photonboy :
Somebodyspecial,
I highly doubt Steam is going to convert their entire game library to work on Linux. Not only would they likely not have the rights to do so, but it would be ridiculously expensive.
There appears to be a lot of confusion about SteamOS standalone versus Steam on Windows-PC's optimized for the living room.
I don't think they need to do 100%, what I'm saying is they need to get at least the ones most people want, and to some degree reverse the current 10% / 90% to more like 80% / 20% (where instead of 90% needing windows, maybe you're only missing 20% and have 80% on linux).
Why would anyone stop them from making a larger audience? There are some REAL windows haters out there. I'm guessing most companies would say "so you want to recode some of our game to make it linux compatible and kill Microsoft's DirectX stranglehold on us?...and either pay us or do it yourselves?...OK. We're in". If that doesn't work, I can see "and we'll only take 15% of your next two games if created on Linux first"...ROFL. Whatever wins the war. Many ways to skin a cat right?
Look at how many games are being converted to MacOSX (check gog.com for instance, or dare I say it, NFOHUMP, GOG is converting their whole library and very fast I might add). It isn't hard. You are not recreating all the assets, just the front end so to speak. Granted MS might block their own games (windows live games, etc that they created, they want consoles and windows to live, not Steamboxes), but I'm guessing anyone else would be fair game.
Think about what I said with money. Valve makes over 1Billion per year now. Most games are made for under 5 million (assets and all, not just frontends). So I doubt they'd pay more than 100K-1mil to convert each game. Since there are no shareholders/hedge funds etc to tell you what to do, valve could just break even for 3yrs and convert the entire library (again except for MS games who would rather block you). 3 Billion over 3yrs and you rule out windows. OF course it goes without saying you create everything from here on out with STEAMOS first and windows 2nd (or pay minimally to get companies to do it). Once the library is converted and companies see how many copies of SteamOS have been installed (I'm sure Valve will shout these numbers out as they increase), they will make SteamOS games first as it's easier to port a linux game to everywhere else (android etc) than to port DirectX games.
Microsoft has shareholders to answer to and can't go break even for a few years to kill someone today. Valve as a private company can do this. There is nobody to answer to but GABE himself really. Torchlight and Torchlight 2 were done for ~2 million (New Torment game, Shadows of Avatar, Wasteland 2 etc all being done for under 4mil this year/next). How much to port stuff like that? 100K-200K? You are only recoding a PATH, not recreating entire games/content. Unreal 3 was ported in 4 days to firefox by a very small team. I don't think it's near as expensive as you suggest when they can do that.
Ideally I'd only convert the 25%-35% that are most popular on steam for windows (work slowly on more after that), say 250mil to convert 250-500 each year (or however many you can for that cost). I'd go the route of funding 100-200 BRAND NEW games/year at 2-4mil (AAA types like torchlight 2 or better) so Windows NEVER gets them and people have a reason to install my OS. Half-Life 3 would be a good start! Make them all in OpenGL/HTML5/WebGL so they can be ported to anywhere else easily if you totally fail with SteamOS, but make them SteamOS only (no consoles, no windows) until you see you've failed in 5yrs or whatever...LOL. Note these will all be funded by Valve in this case so no problem coercing others to allow it being ported to windows if you failed. Worst case scenario is you make DirectX weaker and Linux and everything based on it stronger (which is valves goal at this point according to Gabe - Kill windows/console games - yes he hates consoles too).
Gabe is a long term thinker, so I'm guessing he'll be thinking 5-10yrs out, not if he'll win in a year with a few games. This is a WAR not a battle right? They already have 10% (200+ or so OS certified) and it isn't even out yet. I'd think all games made with opengl can be ported for a few hundred thousand tops (openGL already runs on everything with driver support). I'm guessing that covers about 1/3 or more of their library. Look how many are released for multi-platform (android, mac, windows) simultaneously these days. Those are opengl etc types easily converted. Avoid directX to start with and converting is cheap.
Or make the game with a popular engine that is easily portable:
"With its code written in C++, the Unreal Engine features a high degree of portability and is a tool used by many game developers today."
https://en.wikipedia.org/wiki/Unreal_Engine
"The current release is Unreal Engine 3, designed for Microsoft's DirectX 9 (for Windows and Xbox 360), DirectX 10 (for Windows Vista) and DirectX 11 (for Windows 7 and later); OpenGL for OS X, Linux, PlayStation 3, Wii U, iOS, Android; Stage 3D for Adobe Flash Player 11; and JavaScript/WebGL for HTML5."
4 days from Windows to Java/WebGL for firefox. Considering the engine itself took YEARS to make, 4 days for a few guys is pretty darn good right? More games are made with unreal engine than any other engine (by about 4x). Top games too, Borderlands2, Bioshock Infinite etc...
http://store.steampowered.com/browse/linux/
320 games already on linux now. Surely they all run steamos shorly as it is linux. Ouya doesn't have the funds to do their own crap (why did they close their store? No googleplay etc? Whatever - welcome to limited sales). Valve can get this done and has years of experience with linux/opengl, and as I mentioned the cash to spend yearly getting it done. You get the same support with Id's Engine also as Carmack is a full believer in OpenGL/Linux and even helps Linux with the Utah GLX project etc (linux driver project). So three big engines down counting Source engine
Valve only has to break even on the costs after converting to enhance their goal. Old games don't need to make money here, they just need them converted and break even right? That's what moves us from DirectX going forward, no need to make money on the old library, just get us off windows. They will be sold forever on steam, so likely very few will lose much money over the long haul and 100's of steam sales every year.
Convert now, kill windows in the end. It's a simple formula.
Gabe is worth 2Bil himself. These old library games are just like Nvidia's Shield project. While it may be a wild success in the end (shield rev4/rev5 or whatever may be selling millions as they'll be more powerful than consoles in under 5yrs, 1080p easily on 14nm), but they only need to break even on the 10mil they invested to make Shield to be laughing at how many people they convert to android gaming. Same story for valve. Anything that stops a windows user's gaming is great for them/linux in the end. I'm guessing shield breaks even at 100K units (easily had in a year) at $100 profit a pop (can't cost more than $200 to make it, Vita is $150 and 3DS is $101 to make BOM). Vita has sold 5mil and is far weaker power wise with a smaller library than android, which has all apps etc too for shield, not to mention output to TV etc. I'd think they'll sell 100K just on Nvidiots alone
I can wait for rev2 or rev3, but Nvidiots will buy rev1...LOL. I want a T5/kepler 13in+tablet first
I don't run linux currently, but can't wait for Valve to give me a reason to
Facebook
Twitter
Instagram