledge (though people such as href="http://icculus.org/">Icculus have been known to work very quick and
    very impressive magic).
  • Lack of standards (though the state of video and audio on Linux is better
    than ever and continues to improve).

I don’t necessarily agree or disagree with any of those points. It’s the
game publisher’s decision whether to port a game to Linux. (It’d be really
nice to have source releases, as not everyone uses Linux on x86 chips or Linux
at all — but you don’t really get anyone’s attention by talking about
FreeBSD gaming.) Why should a publisher put out a Linux port if people will
just buy the Windows version and play it under Wine, WineX, or the token
Windows box?

With that in mind, though, it’s also up to potential customers whether
they want to buy the game. This petition will likely have no effect on Valve
(in an industry perhaps most charitably described as “arrogantly adolescent”,
self-absorption is all too common), but it brings up a question Linux gamers
should consider.

Why is Linux good enough to host your game but not good enough to play it?

Are there other questions Linux gamers should be asking?