How could a company justify using it as smartphone OS, where its penetration was exactly zero?
The fact is that Linux has gone from nothing to dominating in several huge markets:
* Smartphones
* Internet infrastructure
* TV/cable/pvr boxes
* WIFI/Routers for consumers

The reason Valve can take linux and make it work is that most of the needed tools and drivers are already in place. Only small changes are needed to deliver a working product, so time to market is only a fraction of what it would be if they started from scratch and of course incredibly much cheaper.

Yes but they have to convince publishers it will be a viable option and that there is a large enough market out there (imho the size of the market and its takeup will cause this to succeed or fail)...
Valve cant just point to their Steam userbase (only a small % are probably linux users i imagine)

Can anyone here recommend (hypothetically) a Linux distro which is good for newcomers to Linux yet very open?

Also does anyone here run Linux on a hard-drive partition while also having Windows on a hard-drive? Are their problems that can emerge from doing this?

There are many. Ubuntu (with all its variations), Mint, openSuSE are very easy ones.

You can easily dual boot Linux and Windows. There's really no limit how many you can install. The install program for the easier distros automatically detects all operating systems and creates a boot menu. All you need is some free space, preferably unallocated, to install. I think the installers can also resize your NTFS partitions but that's always risky business, so make backups if you need to do that.

Try Linux Mint, it will probably be easy for a traditional Windows user to get into. You can use a "live CD" to try it without installing anything.

There shouldn't be issues with installing on a partition with Windows but you do have to be careful. As you may know, partitioning a hard drive isn't without risk so back up your data. The safest thing would be to shrink the Windows partition from inside windows itself. See the first three steps of this tutorial. Then be very careful when installing Linux to make sure you select the right partition.

Some months back they published numbers showing that there were more people running Steam for Linux than running Steam for Mac OS. There are currently far more games available for Mac than for Linux and yet if the statistics are correct - there are for more Linux users running Steam. The numbers were like 1.5% of the userbase or close to it if memory serves correct. Think about that for a minute. :)

With that kind of volume of potential customer interest, it totally makes sense for them to try and expand into the Linux platform outright even without their own OS based on LInux. However there is no one Linux distribution that is definitive out there with which to target as a primary platform or use as a standard base, and so by building their own OS they can create that standard base themselves and update it themselves to suit their own needs, passing along bug fixes and other contributions to the wider Linux community to pick up and roll with on their leisure.

At the same time it gives them the long term opportunity to free themselves from the Microsoft shackles they're currently stuck with.

As for convincing publishers, they port games to platforms that have a large enough userbase which they can estimate they'd see a return on investment for supporting. Some will likely find the existing threshold sufficient to do so - as evidenced by the existing titles available for Linux and the ones that are in varying degrees of being ported already and other general interest that has been stated publicly. Others may decide that they perceive the effort to port their specific game(s) are not worth the risk to reward ratio and wont bother. Still others will choose to wait and see what happens on the platform and defer their decision for the future.

This isn't a simple black and white thing really and wont be a case of seeing all games or no games appear for Linux. It is a serious venture being undertaken, and will be a gradual introduction of more and more games on Linux as the SteamOS platform actually becomes available, and publishers will base their decisions on whether to support it or not based on their perceived estimate of the amount of work involved versus their estimates of the market interest for doing so. It will go both ways, with some choosing to support it and others choosing not to. As the platform grows in popularity and becomes successful (presuming it does, which I do personally), we'll see more publishers take up the platform.

This isn't really much different than Playstation 4 coming out and being a new platform that is completely incompatible with Playstation 3, or Xbox One being a totally new platform incompatible with Xbox 360. These new systems have to prove themselves in the market every time a new one comes out. Valve will too, but what Valve has in their favour is that it is just another PC at heart, plus when you buy a game for Linux - you're really buying the game and not the game for Linux. So you could just as well install the Windows version in Windows running on your PC, or on Mac if you have a Mac and it's supported also. There probably are not going to be any Linux exclusive titles, at least there wont be any from Valve (they've stated this publicly). So in the worst possible case - if this completely blows up in Valve's face and becomes a miserable failure - which seems very unlikely, people still own game licenses for games that are available for existing platforms, namely Windows for sure, and for some titles also Mac support.

