Developers And The Dreaded Platform Listing Of "PC"

Sure they did, Ubuntu has been by far and away the most popular desktop distro for nearly a decade.

Are they as widely used as Red Hat/CentOS or SuSE are in terms of total number of systems it is installed in? Probably not, however, Red Hat/CentOS and SuSE are not desktop distros, they are server and dedicated workstation focused.

They are most defiantly not what the vast majority would pick to use as their living room computer for the kids to mess with.

And YOU think that Valve wouldn't just say that SteamOS isn't Ubuntu based just to avoid THIS SPECIFIC ARGUMENT FROM RAGING IN THE FORUMS FOREVER then you are just naive.

Because that is exactly what I and anyone else in their position would do, lie to the belligerent end users, to save time and money on duplicating effort needlessly.

Quoting: Quote from intokSure they did, Ubuntu has been by far and away the most popular desktop distro for nearly a decade.

Did you read the thread above? For nearly a decade Canonical hyped it as such. I never saw them showing any numbers which could demonstrate it, I doubt they even can. I don't buy hype talk.

Valve shouldn't care about arguments of what distros users prefer. They care about how to develop a system that will serve their goal (to be a good console gaming OS). They can as well make a distro from scratch, but it all boils down to cost and efficiency. You are looking at this from users perspective, but that's wrong, Valve looks at it from system designers perspective and they have their own priorities. Using Wayland based metaditros in the future has all the advantages, using Mir based Ubuntu has none, since Ubuntu own reasons are not logical, or at least not technically reasonable. Valve doesn't even benefit from any Ubuntu specific bits - they aren't going to use Unity or anything the like. So what does Ubuntu give Valve as a base, that Debian or Gentoo don't for example? If you say Valve can follow the same reasoning as Canonical did to think that Mir is so great that it's worth to split themselves from the majority of the Linux world, then we'll just have to wait and see if it's so or not.

Their choice of Ubuntu as a primary support target for Steam was done before all this Mir story came out. If they'd to decide knowing all this, they could chose differently. My point is, nothing prevents them from doing that now, for the future of their Steam OS development. It's not even out yet. Things change and old plan can fall short of expectations.

For example Google based their ChromeOS on Ubuntu originally as well, but later switched to Gentoo. It's not about popularity, it's what works for them.

So netstats, download counts, hundreds of magazine and web articles and hundreds of thousands of Youtube videos all backed up now by Valve never happened?

I take it you think Distrowatch is the only reliable metric?

So, killing 2 birds with one stone is a foreign concept to you? What is the biggest problem facing the indie developers and the biggest complaint of the AAA game devs? The cost of having to test against so many distros. If you are Valve you want to mitigate this to increase sales not just on SteamOS or on Linux but on both. So you base SteamOS on the fattest target and thus minimize the testing requirements needed between the 2 platforms allowing for increased profit.

Wayland, Weston, Mir, it's all open source and nobody with a clue give a god damn what one comes out on top at the end of the day as the code will just be copied and pasted into your pet distro that only 5 people have ever used.

Google switched to Gentoo because it's source based and they are building for various ARM platforms. To get the most out of these platforms they need to compile from source.

Argument over.

Distrowatch is not reliable for any estimations, they write so explicitly. All that you mentioned above aren't comprehensive statistics. It works by the method "if you talk about it enough, people will start accepting it as such". That's how I see the claim that Ubuntu is most used, and until some solid numbers can demonstrate it, I'll keep it as unproven in my view.

Console OS by definition needs to be optimized to the maximum like crazy, that's what all PlayStations and Xboxes do, so actually Gentoo which is bent on optimization makes all the sense for Valve (I'm not using Gentoo, just showing an argument).

Ubuntu won't remain (and isn't, the way I see it) a fattest target, thanks to their shortsighted Mir decision, which will essentially severely isolate them from everyone else. So, from system design perspective, betting on Ubuntu is simply bad. Canonical can of course come back to their senses and drop this Mir obsession, but why should Valve wait for that to happen? Even downstream users of Ubuntu, like Kubuntu and etc. were pretty concerned with it, and contemplated how they are going to maintain Wayland integration on their own. So, why should Valve get involved with this headache to begin with? Just thinking out loud, we of course don't know what Valve actually plans to do. This touches their long term plans, since neither Wayland nor Mir are ready for gaming desktop usage (drivers wise).

