[Linux-aus] Computer users and linux: moving upstream to software producers.

Paul Wayper paul.wayper at anu.edu.au
Fri Feb 16 14:36:12 UTC 2007

Chris McCormick wrote:
> On Tue, Feb 13, 2007 at 12:19:08PM +1100, Sridhar Dhanapalan wrote:
>> Considering that so many commercial games have their server counterparts 
>> working on GNU/Linux, how hard would it be to have the clients working as 
>> well?
> That really depends on the implementation, libraries used, etc. It's
> impossible to quantify that with a general answer, except to say that
> it would be possible with some games.
I heard an epsiode of Leo Laporte's show FLOSS Weekly where they
interviewed the guy who maintains the SDL - which now is an extensive,
portable, graphics and sound library whose primary market is games
writers.  Quake 4 and Neverwinter Nights, for example, are written using
it.  It would allow easy porting of those applications to Linux, and
indeed Quake 4 is available as a Linux download (i.e. you download the
Linux binaries and it uses the game files from your purchased CD).  I
believe Google Earth was also written using SDL, or at least the Linux
port was.

For MMORPG and other games where you need to be connected to a server to
play, the server code need be nothing like the client.  Even for regular
multiplayer games which usually have one person running as the server,
that code could still be portable but the game on top of it be bound to
Windows.  So it really does depend.  I believe that games that are
released on multiple platforms - e.g. PC, X-Box, PlayStation 3 - are
often internal ports with completely separate code bases.

The fundamental problem is that all platforms (not just Linux-based
platforms) are quite different - SDL is obviously trying to bridge this
gap.  Console game designers hate Windows, because you get half a dozen
versions of Windows, all with different service packs, different
versions of DirectX installed, and different hard disk layouts.  They
probably fear Linux tenfold out of a perceived idea that we have even
more variability than that - different libraries, different directory
layouts, different packaging methods, different GUIs.  The truth is less
than the fear - there's been a lot of work in recent times to put
configuration information in the same place on different distros, to
pick one example.

I think this fear stems from the assumption by games distributors (note,
not games developers) that all game users, on any platform, are total
newbs who wouldn't know a graphics driver from a besser brick.  They
want to package up the whole experience so that there's no possible
chance that anyone might blame bad graphics performance, crashes or
other failures on their game.  The reality, on Linux, is far from this:
most people using Linux are still reasonably knowledgeable about what
you have to do to get a program working, and where the fault might lie
if it doesn't work as you expect.  IMO, this extends beyond Linux: there
are lots of forums for games that are filled with advice on how to get
particular combinations of hardware and software working, and what might
be causing your problems.  Very little of this includes pointing the
finger at the games manufacturer.

Sure, the newbs will still blame everything on whatever they can find. 
And there are games that fail the 'expected outcome' test on reasonable
hardware.  But the 'long tail' argument is proving that you don't have
to cater for the lowest common denominator anymore.  I suppose I'd like
to see a statement from someone like id, who has made Quake available
for Linux, to see whether they've had any problems with supporting the
Linux platform.

But overall I think this needs to be seen as part of an overall move
towards awareness of and support for FOSS in the software industry. 
People quote gaming, or business, or CAD, or music creation, as the
'killer app' that, when it 'comes' to Linux, will draw everyone
thereafter.  The reality is that it's all of those and more, but each
one is only a small part of the picture.  And, gradually, it's all
coming together.  I wouldn't be so bold as to posit the Year Of The
Linux Desktop - but then there was no 'Year Of The Microsoft Windows
Desktop' either.  Talking about the games industry is useful, because it
highlights some of the issues that the FOSS community has to address,
but that's not the sum total of the issues.

Enough of my ramblings,


More information about the linux-aus mailing list