As we’ve been working on bringing Steam to the living room, we’ve come to the conclusion that the
environment best suited to delivering value to customers is an operating system built around Steam itself. SteamOS combines the rock-solid architecture of Linux with a gaming experience built for the big screen. It will be available soon as a free stand-alone operating system for living room machines.
Valve goes beyond just building a Linux distribution and grafting Steam on top of it. They are actually working very closely with hardware manufacturers and game developers, which has already resulted in graphics performance improvements. They are also working on reducing input latency as well as audio performance. In other words, they are very serious about upending Windows as the default PC gaming operating system.
In SteamOS, we have achieved significant performance increases in graphics processing, and we’re now targeting audio performance and reductions in input latency at the operating system level. Game developers are already taking advantage of these gains as they target SteamOS for their new releases.
Valve also unveiled that it’s working with the major game developers so that triple-A titles will be natively available on SteamOS. As for your existing Windows games – SteamOS will support game streaming from your existing PC so you can play them on your SteamOS machine in the living room (or anywhere else, of course). ‘Hundreds of great games’ are already available natively on Linux through Steam, too.
This is just the first in a series of three announcements, and it stands to reason that the second one will be a dedicated SteamOS machine from Valve. The third announcement? Well. It’s got a three in it, so Half-Life 3 is pretty much confirmed.
I hope the release is imminent. I would love to try it out.
My, how the tables are turning.
Really?
* Most of the back-catalog won’t work on Linux unless is Valve stuff.
* Most of the games that are going to be released you might as well pick up a console instead of connecting a PC to TV, because you will have a more consistent experience with a console and the last time I picked up a media PC (Asus Revo) with half decent specs (can run Call of Duty 4 level graphics smoothly at decent detail) cost more than a console and I had to buy an additional wireless controller …
* A few people have said that a lot of the mac versions of Steam games rely on Cidre. Ubisoft have traditionally done decent games with shitty PC ports.
Also I like running one OS on my work from home / gaming PC that works perfectly with my kit without any hacks and that OS is Windows.
Edited 2013-09-23 17:48 UTC
Why would it not? Most MacOS X ports from EA and many others already runs via Cider. Look in the package contents in OSX game titles. Nothing stops Steam from doing the same thing for older titles as a stopgap.
Development houses will be lazy and use the compatibility layer. Trust me I am a developer with very little time, I use a compatibility layer quite often on older because I am not allowed to spend time updating it.
Edited 2013-09-23 18:06 UTC
That is why Valve invested in SDl. For me, it is the best “compatibility layer” over there. The speed issues of 1.2 version should be fixed on the 2.0 release that was financed by, no other than.. Valve.
With SDL I ported my little Megaman like game (rockbot.upperland.net) from Linux to Windows, Playstation 2, Dingux/Dingoo, Pandora, Android and PSP. I had a port for Nintendo DS (abandoned only because of RAM limits), and have plans to port to XBox original, Dreamcast, Wii, Gamecube… In most cases, I the only work I have is to add a ifdef on the key mapping and voilá.
If someone with resources port officially SDL to newer consoles (PS3, XBox 360, PS4, XBox One) it could raise the lib to a new patamar, and could end the porting barrier, in theory, to the benefit the most open of all those market-like systems, that is Steam.
And how is that comparable to SteamOS and/or Steambox? The PC you bought quite likely wasn’t optimized for gaming nor was it running an OS that specifically designed for such a task.
Gaming isn’t the most intensive task you can do a computer contrary to popular belief.
Crysis 3 running at Ultra-settings is the only game that even stresses my CPU and my CPU is a first or second gen core 2 duo.
Edited 2013-09-23 17:53 UTC
And yet none of that answers the question.
Yes it did.
I am running the latest general purpose OS from Microsoft on old hardware and the most graphically intensive game that you can buy doesn’t even stress the hardware.
You know what does stress the hardware … Supreme Commander because most of the workload is CPU based.
