The 64th Gamer

:MinecraftGlass: Pneumagination Thoughts 2

Hello, time for more Pneumagination talk.

I am currently completely overhauling the entire game to Godot. I think I’m just fed up with the amount of technical debt Unity has that truly prevents me from working on this game with the features I expect.

My main things are pathtracing and multithreading. The engine was not built for either of these things originally and their current bandaid solutions do not feel satisfying. I’d been using TrueTrace which was great for BIRP, but it was not built for my usecase of customizable shaders.

I understand Godot also possibly will not exactly have what I need right out of the gate, but since pathtracing is the current future of render pipelines I feel this community will pull it off better than Unity’s lackluster implementation of HDRP. If I have to I’ll also just make a custom pathtracer myself.

I knew my next project was gonna switch to Godot but since Pneu is getting this overhaul I felt it was best to go ahead and switch now. I’m always for going open source and I think the engine is good enough to be worth switching to.

Update: Okay began work on the Godot fork and put it in the Github Repo. I’ve currently got a solution for the chunk generation, brush generation, and rendering of the brushes. The meshes do form but they currently do not display due to what I think is a missing material. I already LOVE the fact you can just straight up save generated meshes to a file in a single line to preview them. That’s how I confirmed it was atleast working.

Current gripe is that there’s not any default scene view when playing the game for me to look around in. Atleast there’s a remote heirarchy available- and its only really a downside in these early versions of a project where I don’t have a camera looking around yet. A fine sacrifice so far.

With mesh generation though I think I got an idea on how it’ll work. The world will generate in 2x2x2m blocks, and then an outer layer of blocks will be generated with the (uhhhh the name escapes me right now) method where you check surrounding tiles and that determines what the current tile will be. I’ll check just if its air or ground, and then make tons of sets of prefabs of multiple brushes in a 2x2x2 area that smooth out and curve the landscape. This also lets that first layer be broken faster while the deeper layers are bigger and take longer to dig.