Original Post
I was thinking the other day about "COMPONENT-IZING" my game. So my idea was to provide a rendering engine, a physics engine and a "definition" engine. Renderring and physics are well understood. A "definition engine" would allow plug-in DLLS to populate object classes: like giving a definition for a new tank using turrets and control systems already defined by the game. It could also be a human-layer-device port for controlling specific items in the game. I believe that the main advantage here is allowing for MASSIVELY DISTRIBUTED DESIGN for the game. So a gamer could download some DLL from the internet, install it, then go compete with some other guy. Is there any dangers with such technical design? Can you think of other benefits? Would you be interested in developing something for such an engine (not that I'm asking you to)?