Why is Trigger Rally's Cannon.js is Better?

We compared the cannon.js latest version and the trigger rally’s cannon.js, it appears that the trigger rally’s performs better and don’t stop working easily even when machine over heats (it will eventually crash when the heat is too much).

If only the latest cannon.js could be optimized as the trigger rally’s?

Latest cannon.js (v20)

Trigger rally’s cannon.js

same map but you can already tell the v20 is not as performing as the trigger rally’s.

Hi Ranao! I’m the producer of the neighboring project “Mars Digger 3D”. I saw your comparative test on Cannon.js open-world physics performance—very hardcore!

We are currently building a voxel game with Three.js. As our map scales up and physics calculations increase, we’ll definitely hit the performance bottlenecks you mentioned. I’d love to privately ask you about your experience in physics engine optimization.

However, due to network restrictions on my end, I can’t register for Discord or Telegram. Would you be open to adding me on WeChat or QQ? My WeChat ID is: MarsDigger_Xu. Just looking to make a tech friend and discuss open-world architecture!

u can email me at ranaotaru@gmail.com