Oh my god. This provides a genuine chance for year of the linux desktop in 2014, for gamers at least.
What about pretty much all larger businesses in the world? Windows sure dominate the corporate space.
When I buy a computer with an OS, I tend to pick Windows mainly because I intend to game on it. If I don't intend to game on it, then OS would stop being a major factor in my decision-making. I might spring for a Mac instead, or I might get a Linux box. It's no longer the point.
If I were buying a computer for surfing the web, I can get what I want. If I were buying it for word processing, I can get whatever. If I were buying it for photo manipulation, I can get whatever. For Door #1, I'd be looking at network connection options. For Door #2, I'd probably be looking for compactness. For Door #3, I'd want graphics and RAM, I think. For none of these is my primary consideration, "Well, the program I'm going to want to run isn't going to be available unless I get Windows."
That primary consideration exists for gaming. If I bought a replacement gaming rig today, I'd get a Windows box. Period.
Two years ago I would have said another big reason was Outlook, because no other desktop calendaring app comes close (even though it sucks hairy balls as an email client), but nowadays web-based solutions, particularly Google Calendar, are as good if not better.
Now TF2 runs on Linux, and the perspectives for Linux gaming are very good in general, but I no longer want to switch because I am happy with Windows 7, I hate Unity and Gnome 3 and their dumbed-down UIs that make me want to bang my head against the wall, XFCE is OK but feels like going ten years back in time, and KDE 4... well, that one is not too bad, but I still prefer Windows 7.
It's a pity that all this Linux gaming surge didn't happen back in 2008 or so, when Windows was in its lowest moments with Vista and most Linux distros hadn't yet been ruined by overbearing UI visionaries. I'm sure that desktop Linux market share would have soared in that case.
I would bet a money (albeit a small amount) that, to a first approximation, they're really only using the Linux kernel.
Even if SteamOS uses different toolkits or packaging (which it very well might), the amount of effort to get those titles to run on typical desktop linux distros should be relatively very minimal.
I highly doubt it would be more similar to Android than Linux Mint simply because they have already invested time and money into making Steam run on the Linux Mint family of distros (i.e Ubuntu), and (to my knowledge) none working on the android stack.
The rest is irrelevant; if a game runs on the SteamOS, it will run on my Linux desktop.
When we actually see it, it may be either a desktop-less distro, have Unity/GNOME or some other desktop permutation. My best estimate of what best fits a maximum impact strategy is an OS with a desktop environment pre-installed (even if hidden from casual view) or easily installable.
I have a feeling that we may look back at this seemingly minor product announcement as a significant stepping stone.
Maybe with added Wine.
Nothing in the release said steam only or no desktop, and they call out the ability to create content on the platform specifically Users can alter or replace any part of the software or hardware they want. Gamers are empowered to join in the creation of the games they love. Rather sounds like what they are doing is going to be open source and allow non games to run. Even if they don't ship with KDE or Gnome by default it should be possible to get something up and running.