HMObject design question

[jdowling]

So, I've been doing a good bit of reading about game engine designs and specifically one area was the design of the "game object".

Basically I was wondering why the base object was deisgned how it is in this instance rather than just having one object class that contained a number of component objects.

So instead of having a new object type that is a child of HMObject and implements IRenderable why not just have one object (Just HMObject) that has a Renderer as a component.  Logic that depends on if the object is renderable or not could instead check if the Components hash of the object in question contained a Renderer or (depending on implementation) if thisParticularObject.Renderer == null or what not?

EDIT: or alternatively why not make the Model object a new object that contains a "Renderer" object and a "Material" object as well as a "Position" object (position object containing rotation, position, scaling) that kind of thing.

[Tiago]

as far as i know its just a design desition, i woulnt do it the way you would tho but our code usually reflects the way we abstract a design in our minds so it woulnt be based on the fact that its better to do it one way or the other but rather on how you designed the inner working of the engine in your head.

there are, of course, some ways of doing things that are better than others based on some criteria. your code might be (a little bit) faster, or more readable, or more component based, but since this particular engine is just for learning purposes it doesnt really matter that much. so, in the end, its just a design desition based on the hazy HazyMind's inner workings.

hope that helps