If Playstation fails, your game purchases wont run or be transferred to any other system. Ditto with Xbox etc.

As for convincing publishers though, they're already doing just fine at that with more and more publishers announcing Linux support all the time. If one pays attention, all you see out there is more and more pro-Linux announcements of support happening. There's no reason at this point to believe that this wont continue.

Why do Linux fans keep going back to this?.
I still see this is a might gamble and if it blows up in their face.... well. I think you guys are being a tad overconfident of its success...
Valve should just churn out there own console and be done with.

That's basically what they're doing. However they can't make the hardware themselves so they're outsourcing that. And they can't make the low-level OS stuff themselves so they're outsourcing that. That just leaves them to make the higher level "Steam OS" stuff and design a controller.

Actually, Valve has publicly stated that they are not a fan of Microsoft Windows and where it is going and they're looking at other options, I'm just paraphrasing and repeating what Valve themselves have stated. While I am a Linux user and developer, that is orthagonal to the facts I've presented about Valve's motives.

As for being overconfident of its success that will remain to be seen of course. I will say that in 1994 when I began to use Linux and open source software, a light bulb went off in my mind. It was one of those eureka "ahhhh!" moments where I "got it" so to speak. By that I mean I really "got" the idea of open source software and what it had the power to achieve in the future based on the grass roots nature of the phenomenon. The majority of people at the time, and perhaps even the majority of people still today do not "get" open source software including but not limited to Linux and in their lack of understanding on the subject they predict that it will fail miserably, that it will never catch on, that nobody would ever want to use that for business purposes, that it is only a "hacker's system" and all manner of other statements based out of lack of understanding of the platform and the phenomenon of open source and its benefits.

Nonetheless, it did what I myself wanted it to do at the time and so I used it and found it to be a great solution for my own needs, and I could see that over time by its very nature it would only spread further and further and meet more people's needs as well. I don't think that it will ever solve all problems, nor that it is the best solution for all possible problems, but it is a great solution in general and the areas where it shines will only continue to grow going forward. What Linux/OSS doesn't do very well in 2013 it will do better in 2014 or 2018 or whenever someone sees the benefits of the system and paradigm as a value proposition that makes sense to them to use it as a basis for what they too would like to do - and in doing so, that generally involves contributing back things to the ecosystem as a whole for everyone else to benefit from.

This does not mean that every possible attempt a company or project out there makes at using Linux for something will become an automatic success of course, and Loki Games is an example ultimately of failure to do so. There are no doubt various other examples of failures out there of companies attempting to use Linux for things unsuccessfully. There is no guarantee that using Linux will automatically result in success.

But, things are a lot different now in 2013 in Linux/OSS space than they were in 2001, and Valve is a much larger company attempting to do something like this with huge pockets. If anyone stands a remote chance at pulling it off, it's going to be someone with big pockets that can withstand the blow should the effort turn out unsuccessful. I think Valve has the amount of resources to pull it off, and from what I've seen in the press and interviews they also "get it" and are approaching this in a way that I believe maximizes any chance they've got for success, by interfacing directly with the community, making contributions, joining the Linux Foundation, and various other things they've done or said.

While there is no way to know if the end result will be successful, I think someone infinitely familiar with the Linux community and history, and familiar with Linux/OSS process and culture is more adept to make predictions as to whether something like this stands a hope in hell of success compared to someone from the outside looking in who may or may not truly understand what it's all about and how it works.

Why do I think it'll succeed? Because they're making all the right decisions so far if it has any chance whatsoever to do so, and they're going into it highly-informed about what they're getting into.

There's a good chance MS will go full retard with Windows and try and make it a closed platform like iOS, yes this is very risky and could blow up in their faces if their traditional and business users abandon them because they are accustomed to Windows being open, BUT MS obviously keeps looking at the high profits in the tablet market and the profits from Apple's iTunes store and they want a piece of that pie, or rather they want the whole fucking pie.

