Hello!

I'm doing some sort of high level abstraction for my game.
As I understood the common convention is to have so called game entities, which are simply objects in the scene (either renderable or not like triggers).
Now assuming I create a renderable entity with geometry, texture, material and etc. To render it I have to create vertex buffer for it and then bound the vertex buffer to the device (I'm talking here about DirectX9, but since it a general game question and not dx related i ask it in this forum).
Now I can have very large number of entities, should every renderable entity be created with its own vertex buffer? Should I reuse the vertex buffers somehow?

Sorry for stupid (maybe?) question and thanks a lot for help!

Lots of your game entities are probably going to share the same models, so I would make the VB belong to the model, and then make GameEntities share models.

GameEntity points to a Model.Model points to a Vertex buffer.

Ya thats sounds a lot better! Thanks for this!