Currently it is possible to export the views in equidistant projection and as an intermediate cube map. This is bad, because it makes it impossible to optimize the pipeline, by calculating vram intensive textures like position only at the very end. A better solution would be to have a final shader, which maps the cube-map pixel-correct to a cube-map, calculating the missing measures and keeping the main render-step thin.
Currently it is possible to export the views in equidistant projection and as an intermediate cube map. This is bad, because it makes it impossible to optimize the pipeline, by calculating vram intensive textures like position only at the very end. A better solution would be to have a final shader, which maps the cube-map pixel-correct to a cube-map, calculating the missing measures and keeping the main render-step thin.