Valve isn't "acknowleding" that Linux is a viable gaming platform quite so much as "really really hoping" that it is. Valve is simply trying every last possible thing it can do to avoid the eventual irrelevancy of Steam in a post-Window-8 PC marketplace. Sure, it'll be years and years most likely before Steam ceases to be the primary distribution platform for PC games, but the day is now known to be coming for sure. Heck, given that most gamers will be skipping Windows 8 and waiting for the next "desktop edition" Microsoft OS, Steam probably has another 5+ years of domination on the PC front, but Gabe's cashflow starts to look pretty gloomy after that.
Also remember that Valve is already losing some ground as the bigger, higher-volume, higher-sales companies (Activision-Blizzard and EA) are avoiding Steam more and more and moving towards their own distribution channels (which are also threatened by the Windows Store, of course), in addition to the cannibalization of the "real games" market by cheap iOS games in the current recession. They're already feeling the pressure of competing app stores, and Windows 8 isn't even out in most consumers' hands yet.
Steam fading away is not at all a bad thing, either. For all the doom and gloom of the walled ecosystems of Apple/Microsoft and their respective stores, Steam is a far worse walled garden with worse deals for developers and less choice for consumers. Steam Greenlight was created as a half-ass attempt to support smaller developers who can and do thrive on Apple's ecosystem, and even then it doesn't support anywhere near the same level of openness and freedom that Apple (and soon Microsoft) will provide. Valve doesn't just give you a nice cut-and-dried checklist and straight up fee to be on Steam. You must be personally approved by the Steam gatekeepers (_after_ paying a fee just to try!). You can't even get access to the Steamworks API to do your integration work until after approval. You have no idea what fees you'll be paying, since they can and often are renegotiated on a case-by-case basis. It's a hell of a lot better than the consoles (whose current practices and the recent rise of the small to mid game and indies have more or less destroyed the console market in the last 12 months), but Steam is still a worse deal for just about everyone than the Apple/Microsoft app stores. The only selling points of Steam are that is has the community API (Microsoft and Apple both have their, but Games for Windows Live is a piece of crap and Game Center is lacking in features) and that Steam gives the multiplatform support for every purchase (which is a very nice and appreciated gesture by folks like us, but 99% of consumers currently only one PC-class platform and don't give a shit about the others).
Steam is great because it provided a common platform and distribution channel for PC games, ushering in the digital download and service age for games when the consoles were (and they still are) totally fucking up in that space. The app stores, however, provide 90% of what Steam does, provide it at a cheaper price point and with an easier entry path to developers, and offer better integration and a mostly better user experience for consumers. Steam on Linux is a last-ditch attempt to stay relevant; the best thing that could come out of it would be a possible SteamBox kind of console (which itself may still not work out; all depends on how Sony and Microsoft react and if their console divisions get their heads out of their asses and build console platforms/ecosystems designed around the post-2010 gaming market).
If Linux wants to succeed as a gaming platform, or a desktop OS, or even a tablet OS, it needs to stop relying on big mega companies with ulterior motives to actually do it for them. Google is not making Linux a success on phones or tablets; it's making its largely proprietary OS-with-a-Linux-kernel a success. Valve is not trying to save Linux, it's trying to save Steam. Ubuntu isn't pushing Linux as a desktop success, it's pushing Linux as The Astronaut's personal toybox. If Linux is going to succeed, the people working on it need to actually make a damn user experience that competes with modern proprietary offerings rather than Win98/XP, it needs to make a software ecosystem that is user-oriented rather than (or better yet, "as well as") developer-oriented, and it needs to market a unified developer platform rather than a toolbox to build a bazillion customized ever-so-slightly-incompatible platforms. Steam isn't going to turn Linux into a mainstream gaming OS, just like Ubuntu bug #1 is absolutely no closer to being closed today than it was when Warty Warthog was released, and just like Google has done nothing to make Linux and FOSS (rather than Android and Google) a household name.
I completely agree with you and I think they way you expressed it is admirable. My perspective on GNU/Linux is rather irrelevant when it comes to gaming, because I really don't care about this business very much.
Still, I am interested in expressing my opinion here:
GNU/Linux was, in my opinion, never intended to be a "mainstream" OS and I wonder, why everyone is struggling that much to forcefully aim at this goal. As a result, GNU/Linux in many cases has fallen for corporate ideals (as you perfectly stated) and is actually being prevented from a "healthy" development.
Of course, we do have a solid user-base and developer-community, support from scientific institutions and the corporate sector, but clearly, the product of these huge efforts leave a lot to be desired, as these are divided among themselves.
The goal of creating a viable, independent GNU/Linux distribution has not yet been reached. But is it really necessary to aim at that? To really answer this question, you have to ask yourself, how "mainstream" is defined in our society. Due to the social effects of a mostly capitalistically oriented society, corporations are mostly the ones defining new values. Projects like GNU/Linux, with its ideals and democracy, might be successful on the paper, but in reality, it is not.
