This is exactly why I decided to go through the devs and the HumbleStore. I wanted the linux version, same thing with Don't Starve. I like gog and would have rather helped them out, but it was kind of a deal breaker for me that I wouldn't be able to get the Linux builds that the developer was already making.

Aa bit of a slap in the face indeed. I do wonder how well the Mac OS venture has actually paid off for GOG. It's great that they finally support an operating system that is not part of the Windows family, but how many Mac OS users are there on GOG as opposed to Linux? Yes, Mac OS may have more market share than Linux on a global basis, but where are the gamers at? Going by the GOG Cmmunity Wishlist, there seem to be at least 6.68 times as many willing Linux gamers as opposed to Mac OS gamers. Even if it's not more than 10% of GOG's user base, would you still ignore 10%, Magnitus?

There are plenty of other games I hope to buy for Linux and off GOG, some of which GOG don't even have the Windows clients of yet which I find strange but can nevertheless wait for e.g. Trine 2 and Amnesia: The Dark Descent. I don't want to buy my games from too many different places, and GOG has already won me over with its ways.

Yep, I'd prefer to get those form GOG as well, though I got Amnesia: The Dark Descent for Linux from Humble Bundle already.

I think they would be foolish to ignore 10% of their customer base if it was that amount, but I don't think it is.

I do agree it's a shame that they don't at least provide a Linux version for games that support it out of the box.

Maybe they are worried that if they do just that and nothing more, Linux supporters will be even more annoyed (all or nothing type of situation).

At least it's an unified API allowing game developers to build games which they can trust that they will run on their customers rigs, can be sold for years (API/ABI stability) and will continue to run for decades (platform support).
DirectX (+windows) is a stable game API/SDK/platform where linux failed to create one on its own (I know there are bazillion of concurring own liberaries ...still no reliable platform).
Therefore it makes sense to see Wine as unfiying API extension for the linux ecosystem and should not receive that much hate. This point was mentioned also by codeweaver on the Limbo port shitstorm http://www.codeweavers.com/about/blogs/jwhite/2012/06/05/whining-about-wine

You want a free and open source OS with potential as gaming platform? What about ? [url=http://www.gog.com/wishlist/site/add_haiku_versions_of_games]http://www.gog.com/wishlist/site/add_haiku_versions_of_games (or even ReactOS, optimal compatiblity for GOGs back catalog and still free and OSS )

@shaddim: No, DirectX is a lock-in junk, which people have hard time getting rid of, since it has deep market penetration. OpenGL is as stable and is cross platform at the same time. So stop talking about ABI stability advantages - there are none in DirectX.

Wine is not a unifying API - it's a crutch for the case when no native version is available.

OpenGL is nice...but not complete multimedia SDK like DirectX. OpenGL+SDL+OpenAL -"fragmented distro ecosystem" would be comparable with a DirectX/Wine.

SDL and OpenAL are stable as well, not less stable than DirectX across the same release.

So yes, for any game developer OpenGL+SDL+OpenAL should be the obvious choice.

I'm talkign about the instability of the rest of the linux ecosystem. If games are "virtualized" in a distro-agnostic "exe-container" + stable api (Win32+DirectX), they might be not as hard as native linux games hit by the ecosystem problems. Things like the lib varitation game between distros, packageformat madnesses, flaky driver support, deployment problems as it's binary not adressable, and a general long-time support problematic (no, I dont' care for specific LTS distros...)

Also, I missed in my API list a capable and unified window/GUI/desktop API (like win32).

Talking about the pros of DirectX, DirectX provides now since, lets say DirectX 5, 17 years of binary comaptiblity. THIS IS f**king AWESOME! DirectX 9 is now around since 11 years and still strong as standard for game development. So strong that tries of MS to get rid of it failed (since Vista 2006). This is really impressive success story of a great API. I confess I like DirectX, brought great games to the PC. :)

Most of the issues you mentioned are irrelevant and solved by stable versions of libraries provided (OpenGL, SDL and etc. also can support several versions back for legacy stuff). Packaging - non issue. Drivers? They can be a problem anywhere - Linux is not unique here and this has nothing to do with "ecosystem". Modern drivers are improving though, even for Intel integrated graphics. So, can we please stop the discussion about DirectX - it's going nowhere.

DirectX 9 lingers because Windows XP doesn't support DirectX 10+ (due to technical limitations). It has nothing to do with any strengths of DirectX 9 itself. DirectX 11 is superior in every area (including hardware support--developers can optionally support a partial DirectX 11 feature set on DirectX 9/10/10.1 hardware).

DirectX 9 is not still strong nor is it still the standard; an ever-growing number of new releases require DirectX 10+ and this will become a general requirement for next-generation multi-platform games. DirectX 9 will be gone in a couple of years.

DirectX 9 should be dead since several years, still it is alive and kicking, a common denominator. The importance of stable defacto standards can't be overestimated.

Ubuntu is attempting to address the non-uniformity of the UI stuff with their SDK. Of course, it's called Ubuntu SDK for a reason. LOL.

OpenGL + SDL + OpenAL is clearly the way to go, but I suppose developers are no longer willing to invest in building their tools from scratch.

Unreal Engine was ported by Ryan Gordon, but so far, Epic likes to ignore that it actually exists. Unreal 3 Linux ain't coming out and that's a damn shame.

Yeah this would be still only adress ubuntu, a isle solution. And the fragmentation inside the distro is also not adressed (https://bugs.launchpad.net/ubuntu/+source/software-center/+bug/578045). One problem is , Ubunutu really tries to bring in "Ubuntu" as sole solution for linux desktop problems instead of developing on solutions leading to a fixed ecosystem overall.

The way to go should be: strong LSB platform definition allowing ISVs to deploy apps for the ecosystem (enforcing ONE GUI-SDK, ONE-multimedia library etc), distro-agnostic deployment methods (l, [url=http://portablelinuxapps.com]portablelinuxapps) and in general a clear commitment on binary and API/API stability on all levels (kernel, userland, etc). In short, get rid of the negative consequences of a highly fragmented distro & developer focussed ecosystem.

But I doubt the community and distros are really willing (or able, last 10 years told as not so) to take such hard changes in their ecosystem. Therefore maybe the only way to go is to select from the beginning a desktop OS system already fitting better: , [url=http://www.reactos.org]ReactOS.