The steam machines always felt like a weird move to me. The target audience for those machines seemed very small - gamers who don’t want to buy a console, but also don’t want to build their own PC. I can kinda understand the appeal of not wanting to deal with the hassle of putting all the components together, but I could never personally justify the cost of the machine against the cost of just buying all the parts myself and spending an afternoon with a screwdriver.
You are missing a significant market chunk. People like me. I have over 300 titles on steam. Buying a console now would mean chugging a bit of cash + re-buying/getting new games. Buying a brand new steam console would mean I get ~150 titles without any additional cost (that’s the rough number of the games already ported to Linux from my collection).
I don’t want to build my own PC just for gaming and I don’t want to run Linux anymore on my main laptop/PC. I would potentially consider getting a console from Valve especially since I expect them being better at hand picking & testing hardware against titles available on their platform.
Valve is a long haul player. People HATED Steam when it launched. Give them some time to iterate on the idea.
I don’t want to build my own PC just for gaming and I don’t want to run Linux anymore on my main laptop/PC.
I come from a different direction, that could be served by the steam box: I could build a PC but I don’t want to mess with maintaining the Windows on top of if with system updates, virus scanners, driver updates, malware and whatnot. I used to run Windows for gaming on an external but nowadays I am too tired to care, instead opting to use consoles, since it’s less of a hassle.
The downside of Steam machines is, that it is still running Linux (nothing wrong with it per-se, I run it on all my computers), therefore the Steam games I want to play are often not available for it, at which point I have gained exactly nothing.
In my case I’m moving more and more to an OpenBSD only setup switching only to Linux where work requires it (docker, dart, android yuck). I’m OK with the gaming library available for Linux at this point. I can see myself having a dedicated Steam console to play from time to time. In my case I would gain a machine able to play the games I own, not having to hand pick hardware for it and having more machines free to run OpenBSD :)
From Valve’s perspective, if they had worked, they would have effectively had their own console. It was merely a coincidence that it would have been a general-purpose computer also, and I bet it would have been very badly maintained from that perspective, compared to Ubuntu or Debian. (Yes, I realize that Steam Machines are both a hardware and a software platform; I don’t think that changes it. Being a Linux distro is the hard part of that; PC hardware has been commoditized forever now.)
I imagine Valve sees it as a significant risk that Apple and Microsoft each sell software through their own channels, and that there’s nothing particularly distinctive about Steam as a store. Their community features are desperately trying to add value (I mean, what else is the motivation for Steam trading cards…?), but I personally don’t feel that they add very much. Ultimately, their market position happens to be leading but is not remotely secure; if they could have gotten to a state where most of their users were on Steam Machines, they would have been much safer.
I haven’t been following industry news, but I feel as though consoles in general have been losing momentum in favor of mobile for a lot of things, and general-purpose platforms for things with serious hardware requiremetns. So, for example, I don’t feel like Nintendo’s distribution bottleneck is anywhere near as secure as Nintendo would like it to be, since their platforms have been flopping lately. If my feeling is right, it’s a way that Valve should have known better than to try this. But it’s their loss, nobody else’s, so I’m not convinced I care. :)
I imagine Valve sees it as a significant risk that Apple and Microsoft each sell software through their own channels, and that there’s nothing particularly distinctive about Steam as a store.
That’s how I remember the original SteamOS news being reported in 2013, that it was an attempt by Valve to do an end-run around the possibility that Apple and Microsoft would increasingly lock down their desktop OSs towards more of an app-store model, like on mobile. Since Valve effectively has the same business model (app store that takes a percentage cut of the app sales), they felt they needed their own platform to keep from being eventually locked out.
I haven’t been following industry news, but I feel as though consoles in general have been losing momentum in favor of mobile for a lot of things,
Indeed. The problem for the console manufacturers (and especially Nintendo) is that mobile devices are good enough nowadays for casual gaming. If you want to play on a big screen, devices like the Apple TV (and probably the Nexus Player)  are much more cost-effective, though the number of games is still somewhat limited compared to mobile. The only thing that Nintendo has left for the casual gamer are their franchises, but I don’t think that will be enough to convince the general population to spend more, lose integration with the iOS/Android ecosystems, and deal with their oddball WiiU controllers.
 Since we had a 4th generation Apple TV, our XBox One has only been power up to play some Blu-Rays.
I think Nintendo is being squeezed here more than the others are, since the games are more similar in style to what you can find on mobile. The more resilient group of console gamers are those who either mainly play sports games (e.g. Madden), or those who play FPS type games (e.g. Call of Duty), which are each a pretty huge number. There’s an interesting niche of Americans, especially less wealthy ones, who own two computing devices—a gaming console and a smartphone, but no PC/laptop—and so far they seem to be staying with that configuration.
For Valve to have any position to negotiate with MS from, they have to have a credible threat of being able to walk away from Windows. Mature Linux support and Steam Machines (which are integrated systems) make a credible threat. You could say it’s dead in the water as a consumer product. Maybe it’s not a consumer product though, maybe it’s a negotiation tool.
And, Valve could step on the gas at any moment. I think they have a good shot at shipping the best possible gaming hardware platform by virtue of having a lot of development resources and a strong track record for delivering quality. Maybe they’re just letting the project idle because they’re close to closing an agreement to be the only other MS endorsed app store on Windows?
Even years before its launch, the idea looked a bit like a solution in search of a problem, predicated on the belief that Microsoft would start exercising monolithic control of the Windows software marketplace any day now.
Ironic that they would publish this today with all the news this week about Microsoft’s changes that make it so you can no longer decline Windows 10 updates. Seems rather prescient to me now that Microsoft has made it clear how little they value user consent.
I ran SteamOS for quite some time, worked fairly well! I think that if they had gotten more of the “top tier” games to work it would have taken off!
Heh, that’s true. Big-budget games are always the driver. From my perspective, I never had the slightest interest in SteamOS because I actually care about using a Linux that’s Linuxy, and customizing it heavily in dimensions that the more consumery distributions aren’t good for. I’m not going to give up the things I care about in a general-purpose OS just to increase the inventory of games on it.
But I’m a minority; if you look at today’s buying options for desktops, in particular, you can see that they’re marketed almost exclusively to gamers. Laptops still have somewhat wider appeal. I think most people today who are interested in big-budget games would prioritize them over whatever else they use their computer for.
I actually care about using a Linux that’s Linuxy, and customizing it heavily in dimensions that the more consumery distributions aren’t good for.
You would feel right at home on SteamOS - it’s Debian testing with a different UI. I used it as a file, vm (kvm), media server, stage build machine for trying cross build rust on OpenBSD.. you name it!
Right, but does it have hermetic builds with an immutable root filesystem, with the system defined solely by pure-functional config files? :) Debian is too consumery for me. :)
Are you alluding to Nix? If so I believe it sorta works: https://nixos.org/wiki/Steam
It does! I use Steam on NixOS, and I understand the wrapper enough to modify it.
But that’s the Steam app, not SteamOS. :) SteamOS would produce a level of reliability for games actually running properly, which the Steam subscriber agreement explicitly does not provide for other Linux distributions. (Except maybe Ubuntu - I know that their guarantee was for Ubuntu before SteamOS existed, but I don’t know the current status of that.)
Have you tried using BigPicture mode? Best I can tell on SteamOS it’s just tweaking your X11 stuff to start in BP mode :D
I’m talking about, for example, games that run but don’t show text because they hardcode a font that they erroneously assume everybody has… or that run, but display a blank black window for some reason no doubt having to do with OpenGL versions… or that run, but at 25% their intended framerate… or that run, but only after I figure out which unusual library they erroneously assume everybody has, and add that dependency and a remapping for it to the Steam wrapper, …
These could have been awesome, but they just aren’t ready yet. The software still gets updates, but I have the $800 Dell machine and it’s quite a let down.
I’d be okay with how let down I am with performance on AAA titles, but I have over 400 games on Steam and less than 80 work on the Steam Machine. The real issue here is that they didn’t deliver on the goal to make wine amazing (and therefore play Windows games), and developers aren’t exactly jumping up out of their chairs in excitement to support Linux. Until game studios have strong incentive to build in Linux support, this is a tough sell.
The controller is really neat though, and the Steam is surprisingly smooth.
I don’t think the idea of Steam Machines was doomed to fail from the get-go, but there are at least 2 significant hurdles to overcome for it to be a success.
First, they have to convince publishers to put the time into developing Linux versions of games, when there is virtually no monetary incentive to do so. This could potentially be overcome by chipping in resources to get major games with lasting fanbases ported over, and to get Steam Machine-ready ports released day-and-date released with their Windows counterparts. Not impossible, but not an easy task.
Second, they have to convince every major publisher to sell on Steam. A couple of the biggest games in the world right now (Overwatch and Hearthstone) belong to Blizzard, who is not going to be selling on Steam any time soon. They have a huge fanbase for their games, and they have no trouble getting users to download the Battle.net client as a portal into their games. I can’t really think of a reasonable situation that would make Blizzard willing to offer their games on Steam. Maybe something like featuring Overwatch in a certain position on the Steam store for X number of days, and taking no cut of the sale, and also making end users still go through Battle.net in the end? Doesn’t really seem like a route either party would be willing to go down, and Blizzard isn’t the only publisher to contend with on this front.
I don’t think the failure thus far is a surprise, looking at just two of the big obstacles that need to be overcome to sell Steam Machines. However, the potential payoff for success, and the risk of doing nothing probably pushed them over the edge to give Steam specific hardware a chance. If Steam Machines work, Valve becomes the “App Store” for PC games going forward. If they lose, there’s a potential for Microsoft and Windows to start eating into sales by shifting their gaming focus from Xbox to an Xbox + PC tandem (and it looks like they’re moving in this direction, I’m expecting all major Xbox exclusives to come to the Windows 10 store and release day-and-date with the console versions by sometime 2017). This is a parallel model to what Steam is going for with PC + Steam Machine, and losing out to Microsoft here could potentially eat into future earnings potential.
Even Valve probably put “abject failure” high on the list of probable outcomes, but even so, I can understand how they would think it’s too big of an opportunity not to roll the dice on. Who knows? Maybe with the revelation about the half-step versions of the Xbox One And PS4 coming out, consumers will be disenfranchised enough with the state of the console landscape, and consider the more “honest” approach of Steam Machines as a PC-in-a-box. I wouldn’t bet on it, but I don’t think Steam Machines are necessarily totally over yet, and I don’t think this is the last opportunity that will present itself to Valve.
I hope to God they don’t release Half-Life 3 exclusively for free under their own hardware as a last ditch effort to pump publicity into what appears to be a failed attempt at entering the hardware business.
Don’t worry, it’ll be released exclusively for Hurd.