It matters more what the workload is rather than the OS, because both modern Windows and the Linux kernels use x86 quite efficiently.
I suppose if you have an OS that has less services running and only a few programs running you might save on some RAM. But modern kit comes with 4GB – 8GB as a minimum.
Edited 2013-09-23 18:02 UTC
Not an expert here, but running games is not all about how much RAM and CPU speed you have, but you need a high-end GPU to achieve maximum performance in combination with RAM/CPU. So try to set the maximum graphic resolution you have in your game and it will stress your GPU and not your CPU.
The game doesn’t max out my old 8800GT, everything is set on ultra, the games slows down when there are a lot of units on screen, I used to have the same in 2002 with Rome Total War, lots of units on screen slows down.
Strategy games are all CPU bound for the most part, FPS is GPU bound for the most part and what OS or whether it is optimized isn’t going to suddenly change that.
Edited 2013-09-24 07:36 UTC
And still plenty of games will stress my Geforce 770. Games are GPU dependent.
The point is as you were too thick to see it, is that the OS being “optimized” has little impact when both OSes are on the same arch, similar hardware and both kernels have had a lot of work done on working well on that platform.
I don’t believe for a second the valve team have been able to do something so special with Linux and a regular x86 platform that it is suddenly going to be super fast compared to similar “regular” kit.
http://www.osnews.com/thread?572857
I mentioned supreme commander because if the number of units becomes very high the game lags like hell, and I over clocked the CPU by over a 1ghz to make it last until I can afford an i7/Xeon board. It totally dependent on the algorithms running in the game.
Rome Total War 2 for example, I can’t run on this CPU but the graphics are mid ps3 quality.
Edited 2013-09-23 20:21 UTC
And you have no idea what this means: They are actually working very closely with hardware manufacturers and game developers,
Working closely meaning it includes improving Linux under the hood, graphics stack etc. It’s special because they have the code. I don’t know if you can do similar thing with Windows.
I am not making it a statement of fact that they can’t optimise stuff.
I am asking how much difference will it realistically make, when my monitor can only show 60 frames per second?
They said, “working closely with hardware makers..” something like that, so a hi-res monitors won’t be a problem if games need them.
Because rendering 60+ images per second is exactly the same as displaying 60 images per second.
This is exactly the point that people are making. Your crap hardware is not their awesome hardware. Can your screen to 4K resolution. Some people high-end rigs can, thereby your version of ultra settings are less than their version of ultra-settings. Not everyone’s setup is the same. It’s _hardware dependant_.
Game developers and graphics card drivers are hand in hand. Game developers rely on the graphics driver providing certain features. The game developers probe the driver for what is available and turn on and off features.
Graphics card driver developers work to identify games based on many heuristic characteristics and then optimize the pipeline for running that game specifically. This is the BS that surrounds graphics driver performance metrics.
A metric that says that Crysis 3 runs at “who the hell cares resolution” is total crap because the graphics driver can identify that it is Crysis 3 and optimize the who graphics pipeline to run it at the most capable the hardware can.
If you threw an application at ATI or nVidia that neither of them knew anything about you’d actually get a closer to reality. This is just the way it is and so your statement that X game runs great on your hardware is because, by today’s reprogrammable GPU standards it has been “optimized” to run on your hardware. It still holds zero weight.
And the reason for this is that all current gen games need to run on the Xbox 360 and the PS3. For similar reasons you can play all current gen games with 4 GB of memory with no problems as well. You’d be surprised at how awful specs those old dinosaur consoles actually have by todays standard.
Just wait until you see a game made for PS4 or the XB1. Your Core 2 duo will not be remotely able to run that at the graphical settings of those consoles.
So what do you run?
Even Starcraft II would be very happy with a faster CPU. And that’s three years old.
BF 4 recommended specs is a quad core Intel processor or six core AMD one. And it will use all 8 cores if you have them.
He buys an Atom system with Nvidia ION and talks gaming and using the CPU.
