I'm saying its a proof of concept because it is cool but there are some hefty overhead costs on performance, for starters because the units are actual units and not points of data (entities) every single object in a model is replicated over the network which is not cheap. I could rewrite everything to be entities but that would great overcomplicate the scope of the original frameworks intention and would introduce inconsistencies with how simulated units feel/move/behave vs how player owned and controlled units would behave and would also introduce some inconsistencies with hitboxes and require a complete and total rewrite of projectiles, networking and pretty much everything when I really just want to prototype things and quickly figure out what I actually want to make rather than time sink into another never to be finished or released project. Which this framework has allowed me to do in 4 days. 7 if you include the 3 days to setup the framework, which pretty much consists of me remaking different things from different projects into a re-usable modular component system such as: - custom tools - custom lua based character(s) - custom lua based humanoid - custom per character inventory (I don't use Player.Backpack or StarterPack anymore) - custom character controller (Edited version of PlayerModule to fit with my game) - custom character control logic for dynamic control and server to client handsaking. - light netcode rollback for tools and attributes via custom "NetworkVariable" components - raycast based suspension - modular projectile code - mech builder and IK walk cycle solver - custom proximityprompt and clickdetectors (you have to make your own if you're controlling a character that is not Player.Character!) - custom character "motor" - custom aircraft "motor" - custom vehicle "motor" - custom naval "motor"