how could unlimited speed and ram simplify game code?
imagine its the future, and PCs have become so powerful that gamedevs have access to unlimited ram and infinite clock cycles to burn for all intents and purposes.
how could this simplify game code?
a few things come to mind:
just one generic game object type, and everything is a game object. every entity, every inventory item, every terrain chunk mesh, every particle and emitter, everything.
a flat list of all game objects.
render could simply iterate though the list of game objects: if active and renderable, render it. wouldn't even need culling (unlimited speed - remember?). no backface, no frustum, no occlusion, no possibly visible surface determination, no oct-trees, etc.
update could also just iterate thru the list of game objects: if active, update based on type.
input could ether process input immediately, or store it for later use during update. don't see many things changing there.
there would be no need for separate storage for shared constant data of entity types. each object instance could store a copy of all constant values (weight, price, based damage, etc) - unlimited ram, remember?
all object types could have all possible member variables for all object types (unlimited ram), so you only need one type of "uber" game object type for everything.
obviously, one could also go crazy with new features and exotic processor and ram intensive algos, but thats not the point here.
the point here is not what new complex stuff we could do (we could do anything), its what current complex stuff we'd no longer have to do.
so, how could unlimited ram and clock cycles simplify game coding?
NOTE: this started as a though experiment, but has evolved into "how could such an architecture be adapted for use today", IE what complex stuff we still need to do (like culling).