Is Microsoft Driving Valve and Steam to Strengthen Linux as a Gaming Platform?April 24, 2012 at 1:00 pm | Posted in Decentralism, Operating Systems, Technology | 8 Comments
If I combine a few news items to reach a natural conclusion, for the first time, it makes sense for a large business to fund Linux as a gaming platform.
1. L4D2 runs multithreaded at 250 – 300 fps on my Windows 7 machine. Single threaded it’s ~180 fps. On Linux it runs at 25 – 30 fps with the same NVidia GTX 580 video card (after getting them to fix some things), and around the same on a ATI 6970.
2. Good tools to track down OpenGL performance and help debug rendering issues just don’t exist.
We want to fix this. It would be nice to have people who have experience with Linux work on it.
Followers of Linux gaming have known that Valve had a few employees dabble in Linux before, but not make commitments or investments in it. So a CEO saying “We want to fix this.” Is a major change and improvement.
But why the change? Ah, this is where Microsoft is the danger to Valve. But I’ll back up to explain what Valve does to those who don’t know. Valve started as a video game developer. They saw that a game distribution platform would be useful, and since they couldn’t find anyone else willing to create one, they created their own.
This platform is called Steam and has turned out to be majorly popular and profitable. And unlike the love-hate relationship many have with Microsoft and Apple, users are pretty much in a love-love relationship. In other words, it is viewed as both a must have and good or great on prices.
Steam provides a centralized system to buy games digitally, receive patches and updates, interact with friends who play games, and find people to play multi-player games together.
Steam follows (more accurately, predates) the model of an App Store that is used prominently by Apple, Google, and Amazon for Iphones and Android devices. And unlike Apple at least, it doesn’t do hardly any ideological filtering to remove politically incorrect applications. This is also similar to the repository model of software distribution used by most Linux distros which is much older, but mostly for free without much paid software.
Microsoft’s Windows platform does not follow the restrictive App Store model, but as this model seems profitable, it wants into the format. The Android model is popular for its freedom… you get its benefits, but you can still install separately. The Apple model is much more restrictive, making it difficult to install outside of the App Store.
Microsoft, being a giant monopoly desiring company like Apple, would like to follow the Apple App Store model. It has made indications that it will push toward this model with Windows 8 and the Metro format.
This presents a problem for Valve’s Steam. It currently requires Microsoft Windows, or Apple OS X, which is also pushing the App Store model on the desktop as a future direction for profits. It is not hard to imagine Microsoft locking out or debilitating 3rd party digital distribution stores.
For comparison, think of Microsoft Windows as a government that creates the laws by which the entities (software in this case) are allowed to operate within Windows. But Microsoft does not act as a mere government, it also acts as a competing corporation within the governmental framework. Does this give it an unfair advantage? Absolutely. When Microsoft wanted to get into Office software, it went into competition with Wordperfect, a software that significantly helped make Microsoft’s operating systems valuable. But Microsoft could and did design Windows (the government) in such a way as to advantage MS Office and disadvantage Wordperfect.
There are many other examples like web browsers that could be used, but suffice it to say, such is the Microsoft pattern, and Apple is in no way better. This is bad news for Valve and Steam. If it stays dependent on Microsoft, it is likely to be marginalized to nothing within a decade.
That is where Linux offers an alternative. In the past till now, no large company had a clear and strong financial reason to invest in Linux as a gaming platform. It has developed into an adequate gaming platform, but not great. In order for Valve and Steam to offer an alternative through Linux, it must become graphically competitive.
The prime limitations include OpenGL, while in no means bad, has had its primary financial backing by companies that focus on non-gaming graphical computation, and so is not competitive to DirectX 11. Next is X.org or Wayland display driver. The first is growing archaic and the second is incomplete. These would need developers funded to include gaming performance as an important consideration.
Separately is the Wine program primarily directed by Codeweavers. Wine allows Windows programs and games to run on Linux. However, Codeweavers is a comparatively tiny company that can only improve wine slowly but surely. This is a central place that Valve would need to drastically increase the rate of improvement to provide an alternative for the many thousands of already existing Windows games that it already supports. While there may (I hope) come a time when game developers view Linux as a primary platform, I foresee no transition path without Wine figuring prominently. Nor do I see an intelligent path that does not partner with Codeweavers to rapidly expand Linux gaming.
Microsoft has hinted that Windows on ARM will be locked down to a Metro application store with likely no opportunity for Valve’s current business model for Steam. The potential future lock down on x86/x86-64 as well is a noose around Valve’s future on Windows. It is possible that ARM could continue its expansion against Intel and AMD processors and limit Valve in that direction even if x86 Windows never locks down 3rd party application stores. If so, then Wine on ARM may also be a future focus of Valve to secure its future destiny.
There has been recent discussion that Valve might create a “Steambox”, an alternative to the Playstation and Xbox consoles. While it acknowledged no such immediate plans, if it were to do so today, it would basically be a Windows PC already configured for gaming. If we were to go out maybe 5-10 years in the future, a Steambox could be a Linux based x86-64 or ARM PC that could run the majority of 20+ years of x86 Windows games, and most of those at high frame rates without glitches.
Someone may think that Valve may eventually target Android instead of standard Linux, but the two are compatible enough that converting between the two is far easier than the conversion from Windows. Wine and OpenGL are much bigger challenges necessary for a focus on either standard Linux or Android.
As a post script, Valve is further an interesting subject for this blog because they may be the most decentralized moderately large business in America. If you can believe it from their own statements, they have no management and effectively no job titles.
[POST SCRIPT EDIT: There is further confirmation that most of what I said is in the right direction. Valve CEO Gabe Newell is directly seeking work on Linux and OpenGL. Also note "his negativity towards Windows 8" seems to reaffirm the threat of a Windows lockdown. However, Wine doesn't seem to be figuring into their current equation. Software built directly on Linux without Wine is the best end-stage option if they focus primarily on future games, but I am equally interested in the back catalog of thousands of games.]