Support NeoGAF

[DreadPirateRoberts]

Steam on Linux isn't important. The games have to support Linux, too.

Oh I'm well aware of that, but this will be a massive step in that direction

Expect to be able to play L4D2 natively on Ubuntu this fall.

Amazing.

Edit:

"Not important" probably isn't the term is use for Steam on Linux. But yes, the games library needs to be built, which I'm confident it will going forward (see: humble bundles.) Linux Steam is the necessary first step for any kind of meaningful kind of Linux gaming library, and for that it is very important and even somewhat historical.

This guy gets it.

 

[TouchMyBox]

"Not important" probably isn't the term is use for Steam on Linux. But yes, the games library needs to be built, which I'm confident it will going forward (see: humble bundles.) Linux Steam is the necessary first step for any kind of meaningful kind of Linux gaming library, and for that it is very important and even somewhat historical.

So long as TF2 and Dota 2 are there, that's all that really matters, isn't it? :3

 

[Littlegator]

"Not important" probably isn't the term is use for Steam on Linux. But yes, the games library needs to be built, which I'm confident it will going forward (see: humble bundles.) Linux Steam is the necessary first step for any kind of meaningful kind of Linux gaming library, and for that it is very important and even somewhat historical.

I suppose it's true that Steam on Linux is important, but you understood my point.

However, Humble Bundles isn't an appropriate gauge, at all, for Linux support. PC ports generally only happen if the game was on the 360, as it had DirectX support.

There's no quick-and-easy way to port from PS3 or Wii to PC. There is from 360. That's because of DirectX.

The most important step for Linux support would be for non-MS consoles to adopt official releases of OpenGL, instead of their own modified APIs. Until then, no AAA developer is going to drop the extra dev time on supporting an entirely new platform. Adding Linux support from a PS3 title is directly analogous to adding 360 support. However, Linux support will add a meaningless market share.

Because of the nature of Linux and it's barebones, difficult-to-use structure, it will never be functional enough for a large audience to adopt it. This directly implies that major developers will not adopt the platform.

Mac OS, however, would be a very viable platform, and this is coming from an Apple hater.

 

[DreadPirateRoberts]

I suppose it's true that Steam on Linux is important, but you understood my point.

However, Humble Bundles isn't an appropriate gauge, at all, for Linux support. PC ports generally only happen if the game was on the 360, as it had DirectX support.

There's no quick-and-easy way to port from PS3 or Wii to PC. There is from 360. That's because of DirectX.

The most important step for Linux support would be for non-MS consoles to adopt official releases of OpenGL, instead of their own modified APIs. Until then, no AAA developer is going to drop the extra dev time on supporting an entirely new platform. Adding Linux support from a PS3 title is directly analogous to adding 360 support. However, Linux support will add a meaningless market share.

Because of the nature of Linux and it's barebones, difficult-to-use structure, it will never be functional enough for a large audience to adopt it. This directly implies that major developers will not adopt the platform.

Mac OS, however, would be a very viable platform, and this is coming from an Apple hater.

I don't really see what's stopping Linux from evolving into a more user friendly operating system as time goes on, it already seems like it's slowly on that path as it is.

 

[NervousXtian]

Expect to be able to play L4D2 natively on Ubuntu this fall.

Cool, always like playing 2009 games 3 years later.

..anyhow.. why bother? I honestly wish Valve would put their time into new IP or HL3 and improving Windows-based Steam... the client is pretty janky as is.

No interest in a Steam OS or playing games in Linux, and I doubt most devs would care to bother.

 

[Littlegator]

I don't really see what's stopping Linux from evolving into a more user friendly operating system as time goes on, it already seems like it's slowly on that path as it is.

It's pretty much the nature of Linux. A few distros are getting more user friendly, but if something breaks it breaks bad. No consumer market will buy into an OS that takes as long to fix (and needs fixing as often) as Linux.

 

[DreadPirateRoberts]

It's pretty much the nature of Linux. A few distros are getting more user friendly, but if something breaks it breaks bad. No consumer market will buy into an OS that takes as long to fix (and needs fixing as often) as Linux.