Yeah, great creditability there.
Oh well, at least it was comparable to his consoles or whatever.
(Can it really run Crysis 3 on ultra? In what resolution and frame rate?)
Yes, really. The focus of such a mammoth platform as Steam shifting away from Windows, no matter how long it takes, is a major threat to Microsoft’s business.
Lets see what actually happens because I haven’t seen any numbers from valve on the number of new game purchases that have come from Linux users.
Edited 2013-09-23 18:07 UTC
I’ve not seen numbers from Valve either, and it would be interesting to know. In the meantime, the following page gives the Humble Bundle statistics across the platforms.
http://support.humblebundle.com/customer/portal/articles/281031-pri…
This is jut for interest. Whether these are indicative of the numbers across Steam I wouldn’t like to guess.
The problem about using Humble Bundle purchases as a reference is that (to use a very English term) is a bit of a pissing match.
I’m not entirely sure what you mean by that. Do you mean that it’s just a tiny amount compared to overall game purchases, or that the numbers are biased because people are trying to make a point?
One of the interesting things I notice about Humble Bundle is that many of the games are ported specifically to OSX and Linux purely for the scheme. It’s not the major development studios of course, but this does suggest some motivation for them to do it. Or possibly that they’re all incredibly philanthropic and just want to support charity!
A pissing match is an English term for trying harder deliberately because you are trying to prove something against another group of people for the sake of it.
Linux evangelists typically tend to do this.
I’ve got down-voted for actually explaining some slang in my area. WTF.
Screw Linux. DirectX11 baby lol.
Steam is just taking this route because the Microsoft Store in Windows 8 is eventually going to make them obsolete.
Edited 2013-09-23 21:51 UTC
I think this caused it:
“Linux evangelists typically tend to do this.”
Even though it is true.
And Microsoft tribe says that F/OSS users are cheapskates that don’t want to pay for anything…
Its a warped one. Humble Bundle bundle without Linux has been done. Less Windows copies sold also windows users paid less per unit compared to bundles with Linux.
Seams like a percentage of Linux users buy twice. Once for Windows and Once for Linux or do have better word of mouth. Basically it something. Also the Linux users pay more per unit on average so everything lines up that removing Linux results in lower number of copies of windows and a lower windows unit price. Yes this is also reflected in Humble Bundle with Linux has a higher per unit price for the Windows copies.
So no its not philanthropic from the game makers. It suggests the Linux people are more philanthropic when there platform is supported.
The interest thing of Humble Bundle is the fact Linux users appear willing to pay twice once for a Windows copy and once for a Linux copy if they are running both OS’s.
I don’t disagree with your overall sentiment, but I’d be surprised by this. A single Humble Bundle payment gets you the Windows, OSX, Linux and Steam versions all in one go. There’s no need to pay twice.
Another possibility is that the statistics simply don’t reflect what users actually do. Users can choose which OS they want to be counted for; it’s not based on the actually version downloaded.
As I say, I don’t disagree with your overall comment though. There’s a lot more going on here and it’s not just the raw numbers attracting developers to convert their games to Linux. My guess would be that part of the motivation is to also have the Linux version published on Steam.
I agree with you both here and in the other thread. Gaming is not the behemoth hardware wise it once was. Also, Steam faces the same thing as any other platform: developers, developers, developers, developers. 😛
More in depth, we’ve hit a really fascinating inflection point in the gaming industry. Mobile phones are eating Nintendo’s lunch. PC Gaming is making a resurgence due to budget builds actually, you know, running games that just came out.
Meanwhile in console gaming, there’s a bit of a crisis building. While next-gen consoles are improving, they’ve sort of outpaced the living room theatres they’re built for. HD is standard and QHD or 4K haven’t really made a showing yet. As such, current gen hardware is almost sufficient, with the next gen really digging into the trenches around the current HD and Dolby Digital standards.
