Juho, great call-out with the "Steam-related carrot", I hadn't really thought down that path but you are absolutely right. Valve could easily offer big discounts in their split with the publishers for the first few years for Linux sales.
My (possibly very incorrect) assumption was that the profit from the Linux sales were so low as to not matter, but maybe they can provide a catalog-wide carrot... something more significant (as you suggested) to motivate the publisher en-masse.
This is likely to be coupled with some manner of hardware launch, as well. I am skeptical that the name "Linux" will be a particularly useful term in that context.
Couple that with my skepticism that they're actually shipping much if any traditional Linux userspace software, and it's even less useful to compare potential SteamOS users to traditional Linux users.
Other than that I agree that with Valve as a patron there's a lot more credibility. Steam was a long-term move, and it paid off. So Valve has a good track record and solid momentum. If they can help make it easy for game companies to port their work, taking out a significant or non-trivial amount of work and offer the carrot of sales & paying customers, I could see this taking off.
I was skeptical of Steam back in 2004 but I wasn't thinking long term enough. Launching your own operating system isn't a short term move. It suggests to me that this will only get better and more impressive over time, if Valve is as committed to this as it seems.
"Traditional Linux userspace software" is an informal definition, but let's say something like X. Or your average window manager. Or, say, almost anything from GNOME.
"Traditional Linux users" is another informal definition, but if you use a distro, you're probably a traditional Linux user. Contrast this with an Android user. They run the Linux kernel but the software they interface with is an entirely different breed. The statement "Android is Linux" is true, definitely, but it's inside baseball to anyone but us nerds. The Linux-y aspects are more or less invisible (by design).
I'm not trying to slight either desktop Linux or Android or SteamOS or whatever, just to be clear. I think it's a huge win for Linux, at least in some sense.
I think it's misleading to read "Linux" and imagine "desktop GNU/Linux with all the trimmings." I can't imagine there won't be some way to get that on a SteamOS machine. My skepticism is directed at the idea that Valve will ship anything which, out of the box, resembles a traditional desktop Linux experience. I expect it will be more like an Android experience than Ubuntu, with rather limited access to OS internals, filesystem, package management, etc.
> but let's say something like X. Or your average window manager. Or, say, almost anything from GNOME.
This doesn't apply to embedded Linux, or Linux for servers, which could be argued to be just as "traditional" a use case as Linux for desktops.
Additionally, for the desktop use case, nearly every distribution uses a different window manager anyway. There are such a wide variety of window managers, I wouldn't know how to compute an "average" between them in a meaningful way. Although I use Linux for a desktop everyday, I don't have GNOME installed, and I would barely notice if X was missing and replaced by Wayland or a different component.
> I think it's misleading to read "Linux" and imagine "desktop GNU/Linux with all the trimmings."
Right. It's not Linux for the desktop, or Linux for the phone, or Linux for embedded devices, or Linux for servers, it's Linux for the living room. My point is that due to the diversity of the Linux ecosystem, there really wasn't such a thing as a "traditional" or "average" Linux Desktop in the first place, and that we can't even say Linux for desktop is the "traditional" or "average" use of Linux. Anything running the Linux kernel should be able to call itself Linux, without qualification.
That's fine, but the context was whether "Linux users" will pay for software or not. I basically agree with you in terms of the facts, but speaking primarily in terms of user behavior, each of those platforms is manifestly different.
I think it's pretty unlikely that Xorg won't be a part of it somehow. X+extensions is basically the only meaningful, portable, direct interface to video hardware that exists on linux. At least if you want the hardware vendor's own drivers (which steambox obviously would, Nouveau is not up to the challenge).
It would also mean that all the 300+ games that currently work on Steam For Linux would have to be retooled for whatever proprietary windowing layer Valve would have to invent. And Valve would have to convince nVidia to come along for the ride.
The closeness of the SteamOS experience to that of a typical desktop Linux distro is irrelevant. If it runs on the Steam Box, it will run on a Linux desktop of compatible architecture.
Considering there is already Steam for Linux, its probable that Valve will ensure Linux games run on both platforms equally well.
In any case, if there is no need for "all the trimmings", if Valve just wants to make an entertainment pipeline, then that's fine by me. It's the same kernel; this is a huge win for Linux by any measure.
But if they are optimizing a number of things in their own distribution, it's not far-fetched to think that games would run marginally better on the SteamOS.
- GPU driver improvements (they've already been working with nvidia, amd, and intel) which everyone benefits from
- kernel patches (maybe they fiddle with the scheduler or something) which anyone could pick up
- new/improved subsystems (perhaps they do some low latency input or audio layer), which (assuming they open source it) distros could choose to adopt or not
- improving porting techniques for bringing games or game middleware to Linux based platforms (everyone benefits)
It's possible that some stuff could be foreign enough to the way it has always been done on linux that it may take some time to make it upstream (see wakelocks from Android finally turning up in the kernel under a different name and a different implementation but providing the same functionality), but if it actually improves things, eventually I suspect the mainline kernel or distros or whomever will come around.
I think you misunderstood the point of my comment.
The context was whether or not "Linux users" will pay money for software. In this respect, the closeness is only relevant insofar as it is useful to discuss "Debian users" in the same breath as "SteamOS users."
Android is Linux, but very clearly not "traditional Linux".
"Traditional Linux" suggests to me the traditional distributions like Fedora, Ubuntu, SUSE, Debian, etc.
It'll be interesting to see how much SteamOS will look like a traditional Linux. Considering the many Linux offerings on Steam, I suspect they'd like to have executable compatibility on that level, which Android probably doesn't have. But will it be as wide open as a traditional Linux? I suspect not.
A steam-related carrot with lower fees might be a huge incentive. But when we are speculating a gaming company in the business of selling $60-$100 games might be more worried about their ability to differentiate their products from sub-$10 games. If they believe steambox could provide that then it might be an easy sell.
Even if the market is small, it's critical to be "first" than later. That way you get a bigger share of the market instead of being drowned in an ocean of software like on many app stores these days.
My (possibly very incorrect) assumption was that the profit from the Linux sales were so low as to not matter, but maybe they can provide a catalog-wide carrot... something more significant (as you suggested) to motivate the publisher en-masse.