Redfern: "No, they should wait until they internally feel that the Linux platform is ready for them" you said? i'm sorry, but i feel like you gettin it somehow wrong. First, as you should know, there is no "official linux distro", there is a lot of them, all different.
As a former systems engineer of a major Linux vendor I'm well aware of the differences between Linux distributions. ;oP
That itself is a major problem for gaming companies and distributors and as long as they perceive this to be a problem, it holds back gaming on the platform. When I speak of "the Linux platform" I speak of it in these generic terms that are all encompassing of the majority of distributions that are relevant for gaming on Linux to be successful and not just for 1 or 2 of the most popular distributions.
Redfern: Second, there is not such task in distro developers as "make linux GOG compatible". Games not going to get any special treat.
I haven't suggested anywhere that there is such a task at hand by any distribution vendors either. In fact, there very much isn't any such kind of concerted effort. The effort is going to come from the outside, such as Valve's Linux efforts which are motivated by their desire to build a profitable business model on top of the Linux platform. In doing so, Valve is doing a first handed look into the problems they perceive that exist within the Linux platform and trying to solve those problems and make contributions back to the Linux ecosystem which ultimately will find their way back into all of the major distributions over time as well. Valve is motivated by their business interest and the flexibility of Linux to accomplish their goals, and the work that needs to be done is only likely to come from a company or group of people who are sufficiently motivated in this nature as it is unlikely that major distributions are going to be motivated to do such a thing as there would be no major payoff for them personally up front. For distributors it would be all risk, no perceived gains. For Valve however it is a different story, the risk is acceptable to them and they believe they have the resources to pull it off to a profitable end and everyone else will benefit from the cascade effect that follows.
Redfern: Third, Linux already have all it need to "support" games. It have OpenGL, it have packet managers, main videohardware official drivers. That ELSE gog waiting for?
If only things were that simple. Unfortunately they're not that simple. The X Window system and video drivers are incomplete and a frequent source of instability and unreliability (both open source and proprietary), which is my area of personal expertise (I am a former XFree86/X.Org X server and video driver developer and can state this with some degree of authority and no other X developer will be likely to disagree with me about either). Support for various input devices is also limited with various hardware as well. Audio support is lacking for a lot of hardware is lacking similarly to video and often has problems with latency and other issues. The pulseaudio system which is widely used is still immature and has a long way to go yet. None of the other audio subsystems are any more mature either currently.
Things continue to improve in all of these areas over time, but new hardware comes out just as quickly leaving support always lagging behind the current best efforts of the open source community and hardware vendor contributions to track it all. This means that people definitely do have and will have problems that are quite varied on different hardware across distributions, and it means that the support for specific hardware can vary greatly from one distribution to another and one installation to another.
OpenGL support in the open source drivers is also lagging behind the current standards. Mesa is currently supporting OpenGL 3.x if memory serves correct while current is 4.x, and this is probably going to continue to be the case for the forseeable future. That does not prevent games from coming to Linux so long as they don't hard code requirements of newer OpenGL features, but it does mean that less features are present where people are using the Mesa stack for OpenGL. The proprietary drivers are probably much more current with OpenGL specs however, but I'd have to look that up as I'm not up to date on where things are with ATI/nVidia for that.
Other areas where problems exist that affect developers are in networking (less of a problem), with desktop interactivity issues (a long standing problem to which I have spent time working on myself). These type of issues vary from distribution to distribution as well depending on what kernel is provided and how it is patched and what the default configuration is. Some are affected more than others.
All of these issues and many more I've not enumerated make Linux actually a collection of very similar platforms rather than one single unified platform. So while we can speak of "the Linux platform" in a singular sense and mean something specific from it, the mechanics vary from actual distribution to distribution a fair bit and that has a concrete affect on what it takes for a 3rd party game or applicationi developer to provide their software on the platform and have it work uniformly from one distribution to another, one system to another. The compatibility issues can vary quite a bit even between different releases of a single distribution, but when you multiply that by the differences between multiple releases of multiple distributions it can become quite a maze quickly.
This does not mean Linux is unsuitable for something like gaming, it means that a company considering bringing a game to Linux is going to have to be prepared to handle all of these differences themselves and have to either provide support for a wide range of distributions and setups, or limit their official support to a subset of all of the available distributions and provide limited or no "official" support for other distributions. It's really up to the developer in the end and some will decide to support it while others will decide not to.
For a lot if not most of games at least, it is the game developer who decides whether to support Linux or not on the code side, and then it is up to the distributor like GOG/Valve etc. to decide if they've got the resources to support it at that level as well.
Redfern: And to tell truth, as linux user i can say that MOST source of "halfbaked" ports is BAD ports. Peoples starting to developer games with pure windows-only frameworks and technologies like XNA, NET, DirectX, then after they finish they starting to "port" it to linux. In 50% of cases they try to use different dirty hacks like wrapping existing code in some sort of compatibility layers like Monogame or Wine and call it "port". Great if its more or less polished and working fine, but very often its just fails miserably. So, i dont really see why wrapping awkward code in awkward hacks making LINUX bad platform.
Yes, sadly that is the case that some developers choose to cut corners in this manner seeing it as a quick and dirty cheap way to bring support to the platform with minimal effort. In practice it does not seem to work out well and never has as far back as I can remember so I'd have to agree with you wholeheartedly there. ;o/ Wine has made great improvements over the years and is a great piece of software, but it is also sorely lacking in the audio/video department and other areas, and is a constantly moving target also.
Redfern: Rignt now linux support for gaming stuff is better then it was ever. We even got Unity building native (with mono part) builds for linux which working (unlike MANY ports) out of the box. But if developers will continue to use strange frameworks to make games and after "kinda" port them with no QA - nothing will changes.
Now here I have to agree with you 100% as well. Things are indeed getting slowly better over time and this is something us Linux heads should be very happy about indeed. The battle isn't over yet though, there is a long way to go yet. :) I think it will get there though in the coming years, and that Valve's Linux momentum will be a huge booster shot to see that happen. I also think we'll see other big game companies come forward to support Linux as well, and them combined with the multitude of game developers themselves all coming slowly one at a time to Linux are all going to end up making contributions to X, mesa, the kernel, Wayland, freetype, and all sorts of other areas in the OS stack. I think things are at the point currently where we can say that the scales have been tipped for Linux viability for gaming, and now those with the money and motivation need to (and to some degree are) coming forward to try and solve the remaining major issues that are perceived to have held things back in the past. 1 to 2 years from now I think we'll be having a very different conversation about this, and one which I'm personally quite excited about for sure. I've got a lot of man hours invested in hacking on parts of the system in the past and while I am not active on that front for a few years now it is nice to see things proceeding forward towards more commonplace usage and acceptance of Linux to the wider consumer marketplace.