Most of the improvements will be situated around increasing mult-tasking so the OS can better handle networking, notifications, and other items in the background while not detracting resources from games. At the perspective of a game developer, you have more polygon room, but the end result is still on a 1080p screen, so it’s a bit of a wash. Yes, some people could tell the difference between 720p and 1080p on their monster, multi-thousand dollar screens. So you’ve now solved that problem and anyone putting out a game running at 720p will probably be laughed at.
However, the real improvement is coming out of having more compute time to make decisions and do more back-end loading to remove load screens.
A lot of these improvements are already there on off-the-shelf computing hardware. So if you take an x86 kernel and start optimizing it for a gaming workload and trimming the user space back, you start to see a convergence. This really makes a Linux based SteamBox seem doable.
Furthermore, Steam has a rabid consumer base. One only has to look at the Steam Summer Sale. It’s a fracking meme of it’s own.
That sort of selling power and a built-for-SteamOS box may make a massive run on the console market that we haven’t seen since the original PS stepped up to take on Nintendo.
Personally, I doubt it though. XBox has Live. Sony has PSN. People on Live can’t play with people on PSN can’t play with people on SteamOS. XBox won’t let you run multiplayer over anything but Live, so Live players will always be siloed. Trying to run console players and PC gamers introduces unusual limits that chafe PC gamers. Also, there’s the developers problem.
OTOH, I’m sure SteamOS players could play well with other PC Gamers. So, there’s certainly that going for them. The SteamOS Online service may just be the developer’s PC multiplayer infrastructure, so there’s minimal additional work and the living room players won’t be isolated to their own console.
It’s a big bet, and and the chances of success and failure aren’t favored one way or another. Steam is shooting high for this one, and I really hope they succeed.
Finally, if SteamOS can do something like Ubuntu and reside on the same partition as Windows, you may see some uptake on regular windows machines. Getting the OS installed in as many places as possible may give you the leverage you need to eventually sell a steam box to that customer. It’s like if Microsoft let you install an XBox on your desktop. I’d like to see where that goes too.
I don’t think console gaming is in as much of a “crisis” as you seem to think. This year is expected to have two of the biggest console launches on record, based solely on the pre-order numbers, both in a less-than-ideal economy.
Also, mobile gaming is hardly a threat to Nintendo, at least for now. The 3DS is one of the best selling pieces of hardware Nintendo has released, with over 30 million units sold worldwide already.
That said, I agree that SteamOS has enormous potential. Only time will tell though.
Edited 2013-09-23 20:04 UTC
Steam was afraid the MacOSX and Windows desktop market will be closed like iOS. That is a large part of their business. Consoles are just stripped down/optimized PCs, Xbox and PS4 are both X86.
If you see a big trend around ‘user created content’, especially in games and that these closed systems don’t easily allow for that.
So what do you do ? They are going to make an open alternative. I think they are going to produce a Linux distribution for OEM PC vendors to create consoles.
Here is their presentation at LinuxCon from a couple of days ago, you be the judge:
http://www.youtube.com/watch?v=rCGMiT0CQAI
Did I read that correctly from the slides Steam has 50 million users ? So slightly more than XBox Live ?
I’m currently watching this, an other talk by Gabe:
http://www.youtube.com/watch?v=t8QEOBgLBQU
Edited 2013-09-24 09:05 UTC
The video from the next presentation by Gabe was also interesting:
http://www.youtube.com/watch?v=jhgOqyZHBIU
My loyalty to open source platforms feels exceedingly vindicated when I check the news these days with all the liasons Microsoft and Google have had with the NSA, and it has felt vindicated ever since I migrated from the Commodore Amiga to a platform where the technical merits were driving community involvement without idiotic corporate management killing it.
That’s the kind of environment I’ll develop for, and even modify closed source games on because I like having a real development toolchain at hand instead of a bunch of random tools someone wrote in Visual Basic. I may have relied on Cygwin to get some of the same result back when I did the Sacrifice of Angels mod for Homeworld or the Foundation plugin SDK for Bridge Commander, but these days, my desktop is pure Linux.
