Issue with using freezeActiveMeshes to improve game performance

I’m not using freezeActiveMeshes so I can’t comment on that. But my method for creating mesh previews is dirt simple. I have 2 types of meshes, clones and thin instances. And its fast enough for both.

pseudocode

a) get master mesh
b) clone it and its skeleton, give it a unique id (which is used for disposal)
c) create material for it
d) set its animation
e) reduce visibility. DONE!

Use the gpu picker, its faster.

I have both mesh previews and the gpu picker live at Seedborn. Its a testbed, feel free to break it. Oh btw, I love your visual style! I can see that you are moving into the nitty gritty stuff. Keep it up, because its where things start to get really grueling. What I did with Seedborn was to develop every piece of code in the PG first, keep a copy, this ensures a rock-solid fallback, then port the codes to my dev, test, test and more tests! Iterate for a few hundred cycles and your confidence and speed will go up. tldr: develop in PG then port over. If it doesn’t work in PG, it won’t work on your dev.

Hope it helps, cheers!

3 Likes