Future physics thoughts


I realize physics beyond the simple voxel physics in place now may be some time away and rather undefined, but, I have some general questions.

Will (Should?) physics calculations be handed off to assignment-clients that can do physics calculations, and, if so, utilize more power by using many “physics engines” (at a cost or via having own resources)? Or would it be more as seen in OpenSim where the region server is tasked with handling this, limiting physics complexity to what a single machine can deal with? Or something else entirely?

I really hope the option will exist to leverage more power and have a better experience with physics responsiveness and detail - to whatever extent one is willing to buy assignment time and/or provide own hardware for. Perhaps as this evolves we’ll be able to get an idea of the computing cost of our latest creation and either simplify it or accept its “cost” versus having various hard capped limitations.