I for one have bought recently Skyrim, Civilization 5, Planescape Torment, and X-Plane 10 expressly to play on a Linux-only Core i5 desktop. I’ve been dual-booting Linux since 1994, and I haven’t kept Windows on any desktop since 2009. I vastly prefer the arrangement since I have an OS underneath that’s open source, highly performant, and has a full development toolchain handy. Further, if the motherboard were to be toasted or I had another machine to migrate to, I could transfer the SDD and HDD out of it and straight into any other x86-64 + Nvidia rig and run without a hitch, without having to buy a license for the new CPU.
Back in 2004 I was buying hardware expressly to do my development projects on, and then ran Gentoo with USE flags compiled to keep performance-killing software sound mixing off of my system completely. The result was a sweet run of games like Neverwinter Nights.
I love this platform, and those games which offer Linux binaries I particularly appreciate. To each their own, but I try to run as much of an open source platform as possible while still having cool gaming experiences.
If Valve plays its cards right, folks like me are going to have more great options, and standards in gaming are going to be even more cross-platform so Windows, Mac, and Linux are all well supported. The fringe benefit? Once you’ve got a working Linux binary of a game, it’s even easier to run it on other new platforms – FreeBSD for example. Less lock-in and more freedom is good for everyone.
Again, initially, their Linux focused is not base on how much current market share they will gain for their Linux Steam Client. Their strategy is for the future, wherein, people will buy a steambox from any manufacturer(since it is open), or download SteamOS and install it yourself, so that in addition to your Windows Steam client, you can now stream your windows Games to your T.V. via SteamOS. That SteamOS thing is needed so that you’ll get the full Steam features.
the entire back catalog doesn’t need to use linux. that’s what streaming is for.
Dude…why so negative…
😛
Why did he get modded down for speaking the truth? The cost of porting DirectX to OpenGL is FAR from trivial, so a good 95% of the games ever made will NOT be available, even with the win 8 disaster MSFT owns more than 90% of the X86 market so good luck on getting most big gaming houses to offer more than a token of their catalog, and then there is the rotting elephant nobody wants to talk about…DRM.
Like it or not there is no way in hell the big publishers are gonna support a living room box without DRM yet its practically impossible to use DRM in Linux because of the viral nature of the GPL which RMS himself has repeatedly said is by design.
So what you are gonna have is Valve games and some indies….which everyone can get on Windows WITHOUT having to install and learn a new OS. Believe me I WANT this to work, we system builders have been treated like dirt by MSFT but unless Valve is willing to pony up a good 100 million plus to make Wine into a perfectly seamless translation layer that requires no thought from John Q Public AND be willing to spend a good half a decade in court as no doubt MSFT would sue if it comes a mile of DirectX IP? Yeah…good luck Valve, you are gonna need it.
Common, games were not made to work as Linux Kernel Modules, it’s in user space, so GPL is not an issue.
The cost of porting D3D to PSGL/OpenGL is also far from trivial… and yet it’s being done all the time.
And Gabe will personally come into your home and poke you with a stick so that you don’t run Windows?
Geezzzz…. Just because it becomes available on other platforms does not reduce your level of experience. You reminded me of those people that went crazy over Instagram coming to Android.
That’s quite the pre-mature ejaculation. A custom linux distro with improved closed-source drivers AKA SteamOS has a LONG way to go before it can be considered any kind of competition for Windows gaming.
Just curious, every time somebody on the internet posts a message about how the linux desktop is about to put Windows out of business, you likely buy into it every time right?
Hah!
Edited 2013-09-23 17:45 UTC
I really hope is not Android based.
It’s Linux. Not Android.
Android is a Linux distribution. It is not GNU/Linux though, and quite different from desktop distributions.
About SteamOS, Phoronix speculates that it will be based on Ubuntu Core.
Android is a Linux distribution.
So?
Distributions based in other distribution are quite common in the Linux world.
