Ah, nice sleuthing! Looks like we all had a similar discussion about a year ago about the API also that I don't recall having but more or less said the same things then too hehe.

That's incorrect. They quote response from GOG support - wrong source for this kind of information. Also, you don't quote them precisely. All they say, GOG never planned to release the Linux version (naturally, since it was never completed). They didn't say anything about planning the porting itself. That question only developers can answer.

Plus GOG support already clarified the situation again, saying that they aren't in position to know such plans. See here.

I hope sometime after or during Cyberpunk 2077 they begin with Vulkan development. As much as I like to see Witcher 3 on Linux I want to see CDPR to divert their attention to Vulkan first.
If they adopt Vulkan then it will increase the chance of W3 and other games coming to other OSes.

Yes, if they'll use Vulkan in CP2077, they'll have much easier time releasing it for Linux. But it won't help TW3 anymore. It would be the newer iteration of their engine, and backporting it into TW3 would still be a major task.

Updated the first post with interview with Marcin Iwiński, where he says, that if [Valve] will create SteamOS, they'll be interested in releasing their games for it (since it would be a stable Linux target).

I believe that to be common. However, Galaxy is advertised as being 1.1 beta. Is that marketing speak? Unless they are Google (GMail was 5 years in beta status) that suggests to me that their API is kind of stable. And the API I am talking about is specifically the Client/Server API used for authentication and communication, not the local libraries. I believe that one to be more established, as any change there must be agreed by two teams of developers.

I believe you may have more experience on these matters than I do, and perhaps I may come to change my views on this matter with time. But to external, 3rd parties developers, just the library, download and upgrade API is enough to get stuff going (I should have made this more clear, I'm sorry about that). Everything else is quite secondary for me.

Some people are only looking for some reason to complain. You have some in group A and some on group B. You will never get away with anything, no matter what you choose. But staying in the middle... you may get exposed to both groups.

I don't need Galaxy. However, the API could provide a convenience to write a tool that would integrate Wine and other nice things on my Linux system. LGOGDownloader, gogrepo and PlayOnLinux are already available, but more would be better. :-)

Well, GOG may know what it is doing, but I still think that more transparency or better communication would be better appreciated.
Who is he? I'm sorry, those names don't mean much to me.

CEO and one of the two biggest shareholders in CD Projekt: https://en.wikipedia.org/wiki/CD_Projekt#Founding

Agreed maximally. I think every game developer should be doing this whether they are using a 3rd party engine for Vulkan support or adding it to their own engines. There's no really good reason not to that I can think of at this point, as all the major video vendors are supporting Vulkan now. The capabilities of Vulkan and DX12 are basically identical for the most part, but whereas DX12 is not only a Windows-only API, it is furthermore a Windows-10 only API limiting games that use it to one specific version of Windows to get the functionality it provides, and alienating all users using older releases of Windows or alternative OSs.

Using Vulkan means it will work on all versions of Windows and also on Linux as far as the 3D API side of things is concerned. It may even also work on Mac some day if Apple ever decides to supports it, but Mac is no worse off either way.

Prior to Vulkan's official release along with driver support it could make some sense why early adopters of next-gen 3D API technology might have experimented with DX12 as it was available, but now that Vulkan is also available, just as capable and I believe more capable for that matter, supports a wider variety of platforms, it makes sense for it to be the go-to API to provide that next-gen performance to the widest variety of consumers.

Yeah, both Vulkan and DX12 are majorly different APIs compared to OpenGL and DX11 and older. Games and game engines that haven't been retrofitted to support either yet are probably not likely to see any incremental updates that add support for either API as it is a major effort if it wasn't baked into the original design.

I highly doubt that we'll see a future Witcher 3 patch add Vulkan or DX12 support as an incremental feature patch as the work effort to do that would be rather major and potentially serious risk of destabilization of a product likely in its latter stages of product support.

If they were ever to consider a Vulkan or DX12 version of Witcher 3, I believe it would be something potentially to be entertained a much longer time period from now, perhaps as a remaster edition N years from now post Cyberpunk 2077. Presumably they're revving the Redkit Engine up for CP2077 and adding support for Vulkan and/or DX12, so perhaps in 5 years they update Witcher 3 to use the CP2077 era Redkit Engine and pick up Vulkan+DX12 support along with enhanced graphics or other goodies. I don't see it happening sooner than that though.

If they decide to put out a Linux release of the game in the future, it will most likely be OpenGL based if anything, as I believe it already supports that for Mac.

Sadly, only an extremely small number of games have even announced planned support for Vulkan at this point in time which is somewhat disconcerting IMHO. Only 4 games total:

https://en.wikipedia.org/wiki/List_of_games_with_Vulkan_support

I don't worry about that, since all major / widely used engines are on track to support Vulkan. As soon as they'll be production ready with it, Vulkan usage will naturally increase quite a lot. More exotic / custom engines is a different story, but they'll end up doing it eventually, simply for competitive reasons.

Yes, remaster or enhanced edition was something that was expected to bring back pre-"downgrade" look together with usage of new graphics APIs. But CDPR indicated they aren't interested in it anymore, and their "game of the year" edition will be the last one for TW3.

