Linux games have always been one of the goals of free software. If game developers could only be persuaded to develop for Linux, the daydream goes, the operating system would start to gain serious market share. The last few years have lent hope to the dream, but the progress remains slow -- so slow, in fact, that its realization is starting to look questionable.
The first large scale effort to sell Linux games commercially was Loki Software, which ported games like Civilization and Railway Tycoon around the turn of the millennium. It quickly failed financially, leaving Linux gaming largely to minor free-license games like Pysol and Tux Racer, and to efforts to run Windows games using WINE.
Today, online games sometimes support Linux, but the dream of native Linux games remains. Fueled by months of announcements and speculation, the dream received a huge leap forward in January 2012, when the Valve Corporation officially released support for Linux on Steam, its online game distribution system. Six months later, Valve announced the Steam Machine, the specs for several lines of gaming computers running Linux. Meanwhile, gaming companies that previously ignored Linux began experimenting with it.
On social media sites, these developments received the attention that you might expect from the first alien contact. This time, gamers insisted, things would be different. This time, Linux would come into its own as a gaming platform, after which anything seemed possible.
The Cracks Starting to Show
This current version of the dream has lasted twenty-one months -- not bad, for incarnations of this particular dream. However, it is starting to look a little insubstantial.
One sobering aspect -- although not necessarily a permanent setback -- is the delay of the Steam Machine. A dozen manufacturers are working on their versions of the Steam Machine with Valve's blessing, and Alienware has even exhibited a prototype.
However, while a few beta machines may go out by the end of 2014, none of the manufacturers are likely to have a commercial product until well into 2015. Like most new hardware from new manufacturers, the Steam Machine seems to be suffering from a combination of over-optimism and development and production delays, which raises the possibility that many versions may never come to market.
More concretely, the kindest interpretation of the Steam site's stats is that development of the Linux gaming market is a long term strategy, more likely to be a matter of a decade than a year or two.
Consider: in February 2013, the optional Steam Hardware and Software Survey showed that just over 2% of its users were running Linux. Apologists noted that Linux support was new, and insisted it could only increase, yet for over a year, the general trend has been downwards, until in August 2014, the percentage of Linux users was 1.06%.
Steam is a popular site, with a broad selection of games, so the reduction is unlikely to have much to do with its faults. Nor are its figures completely unusual.
In January 2014,GamingonLinux.com, listed sales and comments from ten gaming companies, most of which relied on Steam for much of their income. In most of the ten companies listed, OS X sales were 2-5 times higher than Linux sales, and in no case did Linux games outsell OS X games. Nor did any company list Linux sales as more than six percent. In fact, in most cases, Linux sales were under three percent.
The one encouraging part of the GamingonLinux.com statistics is that a number of the developers listed expressed continued support for Linux. Most of this support was based on hopes for Valve and Steam, although support did seem to vary depending on awareness of cross-platform libraries and Linux development methods.
However, Puppy Games, one of the companies that responded to GamingonLinux.com's survey, has recently expressed some reservations on Twitter about its support for Linux.
Softpedia noted that Puppy Games has been expressing reservations about continuing to develop for Linux. One tweet from Puppy Games revealed that "To date we've made just **$12,000** from Linux games in total for all time for all four of our games! This will not do!" This figure, apparently, was 1% of Puppy Games' total profits, although later tweets revealed that this figure was only for sales through Steam.
In another tweet, Puppy Games wondered "if other devs have seen similarly paltry sums and whether considering continued economic relevance of Linux."
No companies responded to Puppy Games, but by now many must be aware that Linux is not the untapped market that commentators once imagined. At best, the potential has been overly anticipated. The gaming companies may not be rushing through the door, but some do seem to eyeing the exit.
Waiting on the Bubble
As I write, commercial Linux gaming is a bubble, consisting more of speculation than realized potential. It depends almost entirely on Valve carrying on its support, and at least one version of the Steam Machine being successful.
This might sound ominous, except for the fact that Valve seems to be taking a long-term approach. Valve has emphasized repeatedly that it is interested in an open alternative to Windows and OS X, and this interest might even survive the failure of Valve's current plans.
However, the success of Linux gaming does not depend on Valve alone. The slow development of Linux gaming suggests that the development companies have failed to understand their market. Perhaps they have underestimated the size of the potential market, or its possible appeal to Windows users who are already using what -- to them -- is a perfectly adequate operating system for gaming.
In particular, they may lack insight in the psychology of Linux users. The first figures for Linux users on Steam suggest that the potential audience was curious, but was unimpressed by what they found -- almost half did not become regular users. Perhaps they were disinclined to pay for games, or disliked digital restrictions management, although these complaints have largely gone unvoiced. Or perhaps Linux users are simply disinclined to participate in manufacturer's polls. However, whatever combination of factors is involved, it seems that you cannot sell Linux games the way you sell Windows games.
Yet another possibility is that commercial Linux games are happening too late. Although Steam can offer the popular games, a small but growing number of free-licensed games like 0 AD and Unknown Horizons are starting to emerge, many of them featuring contemporary graphics and game play.
True, these games seem to have small audiences, but they are more in keeping with the free software ethos than any commercial game. Give them a few years – commercial games may have trouble competing with them.
To say that commercial Linux games are a failed concept would be premature. However, we can now say with certainty that their success depends on that of the Steam Machine, and, perhaps, the manufacturers re-examining their approach -- all of which should make for an exciting next couple of years.