Let's break it down
> What is the point of a console?
Very predictable hardware specs & setup.
> What is the point of the console being Linux-based?
Reduction of brand-new OS development costs; potentially zero need for writing drivers depending on hardware setup; very extensive presence of compilers / interpreters / libraries that can be freely usable out of the box.
SteamOS succeeded in forcing Microsoft to clear up its act, and this is the main reason for Valve's investment. Their continued investment is to continue to make sure that they have a plan B. They are doing a lot of behind-the-scenes work to make the Linux gaming stack amazing. This looks "failed" from the outside because they aren't pushing the platform heavily, but it is actually a lot more beneficial that doing a big marketing push before the platform is ready.
The interesting thing with a platform like Linux is that it's not dependent on one big revenue stream to survive. Because of this, it doesn't matter if it takes years to get massive adoption because it's not burning a hole in the pocket of a company. If xbox was in the same position linux was in, you would be correct in saying that it failed, because anything less than 30% market share means gigantic losses for Microsoft which would lead to axing xbox.
PS: Atari's games run fine on Linux. Their retro emulator runs on Linux. They seem to know what they're doing.
[1] http://store.steampowered.com/hwsurvey/?platform=combined
Valve doesn't really care how many users are running Linux right now —they still have their Windows revenue. What they do care about is how many users would switch to Linux if Valve was for some reason forced to drop Windows support altogether (like, Microsoft goes Apple on their users and mandates an app store). That number may be much higher.
You'd also have to keep in mind what's keeping users to Windows. From where I stand, exclusive titles play a big role. Linux has no exclusive title to speak of. Why would it, that market is way too small for such madness. Windows, understandably has loads. My only reason for still using Windows is a couple titles I can't bring myself to renounce.
Then there are drivers and tooling, but that's just the same kind of network effects, really.
The very fact that Asus and others never really bothered that much to sell Steam Machines, shows how much market relevant it was.
> “A few weeks after this post went out, some very senior developers from Microsoft came by for a discrete visit. They loved our post, because it lit a fire underneath Microsoft's executives to get their act together and keep supporting Direct3D development. (Remember, at this point it was years since the last DirectX SDK release. The DirectX team was on life support.) Linux is obviously extremely influential." [1]
And again, market relevance was not required for SteamOS to succeed. Gabe Newell has been very open about the fact that SteamOS' goal was to keep Microsoft in check and have a plan B for when things go south..
[1] http://gadgets.ndtv.com/games/news/steam-s-linux-and-opengl-...
https://www.pcworld.com/article/3045249/linux/linux-gaming-i...
Sadly every piece of software and hardware since win95 has been made with windows in mind and no alternative has been able to come even close to that.
But they've gone a little too far with Windows 10 - for me at least, of course everyone will have their own "measure" where too much will be too much.
Until Microsoft do that, it's not actually very necessary. But it might become necessary at short notice.
If you didn't, what makes game development more difficult on Linux¹, compared to the alternatives²?
[1]: More specifically, mainstream GNU/Linux distributions [2]: Windows and MacOS
Another is that it's just generally a lot more heterogeneous than windows and mac (lots of different distros with variously up to date libraries, drivers, etc).
The same is true on a new install of Windows, though. First stop when the OS is running: Non-Microsoft browser. Second stop: Graphics card vendor site, to replace whichever outdated driver Windows shipped it for the hardware.
I mean, not as bad, usually (outdated official drivers on Windows, versus often-inferior unofficial/Open drivers on Linux).
The game developers culture is based on making money of IP, outsourcing game development tasks, not about sharing the code and making the world a better place.
On commercial platforms, the GPU and OS vendors even fly tech support to AAA studios to work around driver issues.
Developers developers developers as Ballmer once said.
^^ this is something no other game console has.
Just there is not a lot of Linux Desktop users though it appears we are around 3% now.
It depends heavily on the engine being used or if custom engine on how much the devs have relied on stuff designed to wall your product into windows.
There's like a couple of games out of ~100 I got on Steam that I've had big problems with in Ubuntu. A couple others just needed launch options to prevent the game from guessing settings so wrong I can't even fix those in-game and launch a launcher program where I can tweak those before starting the game.
To solve that, I had to build a dedicated living room PC. I didn't go nuts, so it ended up being about $500.
If I could replace that with an handsome Linux powered gaming system for less than $300, then I think it's a no-brainer... especially if it can run third-party marketplaces and apps like Steam or Plex.
If just games/movies/music then what was the logic for getting that over a console? (PS4, XBox1, Switch are around the same price point or below) Just curious, not suggesting that a console would have been better.
This seems like a hard problem, esp. since OP is so emphatic that $300 is preferrable to $500. Let me see if I can figgure this out for us both.
1 child x $500 = $500
1-or-more x $500 = $500-or-more
Oh! That must be it! This solution would be more expensive to implement, and therefore, since cost is a factor, this solution would be less preferable!
Glad I took the time to work this one out, I certainly was very confused as well.
Where it fell short was the overall experience. Most games are of course not designed for this. In PC you always have the keyboard, there are things that are tweaked through Windows. Also it felt that I had to regularly run to the PC to sort out some issue. Disclaimer: I tested just with few games and at least 6 months ago.
- https://en.wikipedia.org/wiki/PlayStation_4_system_software
- https://en.wikipedia.org/wiki/Nintendo_Switch_system_softwar...
- https://en.wikipedia.org/wiki/Xbox_One_system_software
And add to the mix that Playstation and XBox actually are using PC hardware...
http://switchbrew.org/index.php?title=Sockets_services
http://switchbrew.org/index.php?title=Services_API#Service_L...
I feel this is a significant part of the appeal.
"Traditional" consoles used lock-in not because it was good to do, but because they were either too weak and required very close to the hardware APIs (and no portable APIs like that existed), or because vendors were just lock-in inclined.
Atari know that - they tried it with the Jaguar
dude, we're riding the Vulkan wave now!
So not sure if it would generate any considerable profit.
Traditional consoles had fixed hardware which could be exploited through vendor specific APIs. If this is just going to run games programmed with OpenGL, this is not different from any other Linux PC.
Given the introduction of APIs like Vulkan which bings the programmability of PCs even closer to consoles, I don't see a use case.
The only market I can think of the niche of Atari fans.