With the corporate sector defining new trends and society mostly aligning to them, the homogenous crowd of non-tech-savvy people defining a computer by how well they can manage their social relationships and tasks or the biggest part of youth just caring about computer games produced by big corporations is not really interested into the ideals Linux stands for. To cite Immanuel Kant in the context of his definition of enlightenment, we get to the point: "Enlightenment is the escape of human kind from his self-incurred immaturity". If the masses align to "wrong" ideals induced by the corporate sector, it is too easy to say it was their own fault doing so. The big mistake we are currently making is trying to actually adapt to the current trends and losing focus on what GNU/Linux is really about. I don't want to doom every company for influencing the development of Free Software, but in the struggle of profit, growth and existence, they do what they are obliged to do ignoring the ideals we all stand and should fight for.
What gives me confidence is the fact that our ideals and contributions to Free Software will endure, contrary to most companies which are of current relevance in the ephemeral market. History will teach us in the future as it commonly did in the past.
If Linux wants to succeed as a gaming platform, or a desktop OS, or even a tablet OS, it needs to stop relying on big mega companies with ulterior motives to actually do it for them. Google is not making Linux a success on phones or tablets; it's making its largely proprietary OS-with-a-Linux-kernel a success. Valve is not trying to save Linux, it's trying to save Steam. Ubuntu isn't pushing Linux as a desktop success, it's pushing Linux as The Astronaut's personal toybox. If Linux is going to succeed, the people working on it need to actually make a damn user experience that competes with modern proprietary offerings rather than Win98/XP, it needs to make a software ecosystem that is user-oriented rather than (or better yet, "as well as") developer-oriented, and it needs to market a unified developer platform rather than a toolbox to build a bazillion customized ever-so-slightly-incompatible platforms.
First of all, linux is not one person/corporation/whatever with unified goals.
Second, having companies advance free software because of ulterior motives is simply great. Things working as intended. If a company wants to scratch their itch, how is that any worse than a single person? Because they want to profit from it, hm?
So a company wants to save their outdated business model by improving linux. Why in the world should we, other linux users, give a damn why they're doing it? We get a better kernel, better drivers, better integration out of it. New software perhaps, in addition to improvements to existing sw.
It doesn't matter why they do it. It matters that they do it. It could be easter bunny telling a CEO on shrooms somewhere to do it, it just doesn't matter.
the truth is dat idsoftware is no more than a mere shadow of itself since doom 3. and about j.carmack... lol the might has fallen a long time ago.
Actually, I think Doom3 is where they lost it. Personally, I also skipped on Quake3, but that's because I don't like MP only games. But I know many enjoyed Q3, so I hold it as the last influential title from id.
Curiously, while Doom3 is the first id title that didn't impress, it also marked the beginning of developers stopping to license game engines from id; source and unreal engine have been much more successful ever since. There were hundreds of games developed on top of quake2/quake3 engines, but only a handful using later incarnations. Maybe id had more success in the mobile space, I haven't been paying attention.
Well you can play Quake 3 now as Quake Live. Of couse the MP part is what makes you play a game for several days/months even years and not SP. Doom 3 had only minimal MP, that's why it was not that popular. Quake 4 however had a really nice campain, not so dark as Doom 3, also a bit simpler. Really a pity that the Quake 4 source is not available to build 64 bit binaries. 32 bit binaries really suck, especially when you use pulseaudio for hdmi - with Q4 the sound stops after a while (that does not happen with D3 however). Doom 3 BFG native would be cool as well, compared to the original game it is translated now, for non-english speakers this is a real plus. I don't think that the gfx look really better, just more "blurred". Another nice game with doom 3 engine was Prey. All you had to do was to search the way (or use a walkthru) - you did not need to watch much out that you don't die, you always respawned. Basically you could play D3 BFG with wine as well, just like Rage - speed would not be much different to native, but of course wine is a 32 bit binary - all negative things apply to that Also fglrx + Rage was no stable combination, even 12-9 beta crashes soon. I finished Rage with a nvidia 8800 gts 512 via wine.
... Doom 3 had only minimal MP, that's why it was not that popular...
Neah, it was not popular because it departed from its roots. All that shadow with a monster suddenly growling was repeated ad nauseum. Also, because the engine was very demanding for its time, you rarely saw 5 monsters on screen at any given time. To top it off, I could kill at least some monsters faster with the flashlight (forgot to switch weapons in a hurry, you couldn't both see and shoot in Doom3) than using the shotgun. It was really, really poor as a game, but a nice tech demo with its lighting model.
Up to Doom 3 the engine was licenced, so it was basically a game to show what the engine is capable of. Don't you think that Quake 4 was better?
I never bothered with anything from id after Doom3. I think I tried a demo of ET:QW and thought wtf?
Also, Doom3 might have sold a lot of copies, after all there were a bit more PCs out there in in 2004 when it was released than they were back in 1993. I never fall for that marketing trick. And I suspect many copies were sold purely because of its predecessors. But the thing is, it was the first title that didn't make an impact at an industry level. Doom, Doom2, Quake, Quake2, Quake3 engines were used by many other games and heavily modified. Doom3 engine? I can't name a single title that used its engine.