Quoting: Quote from ShmerlDistrowatch is not reliable for any estimations, they write so explicitly. All that you mentioned above aren't comprehensive statistics. It works by the method "if you talk about it enough, people will start accepting it as such". That's how I see the claim that Ubuntu is most used, and until some solid numbers can demonstrate it, I'll keep it as unproven in my view.

Console OS by definition needs to be optimized to the maximum like crazy, that's what all PlayStations and Xboxes do, so actually Gentoo which is bent on optimization makes all the sense for Valve (I'm not using Gentoo, just showing an argument).

Ubuntu won't remain (and isn't, the way I see it) a fattest target, thanks to their shortsighted Mir decision, which will essentially severely isolate them from everyone else. So, from system design perspective, betting on Ubuntu is simply bad. Canonical can of course come back to their senses and drop this Mir obsession, but why should Valve wait for that to happen? Even downstream users of Ubuntu, like Kubuntu and etc. were pretty concerned with it, and contemplated how they are going to maintain Wayland integration on their own. So, why should Valve getting involved with this headache to begin with? Just thinking out loud, we of course don't know what Valve actually plans to do.

I say this as a Mint Mate edition user.

Again, wrong, there where plenty of overly hyped desktop distros over they years, am I the only one here that has been around long enough to have used Lindows and Mandrake? Neither gained more then a passing mention by anyone because while they where easier to make into a workable desktop distro then Debian or SuSE they where still quite broken. Every nerd out there hates Ubuntu for getting things right enough that you could sit anyone in front of it and they could use it.

Not true about console OSs, the only thing that the console does is allow a stripped down OS with a lower level access to the GPU hardware to get more out of it then higher level languages like OpenGL or DirectX can allow.

The history of consoles is one of custom hardware to have a captive market and to extract licensing fees from the game devs. You want your game on the Wii? Cough up the cash for a dev kit and the Wii logo on your box or prepare to be sued.

The last thing SteamOS needs to be is tweaked for performance and as such suffer a loss in compatibility. The whole point is that it's an open platform running on off the shelf hardware. That hardware is already plenty fast. Will it have the longevity of an unchanging set of hardware that determined and talented devs can squeeze blood from a stone with in 5-7 years? Probably not, but that isn't the point, the point is that in 5 years if your CPU has the beef to handle the game you want to play but your GPU doesn't you can order a new one off Newegg and keep the same box chugging along for year to come.

Again, if everyone is already using Ubuntu how exactly are they going to lose anything by going to Mir? The only thing that is going to happen is that as time goes on the rest of the distros will lose compatibility with commercial software and as such slip ever further into irrelevance.

Who cares about down-stream? They will patch it if they want any kind of relevance. Remember, even Mir is open source.

So yeah, this is my last and final post on this topic. You can't see your own hand in front of your face let alone the big picture. Hit me up on IRC if you can come up with an actual argument, but it had better be well thought out.

You make your assumptions on the same claim that "everyone uses Ubuntu". So depending on how much you believe this to be true, you can estimate further what Valve is going to base their system on. I don't consider this claim to be valid, or at least proven. And even if it is, that's not what system design should be based on - it should be based on technological advantages. Wayland is clearly better looking from each angle now. Can Mir suddenly become so good that everyone will start using it? May be, anything can happen. But so far it looks like it will remain Canonical's internal tool only. It doesn't look like you can see a big picture here, if you claim that it makes sense for Valve to use Mir. No one who commented on the matter, so far bought Canonical's claim that it made any sense for them (Canonical) to ditch Wayland for Mir. Let alone for anyone else to do the same thing. Valve didn't voice their opinion on that yet.

I don´t know if it was on Steam forums or github or wherever it was but I read some discussion about changing this, as a PC is a personal computer (like Kristian said it was different in the past) so it has a wide meaning, in my opinion the most accurate description would be Windows/OSX/Linux.