This is why Windows 8 had that horrible Metro touch-screen oriented interface, why it got rid of the Start Menu, and I guess there's support in MS management for giving the finger to desktop users and going for tablets full scale.

That's why it's seen as good for Linux to be built up as a viable alternative, for more porting of games to Linux, for games to be developed with native Linux versions.

Of course there's always the chance MS will backpedal and basically go back to traditional Windows, it could go either way.

I'd say the safest thing would be first to try Linux is a virtual machine using something like VirtualBox. Just to see how it looks like. And then rather than messing with the existing disk, it's better to just plug in another hard drive and install it there (of course if you have a spare one and it's not a laptop or some other setup which doesn't allow it).

Sure using a virtual machine would be safest. However it's almost as safe to use a LiveDVD (or USB) and you don't have to mess around setting up a VM (which might be an issue if you've never done it before). It also doesn't require extra resources like a VM does (which may be an issue on lower spec machines).

Obviously installing to a separate hard drive would be safer then re-partitioning an existing one but I got the impression Crosmando knew that. Still I should probably have explicitly stated that.

You're making the mistake of thinking that Valve is switching to an entirely new and untested platform and abandoning all other platform in an all eggs in one basket switcheroo. In reality, Valve is adding a new platform to their matrix of offerings. They're not abandoning Windows or Mac or anything else that they're doing currently, but accentuating it with something new. So they /are/ staying on well established platforms, but simultaneously providing another option to both gamers and themselves. This is an expansion, not a swap out.

So what you're suggesting, is that if Valve puts out a new Linux offering as an addition to what they do right now, they are putting all of their eggs in one basket and it is risky and likely to fail, the world will come to an end for them, etc., however you would like to see GOG do the same thing essentially however if GOG does it, it is diversifying by not putting all eggs in one basket? :)

Seriously, apply what you just said you think is a good idea for GOG - to Valve. Valve is a much bigger company with more money to throw around and experiment, and a lot more manpower to throw at the various problems. GOG would be wise to wait and see what happens than to assume all of that effort and all of the risk themselves in my opinion, and that's most likely what they'll do too. If and when Valve's efforts show signs of paying off - and I'm sure they will, lots of other companies will be watching carefully and likely following in their footsteps, and many of them - probably GOG included will have already started doing things behind the scenes long before they ever announce it publicly. What's more though too, is that any game developers out there who have their games on both Steam and GOG and are porting the game to Linux already - GOG is likely to get the port for themselves as well once they decide to also support the platform. Valve has a *LOT* more muscle to convince game devs to port their games to Linux than GOG does, and GOG can only benefit from them doing so eventually one way or another.

If I were GOG or some other company in gaming right now, I would be monitoring what's going on with Linux and I'd probably be silently doing things internally with Linux to prepare for the future, and that would likely include porting stuff, making changes to my website code for it, figuring out how to handle tech support etc. but I would most likely be waiting to see whether Valve succeeds at what they're doing or not and sort of ride on the coat tails of that. Lets face it, if Valve's SteamOS/Steam machines actually _fail_, the likelihood of GOG or some other distributor doing similar and succeeding is rather slim and unlikely - so why would the smaller guys go ahead and take such a big risk when they'd have a lot more to lose from it and have a lot less resources?

Yes, GOG should be paying attention to LInux right now but:

1) They almost certainly are to some extent or another (behind the scenes with lips sealed)
2) They should go about it in a low risk way and ride the coat tails of giants rather than dumping significant resources into an unknown. Let the big guys do that and assume the risk.

We don't need another Loki. :)

No, both those are working perfectly well for touch-screens, and if you have one and get used to it, they work much better and faster then the old system. I also found it more unitive, from a HCI perspective, I would say that this was a nice step forward. These are not the points of concern at all.

What Valve and others are worried about is not this at all, but the Windows App Store and all that comes with it.