keep in mind that i have never seen or worked with a 3d rendering system like this, i'm making it all up as i go along.
i was thinking one day and had an idea for making 3D. this might not be in a usable, or even recognizable, state at the time of reading, but i want to get this working at least a bit.
the idea is that you create a 3D cartesian coodinate space, create an "object/box" class that has accessible bounds and can be registered, then step set intervals in a direction for "raycasts" and inform each pixle of the screen by checking collision with the object registry.
GitHub