In article <201301111401.r0BE1nCJ031838 at ls-al.eu>, Sander Reiche <reiche at ls-al.eu> writes: these books and games (see bellow) The books he cited in his list are all books describing how to write your own engine, not using someone else's, so I gather that he's thinking of doing things from scratch and not using anyone else's code. Basically, all the "old school" 3D software engines all boil down to ways of getting the inner rendering loop to run fast enough on your hardware to achieve real-time rates. This is why the monsters in DOOM are sprites and not 3D models and even the representation of the game level is strictly 2D. (On areas where there appears to be a 3D element with ramps, stairs and such, you cannot go underneath any of those raised areas.) The data structures used to hold the maps are custom for the game and basically use a boundary representation (i.e. "B-reps") so that you can cast rays from a pixel and quickly determine what they intersect and at what depth, in order to render the pixel appropriately. They also use limited precision (both spatially and in colorspace) textures in order to simplify rendering. In other words, it's a highly constrained rendering engine and its the constraints that simplify the problem to the extent that it could be rendered in software at interactive rates on a dumb frame buffer with no Z-buffer support (i.e. a VGA card). -- "The Direct3D Graphics Pipeline" free book <http://tinyurl.com/d3d-pipeline> The Computer Graphics Museum <http://computergraphicsmuseum.org> The Terminals Wiki <http://terminals.classiccmp.org> Legalize Adulthood! (my blog) <http://legalizeadulthood.wordpress.com>