Frameworks/drivers/standards for software development for stereo/3D monitors

Question

3D

3D monitors and stereo technologies are slowly becoming popular, and the content is mostly, at best, photos and videos in the form of a stereo pair, and holography loomed on the horizon. Of course, you can't get by with static and recorded video content alone, you need support in games and graphical interfaces.

The question is, what are the standards for developers at the moment, Frameworks and libraries, or drivers?

So far, what I see is that everything is very scattered and cloudy:
1. Equipment
Support is provided at the driver level, as I understand it, only NVIDIA (only anaglyph or nvidia glasses), yes, there is some kind of SDK, but somehow it doesn’t reach the standard, if only because it’s not supported by anyone else (and I don’t really want to to support such a monopoly impudent trick).
1.1. as I understand it, only due to support at the driver level, occasional support for 3D games is possible (instead of 1 camera, we form two side by side), of course, not without bummers (the developers did not think about 3D and, for example, inscriptions may not have three-dimensional coordinates from the point of view of a 3D scene, but means fixed coordinates for the user).
2. iZ3D
As I understand it, also a driver, but only for anaglyph and, it seems, for polarized displays. But there is no binding only to NVIDIA.
3. at the program level
Various players independently try to support all available technologies (and since there is no standard, nothing more than anaglyph and frame mixing (frames side by side, interlacing, ... - this is how some stereo equipment receives an image) is not possible.

ps feedback (technology can be tied to this - for example, the user's location), none of this is observed, i.e. everyone implements and invents a steam locomotive.

Is this basically the whole sad picture at the moment or am I missing something?
Is it possible that by analogy with 'GPGPU Mess' -> 'NVIDIA CUDA' -> OpenCL, there will be development of 'NVIDIA 3D Vision'+iZ3D+'mess' -> ??? (I mean that as a reasonable answer to something exclusive and hardware-dependent, because it is logical that there should be a universal abstraction between the hardware and the program, otherwise it's a bummer)

Answer 1

I'm tinkering with the Qt3D lib there, they made 3d programmatically and added support for the iron separation of the cameras. Well, 3D manipulators. But in fact, it is suitable for small projects.

Answer 2

Yes, it's a mess. Which has existed and developed for 15 years already.
To verify this, just run a demo of some "Avatar" (game) and see the list of available 3D output methods: RealD, Sensio, side-by-side, line-interlaced, full checkerboard, NVIDIA® 3D Vision™, iZ3D and dual head formats. So far, no one is going to bring everything to one standard, because displays with a frequency of 120 Hz are still a curiosity.