Vulkan for Mac is called MoltenVK: https://moltengl.com/moltenvk/

Like shmerl said, I would not worry about games. Worry about engines and tools and consumers being informed of Vulkan. One aspect of Vulkan taking over is that gamers aren't well aware of it.
Take World of Warcraft, it doesn't have DX12 implemented. So maybe just maybe Blizzard might implement Vulkan, this way they would allow XP, W7 and W8 users to be able to use great tech of Vulkan and also Linux users to play WoW natively.

Another thing to worry about is WXP 7 and 8 users, if over course of a year or two those upgrade to W10 then Vulkan will be cornered.
Based on Steam hardware and software survey there is a total of 45% users who play using a Windows that's not W10.
That's a huge number to persuade developers to go with Vulkan over DX12. If that number decreases the weight of Vulkan over DX will lessen.

So as gamers we need to be proactive and promote Vulkan to fellow gamers. Upload youtube videos informing people of why Vulkan is important and other things.

I like to think Vulkan usage will uptick also, and I hope it does, but the statistics don't look good so far at least. It's far too soon to draw any conclusions so I remain hopeful but it still remains to be seen.

For The Witcher 3 future editions though, if you look at every game released to date which have anniversary editions, HD remasters, REDUX releases and other similar re-releases that have varying degrees of new features in them I think if one were to travel back in time to the original developers 3/6/12/24 months after the release of their game and ask them about a future enhanced edition or remaster that many of them will say "No, we have no plans for a remaster or enhanced graphics or anything other future re-releases of our game." because they simply do not think that far into the future or have a specific plan to do that at the time they're asked. However based on a game's popularity, the demand and desire from the fans of the game, the financial situation of the company itself and a variety of other factors, as the years pass on by, companies look over their existing IP and from time to time decide to do a re-release of their game updated to work on modern operating systems of the day, possibly adding widescreen, higher resolution support, HUD scalability and a variety of other improvements and enhancements and then go ahead and release it if they think there is a market for it.

A perfect example of this is Skyrim Special Edition set to come out any month now. The game is about 5 years old and already getting an enhanced Special Edition. I don't think anyone can say categorically that this isn't something that can happen for any game even if the developers have no plans to do such a thing at the moment. It's not something that they necessarily plan for all along the way, but rather always a possibility for new revenue off of an existing title if they think it is worthwhile to pursue and there is a market for it.

So I wouldn't rule out any possibility of seeing an enhanced version of The Witcher 3 some time in the future. I doubt it would be any time im the immediate future as that would be far too soon IMHO, but in 5/7/10 years, that's an entirely different story and definitely a possibility. :)

I pretty much agree with everything you've said above. In the immediate timeframe Vulkan has a high degree of advantage across the board, however as time goes on and the pre-Win10 editions of Windows drop out of favour in PC gaming, the potential customers that would directly benefit and require Vulkan support to get the full experience will drop and DX12 usage becomes less and less of a problem. The number of Linux and Mac users out there are not anywhere near enough to have a big impact on that kind of decision, but the number of people still using Windows Vista/7/8.x are definitely large enough to be important and not ignored right now.

Vulkan is to next-gen 3D graphics as HTML is to an open web accessible by anyone on any device using any browser. Everyone should read up on it a bit just enough to understand how it is a standard worthy of supporting as you suggest and raising awareness of. OpenGL suffered years ago in this regard because the ARB was too slow to adopt to emerging hardware functionality trends and Microsoft came along with D3D and were able to make the hardware functionality available for use by developers immediately without the delays naturally inherent in the slow drawn out processes that OpenGL enhancements go through. As such, OpenGL in gaming fell out of favour more and more over the years.

With the new next-gen 3D interfaces though things are very different because the APIs are much more low level and unlikely to need many changes over time for new hardware emerging. Vulkan, DX12 and Mantle are more or less all doing the same thing for the most part and essentially on par. The problems that held back wider OpenGL adoption in gaming don't really exist for Vulkan, so it'd be nice to see Vulkan get more traction. Not sure what all major game engines have adopted it so far, although I know Source has so games using engines like the Source engine are more likely to use Vulkan.

For games that are small teams or indie developers which have written their own game engine however, it is much more likely that they'll use Direct3D or OpenGL because the amount of effort to use Vulkan or DX12 directly is significantly more due to its low level nature, so those type of games will probably stick with the older interfaces for some time to come I believe.

See https://trello.com/b/gHooNW9I/ue4-roadmap

Check Vulkan entries.

Some people think that Vulkan is the magic bullet of PC gaming. No, it isn't. It's just another API, which in some cases can give you superior performance, but at the cost of writing a lot more code. In some places it might make sense to use Vulkan, but in other places plain old OpenGL/DirectX will stick around for some time. Just think about the effort of re-writing a large game with Vulkan when drawing a simple rotating triangle, which you can do in OpenGL in around 15 lines of code, you need 2000 lines using Vulkan. So developers will have to choose and optimize taking into account the growth in complexity.