Should the t-1 position also be updated so that the velocity after collision goes into the opposite direction (with some dampening) like was done for the walls? in this setup, the balls would still move toward each other after collision.
I am not too savvy with this topic, hopefully Adrian will know.