Android is Linux kernel + Dalvik + Android APIs. The least important component of that stack is the Linux kernel – it could easily be replaced by any other POSIX kernel and still be Android.
O.K. mister smart ass, let me be more specific, just for you.
I hope is not based on the Android versión of Linux.
That’s not more specific at all.
It is, you just can’t get it.
Not so. Android is dependant on items that are not Posix from the kernel level.Suspend blockers or the final form of wakelocks is require for power effectiveness.
Blackberry has tried todo exactly what you described Drumhellar with qnx Yes qnx is a Posix OS. Yes there are Android/Linux particular things at the kernel level under android that you require so stuff works.
Least important component is not exactly true. Kernel working right is key to how long battery lasts. User not seeing it does not remove its importance.
Yes Google has been willing to work with upstream to get the best power management solution they can.
Could the Linux kernel be replaced in Android yes. Will that replacement be easy todo the answer is no its not. If it was easy we would have qnx based android from blackberry that works well. Some android applications talk to the /dev directory and other interface directories straight to the Linux kernel.
Remove Linux kernel you have to emulate it to have all Android applications work. Blackberry android support some applications just don’t run and will never run due to not good enough Linux kernel interface emulation.
SteamOS will be able to do what other Linux distros were unable to – eliminate the overhead of having to support datacenter, IT, and embedded requirements while optimizing scheduling/input/graphics/audio (no more pulse-audio?) for games. This is real and significant. That being said, making the business case to 14-year-olds who have the time to play games may be a challenge…
Edited 2013-09-23 17:54 UTC
i use linux for everything, except gaming.
now with steam os there will be more games for linux(steam OS\box) in general.
So Good Bye Windows(M$).
Are they going to use Wayland, Mir, Xorg, or some other graphics solution? Will an open source box still be a reasonable platform for the Steam client?
That’s the really interesting question. I’d like to see something closer to a fully free and capable environment than Android.
Well, the announcement talks specifically about how end-users can modify both the OS and the H/W to their liking, so I would suppose it’s at least more free and capable than Android. The other specifics we’ll likely learn soon.
It might happen that whatever games target SteamOS are not able to run in other distributions, by relying on SteamOS specific pathnames, libraries, whatever.
Lets see how it really looks like.
Unless StreamOS is limited to some weird hardware platform the worst case scenario is StreamOS chroot (just like i386 support in amd64 distributions). But I really doubt it would take that much of trouble.
It is more than just hardware.
Linux is just a kernel, so the main question is what SteamOS puts on top of it and how different it is from other distributions.
All the Linux-based systems are not that incompatible. You can have a system that combines the software of distinct flavors like, say, Android and GNU in Ubuntu Mobile. Unless Valve does something specifically to hinder compatibility, desktop distributions could assimilate it.
Have you ever had fun trying to run binaries targeted to one distribution into another one?
There are lots of things that can go wrong:
– .so versions
– contents of /proc
– file locations
– scripting language versions
– …
Yes, I have. I almost always used the not so popular distributions, and so I know these problems. Yet, they are trivial to fix – build a bunch of libraries, or – in a worst case – setup a chroot. BTW you may even run most i386 Linux binaries on i386 BSDs. Really, this is a minor annoyance.
For us yes, but not for the users SteamOS is expected to bring to the UNIX world.
And that’s why they’re getting SteamOS, where they are not exposed to this kind of problem, and not a random Linux distro with Steam.
Yes, but nobody said it’s going to be novice-friendly. Valve only promised that users will be able to “alter or replace any part of the software or hardware they want“, which sounds very much like Android on Nexus (or any other pre-rooted device for that matter).
FWIW they promised that “SteamOS will be available soon as a free download for users,” so it’s likely that we’ll have a chance to get our hands on the product and see it for ourselves.
Heavy modification of the core of the OS? I’m a bit worried. I love that steam is on Linux, and I’ve now eliminated all traces of Windows from my desktop. I’d be very worried though if these triple AAA titles built for Steam OS REQUIRED it to run reliably and had shoddy support under other Linux distros.
You have to remember many of the improvements Steam does or helps with end up in the upstream Linux/open source projects.
So with a bit of luck the differences with regural Linux should be minimal.
I dont get it. What are they talking about? Has Linux ever been rock solid? Upgrade your kernel, and things breaks apart. Hardly rock solid.
Lol, yeah I was thinking the same.
s/Linux/Ubuntu/
Never really had this problem on any of my linux machines. However, I don’t compile my own kernels from the kernel.org. That might be why you have/had problems.
On fedora for example there is akmod nvidia drivers and kmod nvidia drivers and running one of those used to give you problems with kernel updates if I remember correctly.
There is post from ask.fedoraproject.org
https://ask.fedoraproject.org/question/9106/what-is-the-
difference-between-an-akmod-and-kmod/
This was not an uncommon situation. Catalyst drivers were a nightmare when I had to use them.
I am sure valve have a strategy for handling updates, but it has to be pretty solid.
Edited 2013-09-23 18:34 UTC
Proprietary drivers, yea those can break on kernel upgrades. Most of my Linux experience comes from headless server boxes. And its being silky smooth ride all those years.(aside from some problems with software packages which is not related to kernel itself)
However, I did use Ubuntu back when it was 8. something and the most problems I had was having to recompile headers for Nvidia driver(or something like that) as I was using the official Nvidia installer.
We all know they can but while the newer intel GPUs are loads better than the old ones, there is no high end GPU that has open source drivers (that perform similarly to Windows) for Linux.
I have perfect Linux updates on my Linux workstation that has everything that is intel, but outside of that I get a mixed experience.
It has got a lot better since Redhat 8, Suse 8.x days … but it can still be problematic.
Maybe you shouldn’t attribute your own failures to the OS.
I am not talking about my own failures. I am talking about the Linux design; unstable APIs. Whenever Torvalds change something, you need to recompile all drivers. Frequently there will be problems after upgrading the kernel. Have you missed all the complaints on this?
How sane is it to have one driver tied to a kernel version? Kernel v3.1.23 runs this driver, and v3.1.31 runs another version, etc. Linux driver model is broken.
If you talk about LTS long cycle distros, they are not better. You want to install a software, which requires new version of a library. So you need to upgrade library too. And then other software breaks, so you need to upgrade them too, etc. It starts a chain reaction where you have upgraded everything, so you are not on LTS long cycle distro anymore.
How can this broken Linux driver design, be “my own failures”? No other OS has unstable APIs, maybe for a reason. My old WinXP drivers, still function fine on all WinXP upgrades, ServicePack1, SP2, etc.
Imagine the situation where Crysis is running fine on WinXP, but if you use a upgraded WinXP kernel with SP1 you need to patch Crysis. And patch yet again for SP2. etc. Is this sane? Is this the “failure of the Windows user”, or is it a broken design? If Windows functioned like this Windows would be so bad mouthed by every nerd. But when Linux functions exactly like this – it is the “user’s fault”. Dont you see the problem? You dont acknowledge there is a problem in Linux driver model? No other OS functions like Linux, for a reason.
As the owner of http://www.GamingOnLinux.com this pleases me to see them using Linux on the SteamOS
, I just hope it doesn’t become incompatible with desktop Linux distributions.
I actually read your GoG article the other week, about them not having the time to support Linux. Was a decently written article.
Looks like the right idea, will come down to execution and how compelling the SteamBox is.
The big take away from this is that worst case scenario Linux has still taken major steps forward in gaming.
Good luck to Valve, this is going to be very expensive for them, and bloody. Ask Microsoft how long the Xbox took.
I know Valve is a company that makes the steam platform but SteamOS sounds too much like a steaming pile of.. code.
I am a huge Valve fanboy though and I wish them luck. Hopefully they see more success than the under powered Ouya did.
Well I guess I called that one. Still surprised as anyone though.
A Linux distro designed to unseat MS Windows. What will they think of next?
Maybe they’ll succeed, maybe they won’t, but it’s not like it’s a new concept or anything. If most devs port their games using Cider, it’s not even a win, they still need the Windows API, it’s more like a draw.
They don’t will have to port anything. Box can stream games from Windows host. But they will, because Valve is doing excellent job with SDL. It was everythere and I hope that version 2.0 will be a big comeback.
With improvements on input and sound latency on the Box part (not the strongest point of Linux gaming now) and customized kernel without hundreds of unused drivers we’ll have good gaming distro. And it’s bad for other distros.
I’m a Fedora fan and I’m understand that Linux games are tested pretty much only for Ubuntu and sometimes for Suse (at least I have RPMs). Now even Ubuntu users will be left with “self-support”, because the only reason for porting games for Linux will be the Box. It’s based on Ubuntu LTS, but I’m pretty sure that, in time, there will be some incompatibility.
As I understand what’s going on now. The development tools will be shared across distributions. As long as the license of their tools allowed for sharing, this would not be a problem.
BRAVO GABEN, BRAV-O.
I’m a big steam user on Windows myself and used to use Arch as my primary OS. SteamOS won’t really challenge Windows, except on the gaming front, which is probably declining somewhat in favor of mobile gaming; If SteamOS is based on the foundation of an existing distribution (probably Ubuntu) then it has real potential to help Linux become a top-tier platform for computer game development and as a result, we may see many more hot game titles supporting Linux natively.
Hopefully SteamOS will also support or bundle Wine software, which supports the Windows version of Steam and hundreds of Windows games that run from within it.
I respectfully disagree. IF linux gaming is going to be a thing there should be as little half-assed solutions as possible.
“An Education”
There are two equally important hardcore game platforms right now: Windows x86 (Xbox One) and Linux x86 (Playstation 4).
Whether SteamOS exists or not, and whether or not you get Linux gaming in the home, Linux has become a more prominent game platform than ever before. Game developers are making games for PS4’s Linux RIGHT NOW. It compiles and runs.
The new consoles being identical to PC hardware and software conveniently speeds their demise. Openness is unstoppable. This decade or the next, SteamOS or something like it is coming to evaporate billions in licensing-stranglehold revenue.
Not really. PlayStation 4’s operating system is called “Orbis OS” and is (supposedly) based on FreeBSD (http://en.wikipedia.org/wiki/PlayStation_4#Software_and_services) which, considering the well known licensing issues, makes sense for a company that rather not share but still play by the rules.
As for SteamOS itself, at the moment there aren’t enough details to express an informed opinion, therefore I’ll just keep an eye on it — but no, no matter how much I would like it to be true, I won’t hold my breath waiting for “all the AAA titles coming natively to SteamOS in 2014.”
RT.
Ill have to widen my statement to posix opengl
A couple of things stood out to me….
“SteamOS combines the rock-solid architecture of Linux with a gaming experience built for the big screen.”
Rock-solid? Really? That’s an amazing claim.. One that countless linux devs and Linus himself don’t agree with.
“Valve goes beyond just building a Linux distribution and grafting Steam on top of it.”
No, that’s pretty much all it is after you remove the PR fluff.
I plan on giving SteamOS a spin. BUT, it has a LOT to prove before I’m willing to take it serious and that view isn’t exclusive to me. Valve has yet to say anything that hasn’t already been said before and there’s nothing tangible that makes me believe SteamOS or the Steam Box can live up to their PR hype.
Be careful, people will mark you as a Troll for quoting Linus and other Linux kernel developers.
They would be idiots for doing so since the linux dev mailing lists are public and it’s well-known Linus isn’t scared to speak his mind. These people and a lot of what they say are very accessible between the mailing lists and freenode irc channels (which are typically logged). There’s nothing amazing or magical about being in contact with them.