Hacker News new | past | comments | ask | show | jobs | submit login

I can't be the only person who has wanted to try bringing the wolfram physics project style hypergraph generation rules into some 3D game engine like unity or unreal but was too intimidated to start.

I've done some cool stuff with generating fractals like these ones made by Softology

https://www.flickr.com/photos/39445835@N05/albums/7215769136...

https://www.flickr.com/photos/39445835@N05/albums/7215768121... using visons of chaos to create .obj files and then importing them as voxels in unreal engine using voxel plugin, but hit some limits quite quickly in terms of data size / memory usage and how large of a fractal structure I could visualize (example: megastructure in this demo of a vr game I was working on https://photos.app.goo.gl/V7NrtG4bCbU3EEwy8)

but would love to try to create some way to visualize and interact with large hypergraph structures in 3d space like the wolfram ones: https://writings.stephenwolfram.com/data/uploads/2020/04/041... https://writings.stephenwolfram.com/data/uploads/2020/04/041...

Wonder if this library could be a good place to start to look into this. you obviously need some language / format to store info about hypergraphs and build them up (what this seems like it could help with?) and then some way to render them or create procedural meshes (which in and of itself will be difficult, although I've seen some people have success with stuff like fractal procedural meshes (https://forums.unrealengine.com/t/wip-procedural-fractal-mes...).

The standard noise generation algorithms (perlin, simplex, iq noise, value, voronoi, etc) and rendering algorithms like marching cubes capture all the procedural 'terrain' / structure attention/hype in games and media, but I think cinematic sets or game environments built of hypergraphs or fractals are an unmined/overloocked resource.

if I could get a team of 2-5 and enough to pay them for a few years, I feel like you could combine hypergraphs and fractals 'terrain' with ue5 nanite to make the memory requirements of the huge / high fidelity meshes lower and make some really cool stuff.




What triggers the initial willingness to implement something with hypergraphs was indeed the Wolfram physics project! https://github.com/yamafaktory/hypergraph/discussions/11#dis...




Consider applying for YC's Summer 2025 batch! Applications are open till May 13

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: