No, the most popular Linux distribution is not the irrelevant part. Try again.

Something weird is going on with the forum - it swallows posts. Android is not a Linux distribution, and isn't comparable to them. So again - open another thread if you are interested in mobile gaming. This thread is about normal glibc based desktop Linux.

For your reference Android is not using glibc, it uses bionic libc, and as such is completely incompatible with normal Linux.

Oh, your fragile elitist beliefs are so fragile. :(
But I'm about to watch Dr. Who so you won't have to put up with any more truth from me for a while.

Peace.

While you have extra time, spend it on learning about the difference between serious OS like GNU/Linux, and mobile toy like Android.

Unfortunately, Linux offers the same features as Windows in most arenas besides gaming, yet people still mostly go with windows for their home or office computer.

To a large extent, Linux is not KISS enough for the average non-geek user (the million of distros out there being the first thing people learn about Linux is not helping... too much choice can be just as harmful as not enough in most circles).

Also, corporations actually like to have a sizable company tied to the OS they are using which they can contact for support.

Thanks for the pointer.

If you try to support Linux with a capital L (considering it one fragmented platform), that is a problem.

However, I think it would be feasible to support a popular subset of the entire Linux ecosystem.

Dependencies, that may or may not be present on the user's system, can be included with the installer as is the case with Windows installers already (I recall one game installing VS 2005 components).

I agree that people re-inventing the wheel in the open source community can be a pain at times. Most of the time, the competing open source technologies are very similar and you'd just like to go with the one that will have the most long term traction, though that is often hard to determine at first.

However, it is worth noting that the private sector is guilty on that front too (ex: the C# and objective-C programming languages... were they really necessary?)... the problems are just shouldered by developers instead of end users (but the end user still gets penalized as it becomes harder for developers to deliver a consistent and platform independent experience). Yes, nobody, but the company who owns the platform can develop official content for their platform so it creates a well integrated experience for the end user (and programmers who want to develop only for that platform), but it gets in the way of developers who just want to create software the works universally across all OSes.

That is a problem for Canonical (as they want a wider adoption of their distro), but from GOG's perspective, that would mostly be a non-issue.

GOG users consider GOG a trusted source, so they don't need GOG games to be added to a trusted repository to trust them (and thus, buy and play them on their Ubuntu distro).

Hmm I recall reading that DX11 did indeed have a fallback and would use DX10 code paths if they were available for backwards compat. However DX10 had the same for DX9. Since DX11 is a superset of DX10...

Hmm, might do some testing when/if I have the time

But then developers would have to develop their own code instead of using the code provided by DX11 APIs and whatever.

So again, it is a trade-off, whether you base a game on DX9 and develop your own code for certain things instead of using already available code in DX10 or DX11, or base a game on DX11 and use all provided code and just develop your own code for things available in DX11 that are not available in DX9 or DX10.

To clarify, I'm not talking about the developers manually implementing equivalents of the unsupported features but rather using the the feature levels to intelligently scale down/disable features based on hardware capability. One such example is games that let you use the DX11 renderer on a DX10 card but disable tesselation-related settings.

The feature levels concept could potentially make it easier for Wine developers to implement DX10+ support but this will probably depend on how well the game developers actually made use of it in the first place (and it probably still wouldn't work with Crysis 3 and any other future games that require the DX11 feature level).

Divinity: Original Sin is probably the newest example of this. A Linux client is definite and the game coming to GOG is also definite, but will the Linux client of the game come to GOG? Nobody knows. I would have loved to kickstart the game if I had been certain that I could get the Linux version of the game on GOG, but alas.

Divinity: Original Sin - just buy the game straight from Larian when it comes out. No point to wait for GOG starting Linux support for that.

The Witcher 3? Please. I want the entire The Witcher series as well as Cyberpunk 2077. Imagine modern games of that magnitude on Linux! The Linux community would go nuts. I know I would! GOG could certainly make up for their ways by releasing the entire The Witcher series for Linux when the third and final game in the series launches for all platforms. That would be a blast!

I doubt there is much point for them to work on Witcher 1 and Witcher 2 ports since they already decently work with Wine (though Witcher 2 might require minimal work to port - it already supports OpenGL in the Mac OSX version).

Witcher 3 and Cyberpunk 2077 however are a different story, since they will be using DirectX 11 and won't be playable with Wine (until Wine project catches up with DX11 support). If CD Projekt Red can release native Linux versions from the start - that would be really amazing and they'll go like hot cakes.

This is why I hate proprietary systems on computers. DirectX being the worst since it seems to lock games into Windows. I also wish Unreal Engine would get on board and support Linux like Unity and Steam are now doing.

I bought Faster Than Light from GOG but I've never played it because I'm on Ubuntu most of the time. I sure could reboot on Windows (and I do for many games) but it feels quite ridiculous considering a Linux version exists (and I'm reminded about that each time I launch Steam). The fact is that I just want to be able to start the game whenever I'm in the mood for a "quick" run and go back to what I was doing as soon as I'm done playing. Yet I can't (with that said I never tried it with Wine, but again why would I do that if it can run natively?)

There are numbers of other indie titles I would buy without any hesitation on GOG if a Linux version was available.

The funny thing is that I got Sanitarium from here a few years ago but the game always crashes at the beginning of the first level when running on Windows XP, and many people encountered similar problems. From what I know, GOG never released a definitive fix (so much for the support), but the community was here to propose workarounds. Alas, they didn't work for me.

Yet I'm still glad this excellent game was released on GOG (and paradoxically I have it running on Unbuntu without any problem with PlayOnLinux).

What I'm saying is that the same would happen if Linux versions were made available: even if GOG is not able to provide full support, people would be able to help each other.

Steam made the move to Ubuntu, so why not GOG? If a game simply doesn't work and people want their money back, so be it. Just make it clear from the beginning that the game will run better on some distros than others and there would be no one to blame.

Or - as Kristian suggested - present Linux versions as bonuses that are not officially supported.

Anyway there may not be one perfect solution but there are many ways this could be done, and be assured that I would buy far more games on GOG if it was embracing Linux :-)