Why move your dependency from Microsoft over to Valve? Valve has innovated steam at a snails pace and has far fewer resources than Microsoft.
From an end-user's perspective, there's really no good reason. Windows works well enough, and while it has some shortcomings compared to Linux (and vice versa), moving to Linux isn't going to represent a fundamental shift in their experience.
From a developer's perspective, it's about survival.
Right now, Windows game developers are utterly dependent on Microsoft. Microsoft provides the libraries, the operating system, the development tools, everything. If Microsoft decides to take their software in a certain direction, game developers have no choice but to follow or exit the market. Up until Windows 7, this relationship has been been beneficial to developers, as Microsoft has served as a generally benevolent steward of the Windows platform by pushing the feature envelope while maintaining solid support and backwards compatibility, and developers have benefited from the stability that Windows has provided. However, Windows 8 has changed that relationship, as Microsoft is not only doggedly pushing Windows in a direction that consumers are rejecting, but they have also signaled a strong desire to move away from Windows' past open ecosystem to a walled garden with Microsoft as the sole gatekeeper.
In contrast to the Windows world, the building blocks of the Linux ecosystem are community controlled, and developers aren't beholden to a single entity. Also, since many of the fundamental components of a Linux desktop are licensed under the GNU GPL, the source code will forever be available to the community for as long as the community is willing to maintain it, which makes vendor lock-in impossible. Linux's open development environment, the ability to customize essentially every part of a Linux system, and the fact that you don't need a license to redistribute it are all big reasons why Linux the platform of choice for so many software "appliances" and embedded devices.
With respective to Valve, you could argue that the new boss (Valve) is the same as the old boss (Microsoft) with respect to locking in developers, and you would have a point. However, if a developer makes a Linux game and distributes it through Steam, moving to another distribution method would be technologically trivial. On the other hand, a developer would be absolutely screwed if they had a Windows game and needed to move away from Microsoft, as they would have to port the game to another platform, which they may not have the money, developer skill, or time to do.
1. A ridiculously large back catalog of content. I'm still able to play games from the Win9X days and XP days on the latest Windows versions with little to no difficulty. Steam is it's own worst enemy here because even though it's making a noble push into linux, the vast majority of it's content is for Windows and 5 years from now people will still maintain Windows as long as the content they purchase today stays functional. For many people, moving away from Windows now is like literally throwing away all the cash they've spent on Windows software since forever.
Linux can run Windows software via WINE. Granted, it's not a very good experience right now because most software developers don't support it, but it provides a viable path toward running older Windows software.
Before you laugh this off, consider the following scenario if Linux gaming does take off:
- New games are developed with Linux compatibility in mind, and will just work out of the box.
- Software distributors like Valve, GOG, etc. will have a strong interest in getting older Windows titles working, so they will have plenty of motivation to fund the development of WINE or a similar compatibility library. Legacy Windows titles would represent a static target, so compatibility would only improve as time goes on.
- Software developers will also have a strong interest in getting older Windows titles working, so they will have plenty of motivation to test their software with WINE or a similar compatibility library. Developers can package their software with known-good settings so that it works right out of the box, much like how developers package older DOS games today.
This is a hypothetical scenario, of course, but it serves to demonstrate that overcoming compatibility problems with Windows software is not as Herculean a task as it seems.
2. DirectX provides a full suite of capabilities for developers for sound, control, graphics, etc. Linux is decentralized. OpenGL is nice and can compete with Direct3D, but it is only one part of a larger equation. Microsoft gaming didn't appear in a vacumn and whether they get credit or not, MS via DirectX has done more for gaming stability than virtually any other platform. Maybe MS has been pushing Xbox more in recent years, but it is the latest Direct3D combined by attrition by numbers that pushes GPU evolution. In economics, it matters more where the money is than the virtue of any particular applicaiton.
Similar libraries exist for Linux, and Valve has hired a number of software engineers to continue developing those libraries. In this sense, Valve will serve the same role for Linux that Microsoft does for Windows.
I do believe linux will have a larger role in the future, but if anyone is expecting some sort of revolution I wouldn't hold my breath over it. There are many steps that have to be taken before Linux would be viable for the average mainstream user.
As has been mentioned before, Linux is already used by the mainstream user on many platforms. Tablets? Smartphones? Set-top boxes? Chances are it's either running Linux, some other type of UNIX clone.
The x86 desktop world is still dominated by Microsoft, but as has been clearly shown by Windows RT, Windows would not succeed on its own merits in today's market. The only thing keeping Windows relevant is inertia, and this can only go on for so long.
Look at today's market forces:
- Non-Windows platforms that use cross-platform libraries like OpenGL are becoming increasingly common, and developers are increasingly targeting multiple platforms to increase their potential market.
- Web applications are becoming more and more functional, and as rich web applications move away from Flash to HTML5, these applications are becoming inherently cross-platform.
- Tablets, smartphones, and other cheap computers are growing in market share, and as OEMs seek to decrease the cost of their products, Microsoft will be under increasing pressure to decrease the cost of Windows.
- As the market for full-priced Windows licenses shrinks, Microsoft will look for other means to generate revenue, which will provide a huge incentive to lock Windows applications into their App Store. If (perhaps when) this happens, developers will increasingly look to other platforms.
Microsoft is being pushed from multiple angles, and they are being pushed hard enough for them to stumble in the market after nearly 15 years of unquestioned dominance. While they undoubtably remain a giant in the x86 PC world, the market is in the process of tipping Microsoft over, and if Microsoft can't right themselves, they may wake up one day to find that Linux has reached the tipping point of mass acceptance.