Steam and Linux

If Steam releases this on Linux, I will make a full switch to Ubuntu and just have a tiny partition of Windows I can boot just for cross platform development, since I'm hugely lacking Linux development experience, but I would do cross platform programming with Linux, to help the community including that most software packages out their are open source, and I love collaboration, plus hey if you make commercial software for it too, more money too. =)

I do think that this is going to be a major milestone for Linux gaming, its going to mark the first earnest support from a sizable commercial games company for Linux, and one that brings with it not just first-party titles, but a broad distribution channel for third parties, big and small. The hardware vendors will, in turn, take driver support more seriously, and I expect a slow-burning-but-consistently-positive feedback cycle to occur.

Valve is doing this because they believe that the other two platforms, Windows and MacOS, are heading towards a walled-garden, where the computer is no longer a general-purpose device under the control of the user, but a platform who's ecosystem will be more-tightly controlled by their vendors. It would be foolish for a digital distributor to put all its eggs in a basket where the message is so clear that the vendor want's to own, control, and tax the only channel in the long-term. It would be like AMD or Nvidia relying on Intel to manufacture their chips.

If Steam releases this on Linux, I will make a full switch to Ubuntu and just have a tiny partition of Windows I can boot just for cross platform development[/quote]
There's still the important issue of the games themselves being Windows-only. Wine is nice, for the most part, but it still isn't perfect by any means.

will this influence your decisions regarding targeting Linux as a platform for your ongoing / future projects? If so, why or why not?[/quote]
I do already, but that is mostly a side-effect of using Linux as my preferred development platform.

I expect a slow-burning-but-consistently-positive feedback cycle to occur.[/quote]
Minor off-topic note, but I sincerely hope so. I also sincerely hope some of the more .. erractic (read: insane) ... elements of the FOSS/Unix world don't drive them away.

Valve is doing this because they believe that the other two platforms, Windows and MacOS, are heading towards a walled-garden, where the computer is no longer a general-purpose device under the control of the user, but a platform who's ecosystem will be more-tightly controlled by their vendors. It would be foolish for a digital distributor to put all its eggs in a basket where the message is so clear that the vendor want's to own, control, and tax the only channel in the long-term. It would be like AMD or Nvidia relying on Intel to manufacture their chips.

This does make sense. I'd be curious to know where you have read about it though.

I still think Windows will remain the dominant platform, at least for quite a while yet. It's all too easy to go buy a machine from the nearest big box store that is loaded with Windows. Until Linux becomes more available (and less scary) the the average consumer, I don't see it taking off as much. Also, how is the video card driver situation for Linux? Has it improved much in the past few years?

Developing for a potentially growing platform is always a bit of a chicken vs. egg scenario. No one wants to develop for a platform that isn't popular, and a platform that isn't popular won't become popular unless it has reasons to move to it.

Gabe Newell has said as much, directly. I believe it was in the Phoronix article.

Windows is and will remain dominant for the foreseeable future, and its true that valve could not survive on Linux alone -- at least not as we know it today. However, they are essentially taking control of their own destiny with this move, because Apple and Microsoft could (and will, if they believe it makes business sense) lock out other distribution channels entirely, or perhaps charge third-party channels for the privilege.

If push came to shove, only Valve, EA, and maybe Activision are large enough to resist, practically speaking -- by which I mean that only these few are large enough that they could conceivably launch their own platform (E.G. game console, or internet-based game streaming) if they had no other choice. NVidia (most likely) and AMD (less so) could also conceivably shake-up the space with their own hardware, but its far less risky for them to be a supplier, and they'd probably just be another walled-garden anyhow.

I hope it does result in more people producing games for Linux.

I do fear it might just result in an OS/2 games type situation though. (OS/2's support for DOS was so good that games companies just released their DOS games with a configuration file to use when running under OS/2. As a result very few native games were produced. I could imaging them doing the same for Wine.)

I always use only cross-platform technologies (C++11, OpenGL) for all my work, so they can run on Mac, Windows, Linux.

Most people think that the source of this is fake (why would Gabe reply to some random stranger about such an important question when Valve has said nothing publicly?), so don't get fooled. Though I think that Source and Steam being ported to Linux was more or less confirmed (not sure). Then again, Valve Time...

Valve is doing this because they believe that the other two platforms, Windows and MacOS, are heading towards a walled-garden, where the computer is no longer a general-purpose device under the control of the user, but a platform who's ecosystem will be more-tightly controlled by their vendors. It would be foolish for a digital distributor to put all its eggs in a basket where the message is so clear that the vendor want's to own, control, and tax the only channel in the long-term. It would be like AMD or Nvidia relying on Intel to manufacture their chips.

Either this or the whole "Steambox" thing (it'd be cheaper to bundle Linux than Windows). Or both... I doubt Windows 8 will be a full walled garden, but I wouldn't be surprised if for Windows 9 they attempt to remove support for non-Metro programs in home editions :/ (seems unlikely, but it could happen - depends on Windows 8's reception, I guess).

Minor off-topic note, but I sincerely hope so. I also sincerely hope some of the more .. erractic (read: insane) ... elements of the FOSS/Unix world don't drive them away.

I'd be more worried about them giving up with the state of OpenGL drivers on Linux. The only ones that work reasonably fine are the proprietary nvidia ones, the rest all seem to have issues as soon as you go beyond the basics (or at least far enough for the kind of stuff Source does). Also the state of wi-fi drivers on some hardware, though no idea how bad is it.

will this influence your decisions regarding targeting Linux as a platform for your ongoing / future projects? If so, why or why not?

Depends; if I'm using an engine which makes it seemless then I'd just release.

If was building from scratch then unless an API which was as good as DX11 was released I'd say hell no.

Given the OpenGL ARB are about as useful as a choclate teapot in the depths of hell I'm not holding out much hope 3D wise... OpenCL is better but still has gaps and support issues, heck even finding out if NV supports 1.2 on Windows isn't easy.. as for sound... I've not played with Windows sound and I don't know what exists on Linux; I use to like OpenAL but XAudio seems like a saner API... same for input; if I was making a game which used the 360 controller than XInput is easy, even windows messages is easy; last time I tried to deal with in on linux I woke up 2 days later hung over and with a gap in my memory.

So.. yeah... APIs if not using an engine... MS still seem to rule the roost there...

Given that Gabe Newell says that a steam client for Linux will be launching later this year (assuming this isnt like other Valve long-delayed projects), will this influence your decisions regarding targeting Linux as a platform for your ongoing / future projects? If so, why or why not?

For me, I think this is will push me multi-platform and is one reason I'm looking into things like mono and monogame.

It will influence my productivity when working on Linux

Given the OpenGL ARB are about as useful as a choclate teapot in the depths of hell I'm not holding out much hope 3D wise... OpenCL is better but still has gaps and support issues, heck even finding out if NV supports 1.2 on Windows isn't easy.. as for sound... I've not played with Windows sound and I don't know what exists on Linux; I use to like OpenAL but XAudio seems like a saner API... same for input; if I was making a game which used the 360 controller than XInput is easy, even windows messages is easy; last time I tried to deal with in on linux I woke up 2 days later hung over and with a gap in my memory.

To be fair, with SDL you can easily cover about all I/O stuff a game may need (network aside, but on Windows sockets aren't very different), so that pretty much makes things a lot easier, especially for SDL 2.0 which is under the zlib license (not LGPL). Your only issue would be OpenGL... which is what I was worrying about, since that part doesn't work very well in the general view of things.

Given that Gabe Newell says that a steam client for Linux will be launching later this year (assuming this isnt like other Valve long-delayed projects), will this influence your decisions regarding targeting Linux as a platform for your ongoing / future projects? If so, why or why not?

For me, I think this is will push me multi-platform and is one reason I'm looking into things like mono and monogame.

I think this is a good strategic move for Valve and linux gamers will benefit from it, but I have my doubts about the cost vs. benefit ratio for game developers. If I make a game in XNA/C#, then I'd have to port it over to some comparable platform on linux and run through the whole gamut of QA, from top to bottom. Therefore, it would be much better to decide to target Linux and Windows from the outset and choose a platform which is supported on both OSes (Java? C++? ...Flash?). It'd also be good to look for targeting mobile devices, depending on the type of game being developed.

It's also interesting to see the competing interests. Microsoft and Apple are trying to get vendor lock-in within their markets. Financially, it makes sense from their perspectives and we can see that evidenced in the API's and business models they release to developers. Developers, on the other hand, want to get as much market exposure to their products as possible so platform agnostic languages/API's let that happen easier. If Valve opens up their digitial distribution network to more platforms, its a big win for the developers who support multiple platforms and the consumers who don't necessarily want to be married to a particular corporation. I think that the long term effect of making the linux gaming market more accessible to developers via steam may cause a gradual shift away from windows and make it easier for gamers to say "yes" to running Linux. What I'm really curious to see is Microsoft's response to this move. Here's a list of the possible moves I see for Microsoft:
1) Ignore this (they probably can, to a large degree)
2) Attempt to saturate the desktop & gaming markets further with windows OS (to maintain and grow market share)
3) Continue to make their C#/XNA/DirectX API's the easiest to use compared to alternatives and make sure they have the shortest dev cycles
4) Release a competing digital distribution service (XBLA doesn't seem to compare very well to Steam)

I think the real meat and potatoes in the future is going to be about who can capture and secure the digital distribution market for phones and tablets. Currently, the apple app store is dominating the apple hardware and the Windows7 phone app store is pretty dismal. I think a MAJOR play would be if Steam gets released for mobile devices across multiple devices (iPhone, iPad, WP7, etc.). It could seriously undercut the app store and apple would probably try to cut them out of the market by denying steam on apple devices, so valve would have to pull a pretty tricky move to get market access (anti-trust lawsuit? business deal with apple? some other shenanigans?). It would be a huge win for developers and consumers though.

To be fair, with SDL you can easily cover about all I/O stuff a game may need (network aside, but on Windows sockets aren't very different), so that pretty much makes things a lot easier, especially for SDL 2.0 which is under the zlib license (not LGPL). Your only issue would be OpenGL... which is what I was worrying about, since that part doesn't work very well in the general view of things.

Honestly, I'd rather stab myself in the eyes with a rusty spoon than use SDL for anything... ever.

If this does go ahead, then I see one other nice benefit coming out of such a move: Valve supported APIs for Linux.

They're not going to do this out of the goodness of their hearts. If they make such a move, then they are going to push it, and they are going to supported, otherwise it is a wase of money to them that produces nothing. If no one is making games for them to sell on their service, then they won't be selling games over it. If people aren't making games because it is too hard for them, then Valve is going to have to step in and help make things happen.

As for windows and the walled garden? I doubt we will see massive changes in Windows 8 or 9 in that regard. Mostly because they are going to need to maintain some backwards compatability with existing software. If it becomes known that you can't play your not-exceptionally-old video games on the new OS, then you're likely not going to want to buy it. So complete walling off and limiting what can be run in windows, and where you can source your software from is going to be a long way off.

[quote name='Sik_the_hedgehog' timestamp='1339325742' post='4947878']
To be fair, with SDL you can easily cover about all I/O stuff a game may need (network aside, but on Windows sockets aren't very different), so that pretty much makes things a lot easier, especially for SDL 2.0 which is under the zlib license (not LGPL). Your only issue would be OpenGL... which is what I was worrying about, since that part doesn't work very well in the general view of things.

Honestly, I'd rather stab myself in the eyes with a rusty spoon than use SDL for anything... ever.
[/quote]
To be fair, coming from classic Allegro (i.e. before Allegro 5), I thought using SDL was going to be a pain. Turned out to be much easier than I was expecting, and this is taking into account I'm using raw SDL (with none of the helper libraries). In fact, I was expecting to go insane implementing sound output, and in the end I got it running in like no time. I'm having more issues with Vorbis files than with SDL...

will this influence your decisions regarding targeting Linux as a platform for your ongoing / future projects? If so, why or why not?

No. Multiplatform will always remain an important possibility, but only a possibility. No effort, for the time being, will be spent on removing platform dependancies. For the time being, if I look at my long-term roadmap, I have a component that mandates the use of D3DX. I have spent weeks in searching for a multiplatform replacement with no success.
It seems extremely likely Win will remain the center of operations for quite a while. Because of my past experiences with Linux, it will be left as a last target. Both Apple and Android seems much better candidates to me for the time being. Sorry dudes, that's my sincere decision after having used Linux for close to 10 years.

Honestly, I'd rather stab myself in the eyes with a rusty spoon than use SDL for anything... ever.

Agreed.

As for windows and the walled garden? I doubt we will see massive changes in Windows 8 or 9 in that regard. Mostly because they are going to need to maintain some backwards compatability with existing software. If it becomes known that you can't play your not-exceptionally-old video games on the new OS, then you're likely not going to want to buy it. So complete walling off and limiting what can be run in windows, and where you can source your software from is going to be a long way off.

Given Microsofts OS marketshare i doubt they're even able to force their own download service on consumers without being slapped with an anti-trust lawsuit so any such move is likely to be far in the future.

It seems extremely likely Win will remain the center of operations for quite a while. Because of my past experiences with Linux, it will be left as a last target. Both Apple and Android seems much better candidates to me for the time being. Sorry dudes, that's my sincere decision after having used Linux for close to 10 years.

Not to mention, it's very likely that the amount of Steam users that would migrate to Linux is probably going to be below 5% (heck, even below 1%, I dare say). Users will not switch unless forced to, which means that at best Linux may get some better drivers, and at worst it'll just be let to rot as it has been and eventually Steam for Linux may get withdrawn.

Now, if something serious happens that makes people go away from Windows, that would be more interesting... though I'm afraid PC users would be more likely to migrate to mobile than to Linux PCs, while for core gaming there's always consoles (assuming the console manufacturers don't go idiotic with the next generation, that is).

Given Microsofts OS marketshare i doubt they're even able to force their own download service on consumers without being slapped with an anti-trust lawsuit so any such move is likely to be far in the future.

They have already said they're going to do it for Metro apps though. The issue here will be what will they do with support for non-Metro apps.

While I believe that games on Linux will be the norm one day, I don't know how well Steam will actually do since the FOSS community is quite a bit more aware of the limitations of the online DRM deployed by Steam than the typical Windows user and are quite a bit more likely to boycott it.

Unfortunately other companies might see Steam's failure to take the Linux market as lack of demand which will ultimately cause more harm than good.