It would be great if we could have a version of the GVR SDK that, just like Apple did with ARKit, allows us to create a Session object that is independent of any rendering APIs in the device, and then either poll or get notified of any new Frames generated by this session, and then get projection, head and eye transforms from this particular frame.
Looking at the code for GVRKit that was still present in the repo for 1.150, it looks like there is already a similar concept implemented in the native C/C++ SDK; the challenge would then be to find a way to separate the OpenGL-specific code from the session, and send it to some kind of renderer "bridge", while also giving end users a way to implement their own code for the bridge to allow them to integrate with different rendering APIs.
It would be great if we could have a version of the GVR SDK that, just like Apple did with ARKit, allows us to create a Session object that is independent of any rendering APIs in the device, and then either poll or get notified of any new Frames generated by this session, and then get projection, head and eye transforms from this particular frame.
Looking at the code for GVRKit that was still present in the repo for 1.150, it looks like there is already a similar concept implemented in the native C/C++ SDK; the challenge would then be to find a way to separate the OpenGL-specific code from the session, and send it to some kind of renderer "bridge", while also giving end users a way to implement their own code for the bridge to allow them to integrate with different rendering APIs.