Well I think the existence of android kind of blows a hole in that. Even if you credit that to modifications made to the Linux Kernel, they have actually started including android code in the Linux Kernel.

http://www.xda-developers.com/android/linux-kernel-3-3-to-once-again-include-android-code/

 

[Littlegator]

Well I think the existence of android kind of blows a hole in that. Even if you credit that to modifications made to the Linux Kernel, they have actually started including android code in the Linux Kernel.

http://www.xda-developers.com/android/linux-kernel-3-3-to-once-again-include-android-code/

Android locks down root.

 

[kuroshiki]

Before creating OS they should perfect their steam client.

Steam client is not exactly well built software. Offline bugs, and there are some bugs that just crashes the client.

 

[DreadPirateRoberts]

Android locks down root.

Which can be unlocked quite easily. I'm not really sure where you're going with this.

 

[Littlegator]

Which can be unlocked quite easily. I'm not really sure where you're going with this.

The reason it's so widely adopted is because it's so locked down. Yeah, it can be unlocked, but it's by no means done by the majority.

Linux gives root access. Root access breaks things for the unknowing.

 

[DreadPirateRoberts]

The reason it's so widely adopted is because it's so locked down. Yeah, it can be unlocked, but it's by no means done by the majority.

Linux gives root access. Root access breaks things for the unknowing.

https://help.ubuntu.com/community/RootSudo

 

[Nilaul]

Sigh at al those comments from people that havnt used linux.
A lot of linux distros a very friendly. Mint particully is more user friendly then windows and its start menu. That's just a oppinion.

Now root access is like on windows, sure its gonna break things if you start deleting everything much like if you delete some windows files of your computer. Its easy to locate your program files, and any person with common sense what start deleting things that have any technical names. Its pretty obvious to understand which files shouldn't be deleted, in some cases these files make a warning sign if you try to delete them. Especially in newer distros.

 

[Shining Sunshine]

Why would they even want to? They should be on every platform.

 

[DreadPirateRoberts]

Sigh at al those comments from people that havnt used linux.
A lot of linux distros a very friendly. Mint particully is more user friendly then windows and its start menu. That's just a oppinion.

Now root access is like on windows, sure its gonna break things if you start deleting everything much like if you delete some windows files of your computer. Its easy to locate your program files, and any person with common sense what start deleting things that have any technical names. Its pretty obvious to understand which files shouldn't be deleted, in some cases this files make a warning sign if you try to delete them. Especially in newer distros.

I've been meaning to try out Mint actually, would you say it's a pretty easy transition from Ubuntu?

 

[hardcastlemccormick]

There's no way Valve would waste the manpower to make a Steam Linux distro. It would take an ungodly amount of time and effort and wouldn't pay off much. Gamers don't want it and there's literally no reason for it to exist.

It's pretty much the nature of Linux. A few distros are getting more user friendly, but if something breaks it breaks bad. No consumer market will buy into an OS that takes as long to fix (and needs fixing as often) as Linux.

Not true. Unless you're using a poorly built distro (Ubuntu is this way for some of its releases), most problems can be fixed pretty quickly by someone who knows what they're doing. The main difference is that because Windows is the most popular, it has the most help on the internet, and the familiarity is there. And OSX has Apple's excellent documentation.

The "needs fixing" is plain nonsense though, as is most of what you're posting about Linux in general. Can we please discuss the topic without you banging your fanboy war drum?

As for the market, no consumer market will ever install an operating system. Unless it comes pre-installed, you won't find it making inroads ever.

I've been meaning to try out Mint actually, would you say it's a pretty easy transition from Ubuntu?

Mint was Ubuntu. I used to be just a slight remix, then became more substancial. Even if they went away from Ubuntu in later releases (can't remember), it's definitely worth trying.

 

[Nilaul]

I've been meaning to try out Mint actually, would you say it's a pretty easy transition from Ubuntu?

Yes, if you remember that everything is under your start menu. Much like windows. It basicly a cross between Mac and windows, very light on your computer.

Soon android and linux will be completly remereged and running android apps will be possible.

 

[ThatObviousUser]

This is serendipitous since I got my new System 76 laptop for work today.

Who says Linux can't be pretty?

Spoiler

Nerdy specifics:
- Ubuntu 12.04 LTS
- Gnome Shell (Gnome 3)
- Adwaita Dark theme (Oxygen icon set)

 

[Vormund]

As already said; Windows Embedded.

In a nutshell, a version of windows that allows you to put in whatever components you need.

Windows Embedded Compact 7 would be the platform of choice, although a windows 8 version would be good considering how quick it boots up.

 

[Emily Chu]

NO

FUCK NO

don't make people dual boot this shit it'll be a fucking disaster

 

[Littlegator]

The "needs fixing" is plain nonsense though, as is most of what you're posting about Linux in general. Can we please discuss the topic without you banging your fanboy war drum

Err, wtf are you talking about? Or are you insinuating I'm a fanboy because it can take 5 to 10 hours to get a network card or GPU working when you know what you're doing?

Linux just isn't friendly to the masses. That's not going to change until hardware developers have a massive push for driver compatibility, but that's just not going to happen with so many distros.

And what the hell did I say that's nonsense? Everything I said is pretty much a matter-of-fact and not opinion.

 

[XtremeXpider]

This is serendipitous since I got my new System 76 laptop for work today.

Who says Linux can't be pretty?

Spoiler

Nerdy specifics:
- Ubuntu 12.04 LTS
- Gnome Shell (Gnome 3)
- Adwaita Dark theme (Oxygen icon set)

I have used gNome and I don't like it, in an screenshot it looks like Windows, and it is a "light" desktop, but it is not as good as Windows or OSX desktop.

Linux have a great problem as a desktop OS: fragmentation.

 

[whalleywhat]

Only if it's John SteamOS.

Cut it out.

 

[alexandros]

Why would they even want to? They should be on every platform.

One doesn't preclude the other. Valve could come up with their own Linux distribution for use by hardware makers and DIY enthusiasts. Valve provides the OS, the software and the spec sheet, hardware makers build the machine or enthusiasts assemble it and there you go: One Steam Box, ready to go! It would complement the other platforms, not replace them.

Edit: Plus, it would allow hardware makers to claim a piece of the console hardware pie, a road that is now closed to them.

 

[Vormund]

There's no way Valve would waste the manpower to make a Steam Linux distro. It would take an ungodly amount of time and effort and wouldn't pay off much. Gamers don't want it and there's literally no reason for it to exist.

If the guys working on XBMC can do it, I'm fairly sure Valve can.

I'm not saying it'll happen however.

 

[MrNyarlathotep]

PS3 and Wii do use modified versions of OpenGL, but a SteamOS pretty much makes developers learn another very specific API

Which is why most PC games use GFWL which is the same API as Xbox360 development, rather than a different one like Steamworks, right?

without allowing the performance optimizations of a unified hardware platform.

There are no performance optimisations for a unified hardware platform on the PC.

That's basically the entire point of the PC.

That's also on top of the fact that Valve would then have to maintain their own API.

Like Steamworks. The API they're already maintaining.

Regardless, DirectX is pulling ahead in capabilities, and we'll probably really start to see that next gen.

Pure speculation, and the move to the-UI-formerly-known-as-Metro would suggest otherwise.

Tesselation has been available in OGL for substantially longer than it has been available in DX11 for example.

There's no quick-and-easy way to port from PS3 or Wii to PC. There is from 360. That's because of DirectX.

Yes, if you're a developer and the time and effort you want to put into a PC port is:
makefile [360_code] [PC_MSI]

...and those end up selling real well on the PC, GFWL integration, press start to start, no using mouse on the menus and all.

If you're adding Steamworks support anyway - which increasingly more titles are - and you're changing your network, achievement, save & load files and DLC store / additonal content checks to using Steamworks anyway, changing your renderer includes to be a Steamworks OGL wrapper is not a huge endeavour.

The vast majority of time and work in games development is in asset creation, which is platform agnostic.

There's no way Valve would waste the manpower to make a Steam Linux distro. It would take an ungodly amount of time and effort and wouldn't pay off much. Gamers don't want it and there's literally no reason for it to exist.

I think most gamers would want it if it was done well;
- No more of that shitty installation of DirectX with every game you buy
- Generally faster performance in games; even ignoring the arguable render speed increases OGL offers over DX, having an OS specifically built to run Steam as a file manager / 'desktop' and one 'app' at a time, with friends / chat / webkit browser as optionla extras via the Steam overlay, you are still looking at an order of magnitude fewer concurrent processes than an idle version of Windows running nothing.
- Major new GPU features being rolled out without paywalling into a new version of Windows (as MS have traditionally used major revision DirectX versions as carrots to get gamers upgrading to the newest version of Windows)
- Free

 

[sazzy]

Given their client, would we even want one?

This.

My answer is HELL NO.

 

[alexandros]

Which is why most PC games use GFWL

Edit: Misread, never mind.

 

[szaromir]

Steam is a terrible client, why would anyone want a similarly terrible OS?

 

[Valnen]

That's basically the entire point of the PC.

I would argue that modding and all the other functions a PC can do are more important than playing with menus.

 

[DarkoMaledictus]

So this is total speculation, but let's assume that SteamOS is a real thing.

Would it work? Don't most games require DirectX to run? Is it even legal/possible for Steam to make DirectX work on Linux or would they have to use something like Wine?

I dont think SteamOS would work if it relied on companies creating games to natively run on it (at least at the beginning) so it would have to be Valve making the Windows games run on their OS, and if that's a really huge challenge, then this is a total nonstarter

Nah if that rumor would come to fruition total failure on the pc and VALVE goes bye bye... nah not in a million years.

 

[Omikaru]

I don't think you'd ever get 100% of the library over to Linux (which is what any SteamOS would be based on), but could you build something like Wine or some other implementation into Steam so that legacy games run well, with future games being made for multiple operating systems?

Yes.

I think you could get a significant portion of your Steam library running on Linux and possibly even MacOS, even if the developer/publisher doesn't want to port it.

Though I don't think things like Wine are viable for new releases, unless developers start optimising their games to run well on it, but even then that's not ideal. In the long term you want developers making games that run natively on Linux, with older games running in Wine bottles built into Steam, not all that dissimilar in effect to how we run older DOS games on DOSBOX and ScummVM, or older console games on their respective emulators (though before anyone chips in, I know Wine is not an emulator, and that's not what I'm implying).

Potential roadblocks are things like DRM, especially the likes of GFWL and Securom. I don't expect any GFWL game to ever work on Linux, for starters, and I expect there will be issues with getting DRM to work well because of how secretive they are, and how deep they embed themselves into the system.

 

[unrenowned]

Not a complete OS, no chance. Too much work for both Valve and 3rd party developers and not enough people willing to migrate. A store for the several major OS distributions is a big enough commitment.

Not to mention all programs people use that won't be available on Steam OS if it's a gaming OS only. I can't even imagine how many years it will take to get support from developers for all kinds of games/applications (thinking about the switch from XP to Vista and how long it took for some major applications to be made Vista-ready).

 

[Atolm]

Just license Windows and use an embedded version, and then put Steam on top of it.

Edit: Beaten by years.

 

[MrNyarlathotep]

Potential roadblocks are things like DRM, especially the likes of GFWL and Securom. I don't expect any GFWL game to ever work on Linux, for starters, and I expect there will be issues with getting DRM to work well because of how secretive they are, and how deep they embed themselves into the system.

Which is a good thing, because a title using Steamworks has no need for either of those things anyway.

 

[Angel Abaddon]

In the short term, turning round saying "Here's SteamOS!"? No. But I think all the recent developments of BPM, Steam Linux, Steam offering other software in addition to games and the rumours of Steambox, alongside Gabe's criticisms of Windows 8 are all funnelling toward a Linux Distro that is considered SteamOS.

I see the Linux client as a test bed of sorts, try and get people and devs on board and build the tech around a DX layer that people have mentioned here so the back catalogue can be offered. Should Steam for Linux fail, they can easily put it down as a failed experiment and shift onto something else, but I forsee Valve seriously pushing the Linux client far more than they have the Mac.

And for those asking why not just stick with Windows (Beyond the less than enthusiastic response Valve have had to Win8), a Linux distro built solely around gaming would likely give performance a boost I'd imagine.

 

[Littlegator]

Glad to see you don't know what you're talking about.

Which is why most PC games use GFWL which is the same API as Xbox360 development, rather than a different one like Steamworks, right

Porting netcode to Steamworks is a hell of a lot easier than rewriting your entire graphics engine.

There are no performance optimisations for a unified hardware platform on the PC.

That's basically the entire point of the PC.

My point was that they would have to learn ANOTHER API just to get the same performance as Windows, for the exact same consumer base.

Like Steamworks. The API they're already maintaining.

...wat

Steamworks isn't a graphics API. Having to maintain a graphics API probably takes more manpower than Valve has.

Pure speculation, and the move to the-UI-formerly-known-as-Metro would suggest otherwise.

Tesselation has been available in OGL for substantially longer than it has been available in DX11 for example.

Metro is 100% irrelevant. What are you even talking about? Metro is diminishing DirectX? wat

Anyways, the core of OpenGL isn't all too compatible with multicore processing, which is going to take a total revision to change.

Yes, if you're a developer and the time and effort you want to put into a PC port is:
makefile [360_code] [PC_MSI]

...and those end up selling real well on the PC, GFWL integration, press start to start, no using mouse on the menus and all.

If you're adding Steamworks support anyway - which increasingly more titles are - and you're changing your network, achievement, save & load files and DLC store / additonal content checks to using Steamworks anyway, changing your renderer includes to be a Steamworks OGL wrapper is not a huge endeavour.

Changing wrappers is simplifying it a bit too much. Wrappers pretty much only work with maybe DX9. DX10 and up are used so vastly differently from OpenGL that a wrapper isn't all that feasible. We're talking a rethinking of how it works at the core.

Not to mention that driver support for OpenGL is generally shit compared to DirectX.

I think most gamers would want it if it was done well;
- No more of that shitty installation of DirectX with every game you buy
- Generally faster performance in games; even ignoring the arguable render speed increases OGL offers over DX, having an OS specifically built to run Steam as a file manager / 'desktop' and one 'app' at a time, with friends / chat / webkit browser as optionla extras via the Steam overlay, you are still looking at an order of magnitude fewer concurrent processes than an idle version of Windows running nothing.
- Major new GPU features being rolled out without paywalling into a new version of Windows (as MS have traditionally used major revision DirectX versions as carrots to get gamers upgrading to the newest version of Windows)
- Free

Your bullets may be logical reasons to support it, but here's a logical reason you're wrong (and the only one necessary):
-gamers want Windows

No shit? Windows is the lead platform. The public loves Windows. The internet is the only place that Linux followers are vocal. Go ask the Steam audience if they would want to switch from Windows to use Steam. It's just not in the cards.

This is what I always get at in these threads. It's not in the cards. People don't want Linux.

 

[oneils]

That's running on Ubintu, its a really heavy system. Now imagine if you run it on a much lighter mint or another distro that runs with less then 256 mb of ram. The difference will be a lot larger.

people play video games with 256mb of ram? I am guessing your posts are mainly about the difficulty of getting older/legacy games onto a theoretical steamos that uses opengl?

Personally, I don't really see the point of a Steam OS. I think "Steam OS" is just going to be a GUI on windows or linux. If it is only on linux, I can't see how they continue to make billions.

 

[oneils]

Cut it out.

oh snap. took me a while...

 

[MrNyarlathotep]

Glad to see you don't know what you're talking about.

Well I know enough that different opinions on a subject don't equate to ignorance in a subject, so I'm apparently further ahead of the curve than you on that one.

My point was that they would have to learn ANOTHER API just to get the same performance as Windows, for the exact same consumer base.

...

Steamworks isn't a graphics API. Having to maintain a graphics API probably takes more manpower than Valve has.

...

Anyways, the core of OpenGL isn't all too compatible with multicore processing, which is going to take a total revision to change.

You keep asserting (in this and other topics) that the only way anyone could possibly create OGL renderers that compete functionally with DX would be if a company dedicated to doing so and with the express business model of charging for that work chose to do so, and that it could never be done for free, when this is demonstrably wrong.

Valve are already working on improving OGL rendering.
Intel are already working with them on it.
These are things that are already happening.

You also keep asserting that learning how to implement a new API is such a huge task that nobody could ever do it, but developers do it all the time.

Every game that uses middleware (which is every AAA title out there) has to learn new APIs.
Adding Euphoria to a title instead of Havok? New API.
Adding Twitter / Facebook functionality? New API.
Developing for a new console? New APIs.
Integrating Steamworks? New API.

You also make it sound like changing to using an OGL renderer core instead of DX would break every pipeline out there, when it does nothing of the sort.

If you're modelling, if you're animating, if you're scripting, if you're building levels, if you're designing audio, if you're creating AI, you're using the same damn tools you already use. What the compiled package uses as its includes doesn't matter.

Would it be trivial to recompile to use a different graphics API? No.
Would it be unfeasible to recompile to use a different graphics API? No.

Do developers already have forked engine branches and builds for different APIs?
Yes, they demonstrably do even if you've never actually stepped foot in a development studio, because the only platform that only uses DirectX is the 360.

Your bullets may be logical reasons to support it, but here's a logical reason you're wrong (and the only one necessary):
-gamers want Windows

No shit? Windows is the lead platform. The public loves Windows. The internet is the only place that Linux followers are vocal. Go ask the Steam audience if they would want to switch from Windows to use Steam. It's just not in the cards.

This is what I always get at in these threads. It's not in the cards. People don't want Linux.

Except you're ignoring iOS, OSX and Android, which are all built using a *nix kernel.

You might think Windows 8 is going to tear shit up in the Tablet and Phone market, but right now the majority of customers in those markets are very clearly saying they don't give a fuck about 'brand loyalty' for Windows, they just want something that works.
Why are those markets important? Because those markets didn't start out with a Microsoft monopoly.

In the tablet and Phone space, 'the public' seem absolutely fine with Linux. So why is Linux automatically impossible even in concept for mass market usage, when it's already there?

A SteamOS as hypothesised in the OP would just be a computer that runs Steam.
It would boot into Steam, it would be big picture mode, it would run games in your library, it would have chat / webkit based browsing / facebook posting and youtube uploading.

It would just cost nothing and run games faster without any hardware upgrades.

Are gamers going to go "FUCK THAT NOISE I LOVE WINDOWS" if that was the proposition offered to them? A proposition offered for free to boot?

You say yes. I say no.

I'd also say you must have never actually spoken to many people who work in an office in your life if you think "the public loves windows" and the only complaining about it is on "the internet".

 

[Basileus777]

Comparisons to mobile don't do Valve any favors here, the mobile market has shown just how incredibly difficult it is to create a viable ecosystem around an OS. Valve doesn't have the manpower and resources to really get it done. Valve getting into the OS or hardware business mostly sounds like a good way for them to lose money.

 

[ThatObviousUser]

I have used gNome and I don't like it, in an screenshot it looks like Windows, and it is a "light" desktop, but it is not as good as Windows or OSX desktop.

Linux have a great problem as a desktop OS: fragmentation.

Gnome 2 is Windows-ish but Gnome 3 is an entirely different beast and quite a pleasure to use.

Valve seems unfazed by the fragmentation issue, probably because it's so overblown. Aside from that, you can sidestep it entirely by focus in on Ubuntu, which is what Valve is doing early on.

 

[Tacitus_]

So Nyarlothep, who is going to foot the bill to port over the majority of the existing steam library? Because lets be honest, it's the huge library (and the sales of said huge library) that draws in most of the users. Who is going to convince the devs to develop to OGL over DirectX? Who is going to pay for the middleware solution ports?

 

[MrNyarlathotep]

Valve getting into the OS or hardware business mostly sounds like a good way for them to lose money.

I suspect most OEMs would love to sell a console style games PC that meets certain hardware specifications and doesn't require upselling a windows licence, and I suspect Valve would love to see hardware sold that can only access Steam.

EDIT:

So Nyarlothep, who is going to foot the bill to port over the majority of the existing steam library? Because lets be honest, it's the huge library (and the sales of said huge library) that draws in most of the users.

Why is it an all or nothing proposition?
Would the success of a SteamOS be predicated on every single title in the library working perfectly under Linux? (Particularly when there are a number of titles that don't work perfectly under Windows in the first place).

A SteamOS 'LiveCD' would be the first step, adding steamplay through steamworks as windows / mac / linux would be step 2.

Who would foot the bill for future releases? Publishers that like money.

 

[Tacitus_]

Why is it an all or nothing proposition?
Would the success of a SteamOS be predicated on every single title in the library working perfectly under Linux? (Particularly when there are a number of titles that don't work perfectly under Windows in the first place).

A SteamOS 'LiveCD' would be the first step, adding steamplay through steamworks as windows / mac / linux would be step 2.

Who would foot the bill for future releases? Publishers that like money.

Because people expect their games to work? Valve and some indie titles would have crossplay but what about the rest of the games? Who is going to install a gaming only OS with barely any games?
Why would publishers spend additional money to port their games to SteamOS? It's hard enough to get PC ports from some genres and now you'd want two different PC ports?

 

[MrNyarlathotep]

As I said;
- a bootable 'LiveCD' would be step 1. Nobody would replace windows.
- steamworks fully allowing steamplay would be step 2. That's the opposite of making '2 seperate ports'. That's making one port that works on 3 different target platforms. That's the sort of value proposition publishers fucking love.

EDIT:
Need a bigger incentive as a publisher? "Valve are proud to announce the release of Steamworks yadayadayada, and for a limited period any titles patched to support the new SteamOS we will waive our transaction fees, so all proceeds of sales go directly to you"

 

[Rolf NB]

Err, wtf are you talking about? Or are you insinuating I'm a fanboy because it can take 5 to 10 hours to get a network card or GPU working when you know what you're doing?

Linux just isn't friendly to the masses. That's not going to change until hardware developers have a massive push for driver compatibility, but that's just not going to happen with so many distros.

And what the hell did I say that's nonsense? Everything I said is pretty much a matter-of-fact and not opinion.

This is complete insanity. Hardware support is Windows' 100% total knockout weakness compared to Linux. You can do NOTHING without getting some random EXE from the internet and double-clicking it with your local admin user account. And every hardware vendor has their own installer methology that vomits all over your drive, their own control interfaces, their own taskbar icons, their own idiosyncrasies. I can't recount the tales of suffering I had to endure to get trivial shit like SATA controllers, sound, USB WLAN sticks, or indeed just plain ethernet cards to work on Windows machines.

All that shit just works, out of the box, hands-off, the moment you boot a recent (X|K|)Ubuntu LiveCD. Plus it's configured always in the same way, no matter who manufactured the chips or had them soldered on a board in China and put them in their own ugly boxes.

OTOH I can't even plug a USB keyboard into a running Windows 7 machine without it informing me that it's "Installing drivers", whatever the hell that means. Or use my USB stick plugged into the USB hub (hub works; stick works; just not in combination). Or try moving a harddrive to a different SATA controller, or *gasp* turn off EHCI emulation in the BIOS after Windows has been installed. Do tell me how that goes.

Now of course you mean graphics card drivers with 3D acceleration. They're in the Ubuntu repositories for Nvidia, Ati and Intel users, and pulled in by default. Which means they're already inherently easier to install and easier to keep current than whatever process you go through in Windows.

 

[Littlegator]

You keep asserting (in this and other topics) that the only way anyone could possibly create OGL renderers that compete functionally with DX would be if a company dedicated to doing so and with the express business model of charging for that work chose to do so, and that it could never be done for free, when this is demonstrably wrong.

Valve are already working on improving OGL rendering.
Intel are already working with them on it.
These are things that are already happening.

You also keep asserting that learning how to implement a new API is such a huge task that nobody could ever do it, but developers do it all the time.

Every game that uses middleware (which is every AAA title out there) has to learn new APIs.
Adding Euphoria to a title instead of Havok? New API.
Adding Twitter / Facebook functionality? New API.
Developing for a new console? New APIs.
Integrating Steamworks? New API.

You also make it sound like changing to using an OGL renderer core instead of DX would break every pipeline out there, when it does nothing of the sort.

If you're modelling, if you're animating, if you're scripting, if you're building levels, if you're designing audio, if you're creating AI, you're using the same damn tools you already use. What the compiled package uses as its includes doesn't matter.

Would it be trivial to recompile to use a different graphics API? No.
Would it be unfeasible to recompile to use a different graphics API? No.

Do developers already have forked engine branches and builds for different APIs?
Yes, they demonstrably do even if you've never actually stepped foot in a development studio, because the only platform that only uses DirectX is the 360.

I was talking about needing to charge to maintain a fork of OpenGL, not OpenGL itself. It's totally possible to just use OpenGL, but in the other topics people were specifically talking about a new renderer.

You say I'm asserting that "learning how to implement a new API is such a huge task." No, it's not. Implementing a new API is such a huge task. It's not like the programmers learn how to use OpenGL and BOOM IT'S DONE. It's thousands of hours to write a rendering engine, and to port anything from DirectX 11.1 to OpenGL 4.3 would require a major rewrite. It would take a restructuring of a lot of the engine. It's not on the scale of thousands of hours, but it's a solid chunk of time and money that just plain isn't worth it. It's not just "compiling with a different renderer" like you're insinuating. If it was so easy to ship with OpenGL, people would do it.

Except you're ignoring iOS, OSX and Android, which are all built using a *nix kernel.

You might think Windows 8 is going to tear shit up in the Tablet and Phone market, but right now the majority of customers in those markets are very clearly saying they don't give a fuck about 'brand loyalty' for Windows, they just want something that works.
Why are those markets important? Because those markets didn't start out with a Microsoft monopoly.

In the tablet and Phone space, 'the public' seem absolutely fine with Linux. So why is Linux automatically impossible even in concept for mass market usage, when it's already there?

A SteamOS as hypothesised in the OP would just be a computer that runs Steam.
It would boot into Steam, it would be big picture mode, it would run games in your library, it would have chat / webkit based browsing / facebook posting and youtube uploading.

It would just cost nothing and run games faster without any hardware upgrades.

Are gamers going to go "FUCK THAT NOISE I LOVE WINDOWS" if that was the proposition offered to them? A proposition offered for free to boot?

You say yes. I say no.

I'd also say you must have never actually spoken to many people who work in an office in your life if you think "the public loves windows" and the only complaining about it is on "the internet".

I'm ignoring iOS, OSX, and Android because they're against the spirit of nix. Nobody cares that they're nix, and nobody is jumping on the nix bandwagon because of them. They're locked down, proprietary operating systems that serve a purpose directly contradictory to the ideals of the GPL.

I think Windows 8 will flounder in the tablet and phone space until Microsoft rallies up their support for parallelism between devices. That's a long ways off, and anyone who thinks they can make a prediction that far ahead is just kidding themselves.

Once again, the public is fine with them in the tablet and phone space because they're lightweight, easy to use, locked down (read hard-to-break) interfaces. They're not using their tablets and phones to do most of their work, they're just using them as a connection to social networking and texting. If they had it on their desktop, they would hate it. Or are you saying you LIKE Metro, now?

Are gamers going to go "FUCK THAT NOISE I LOVE WINDOWS"? No, that would be stupid. It's not like they drink Windows' kool-aid, they'll just make a logical decision. They'll look at SteamOS, look at all the trouble of partitioning, installing a bootloader, dealing with hardware issues (because no matter how much you praise your package manager, hardware never just works in Linux), and dual-booting, and they'll give it a pass. Yes, they'll give it a pass "for free", because it still comes with major, major inconvenience.

And yes, I've never talked to anyone in an office. Like none of those 21,000 employees in my office. I've never talked to them. None of them in the software development division that I work in. I don't talk to those people.

Everyone there prefers Unix for development. It's easier to develop in. We work in embedded systems. It's reasonable. However, other than the few who follow the Church of Linux Christ, they mostly all use Windows as their primary operating system at home. Why? Because it's familiar, it's convenient, and it works. It's the same goddamn reason the rest of the world uses Windows. In fact, the developers use Windows on their workstations, and use Visual Studio whenever possible.

Honestly, it's tiring how many people preach the glory of Linux. It's not easy to use. Get over it. You know that it takes learning to use. And it doesn't just take some learning, it takes a shitton of learning. Linux is for the experienced. That's how it is.

This is complete insanity. Hardware support is Windows' 100% total knockout weakness compared to Linux. You can do NOTHING without getting some random EXE from the internet and double-clicking it with your local admin user account. And every hardware vendor has their own installer methology that vomits all over your drive, their own control interfaces, their own taskbar icons, their own idiosyncrasies. I can't recount the tales of suffering I had to endure to get trivial shit like SATA controllers, sound, USB WLAN sticks, or indeed just plain ethernet cards to work on Windows machines.

All that shit just works, out of the box, hands-off, the moment you boot a recent (X|K|)Ubuntu LiveCD. Plus it's configured always in the same way, no matter who manufactured the chips or had them soldered on a board in China and put them in their own ugly boxes.

OTOH I can't even plug a USB keyboard into a running Windows 7 machine without it informing me that it's "Installing drivers", whatever the hell that means. Or use my USB stick plugged into the USB hub (hub works; stick works; just not in combination). Or try moving a harddrive to a different SATA controller, or *gasp* turn off EHCI emulation in the BIOS after Windows has been installed. Do tell me how that goes.

Now of course you mean graphics card drivers with 3D acceleration. They're in the Ubuntu repositories for Nvidia, Ati and Intel users, and pulled in by default. Which means they're already inherently easier to install and easier to keep current than whatever process you go through in Windows.

Maybe you wouldn't have USB issues if you didn't fuck with EHCI emulation. Just a guess.

And you know that it's complete horseshit that Linux works out of the box. That might be true for manufactured PCs, but it's definitely not true for piecewise PCs, you know, the ones PC gamers have.