Closed FailSpy closed 6 months ago
Oh, worth noting: I also added a check for CSGCombiner3D, as that would cause a crash similarly to ArrayMeshes
It also will try to grab a parent MeshInstance3D to the Collider (RigidBody or StaticBody) before trying the children, as that generally tends to be how I layout my Godot nodes, and I know other people do similarly.
I finally had time to at least read the code, looks all good to me. I would like to test it with Qodot when I get a chance as both Bezoika and myself us it. @Bezoika have you had a chance to see if this works with Qodot?
I finally had time to at least read the code, looks all good to me. I would like to test it with Qodot when I get a chance as both Bezoika and myself us it. @Bezoika have you had a chance to see if this works with Qodot?
Thanks for reviewing it. If issues do arise/are discovered with this system in particular I'm happy to jump in on those.
Pinging @ac-arcana here for a look, as well as @Bezoika
This is my attempt at implementing handling multiple surfaces. I don't currently have a way implemented to handle multiple MeshInstance3D children, so just go for the first one.
However, with a MeshInstance3D that has an ArrayMesh or multiple surfaces, it should work. I've tested this with as many cases I can think of, but I'm doubtless missing some.
This will try very hard to validate that there's more than one surface material needing to be thought about, and if so:
This is far from perfect in implementation, but it's surprisingly performant on my machine, curious to know how it works for others. I did a similar stress test to that video Bezoika posted of a heavily subdivided plane with two surfaces, and got an occasional stutter but really didn't feel much other than that.
There's a whole bunch of ways I could imagine optimizing it, but again, want to know how it works for others before spending the time optimizing.
This would fix #65