In case of G3D, we don't get access to Vertex Array alone, it can be done by copying the values and for different objects, this may make some sense. In fact, I guess the rendering happens by copied values into a G3D::Surface object. But this limits the interaction with Deformable/Destructive bodies as they might require access to the raw data because of geometry manipulation.
Exploratory tasks:
Figure out if the Models which the entities use are unique or not? If so, then we may have to use custom models or different workflow for Deformable bodies.
If they are not unique, then how do we communicate all the information back to them, we can have a custom memory mapped array which can be used to do this behavior as copying the values would result in performance decrease.
In case of G3D, we don't get access to Vertex Array alone, it can be done by copying the values and for different objects, this may make some sense. In fact, I guess the rendering happens by copied values into a G3D::Surface object. But this limits the interaction with Deformable/Destructive bodies as they might require access to the raw data because of geometry manipulation.
Exploratory tasks: