Rendering in DXFramework is pretty simple, during the render phase you
call render on each of your sprites, the first ones you call render on
are drawn first and the later ones on top. There is no Z-buffering
mechanism. It's designed to be simple and geared for 2-D.
The back buffer can be set to clear each frame or it can be set to
stay the same. If it is cleared each frame, you must redraw everything
in the area that was cleared every frame. If not, you need to manually
control what to redraw based on changes.
There are many, many ways performance in DXF could be improved.
Performance was not one of the framework's design goals--time to
develop was emphasized instead.