If you've got 3.5 hours to kill, John Carmack's Quakecon keynotes are always fascinating. id games may not be the greatest titles around, but Carmack's knowledge and skills continue to shape the future of gaming across multiple platforms. There's been a great deal of rumor and speculation about Valve's plans for Linux support, including the Steambox (which trusted sources have told us is an actual project). Gabe Newell's distaste for Windows 8 is well-known, and while I happen to disagree with him regarding Windows 8's likely impact on the game industry, there's no denying that Valve's actions will have an impact.
Carmack thinks Valve's move could shake things up and spark fresh interest in Linux as a gaming platform, but doesn't ultimately believe that Steam will change much. His remarks are below, slightly edited for clarity:
Linux is an issue that's taken a lot more currency with Valve announcing Steam for Linux and that does change things a bit, ...but we've made two forays into the Linux commercial market, most recently with the QuakeLive client, and that platform just hasn't carried its weight compared to the mac on there. It's great that people are enthusiastic about it, but there's just not nearly as many people that are interested in paying for a game on the platform. And that just seems to be the reality.
There's at least one explanation that fits Carmack's experience without torpedoing the idea of Linux as a gaming OS. Historically, Linux and gaming have been viewed as nearly mutually exclusive. That doesn't mean there aren't some great Linux games with really talented people working on them. It doesn't mean there aren't any ports of existing popular titles. What it does mean is that gamers interested in switching to Linux have always been told not to expect parity in terms of the number of titles, frequency of updates, or launch windows. There are plenty of Linux or Mac users who maintain a Windows box for precisely this reason.
If the population of Linux users has been self-selected from those who don't game, then it follows that id would have trouble launching a commercial Linux game. A handful of titles -- even really excellent titles -- isn't enough to create a sea change in the user base. This is where Valve could make a much bigger difference if it can convince other studios to port to OpenGL. This will be critical for any sustained movement. Valve, like id, has a relatively small stable of games, most of which are old and have already been played by self-identified gamers on other platforms.
When Microsoft or Sony releases a new console, there's a push-pull effect. The 'push' comes from knowing that the previous system has reached end-of-life. New games won't look as good or play as well and eventually titles will dry up altogether. The 'pull' is obvious -- buying a new system gets you new and improved games plus whatever other goodies have been baked into the platform.
The problem with Steam on Linux as opposed to Steam on Windows is that it's going to be hard for Valve to offer much of a push. Any attempt to give the Linux client preferential treatment will have Windows and Mac users crying foul. That turns the whole affair into a careful balancing act, with Valve trying to encourage Linux development without leaving its primary users feeling like second-class citizens. I think Carmack is ultimately right to be dubious on the possibility of a major Linux gaming push -- having better titles on Linux is fabulous, but building a compelling platform without alienating other users will be extremely difficult.