TinyEngine VS territory

The full code is available here. Written in C++ with TinyEngine.

You can find an article explaining how this works on my blog here and a terrain-based implementation here. For more detailed source code on the LBM method, check out the examples here.

This is a simulation of wind flow over procedural terrain using my own GPU accelerated implementation of the lattice boltzmann method. All implemented and visualized in TinyEngine (by me). Terrain generated using SimpleHydrology (also by me). The source for this is not released (yet!). This scene is 100% procedural.

[Source Code]

I went ahead and made you a reference implementation. You can find it here. It's pretty short, and by the shaders you should understand how it works. Instance render 3D quads using centroids as attribdivisors, then reposition the quad vertices correctly, manually write to the cubemap depth buffer and whatever you want to the colorbuffer.

Here the link to the github repo: https://github.com/weigert/TinyEngine I thought it seemed neat and think I may use it for some future projects.

I have already integrated this into my personal engine TinyEngine (simply by making all file loading interfaces use stdio.h FILE* style interfaces). Any program built with this system can use either a real, relative file system or be built to have a virtual snapshot embedded into the executable!

The visualization of the dynamic data was made using my own TinyEngine.

So I wrote an article on it here. You can find the code here.

A practical implementation in a real voxel project is actually here. I never actually hit any limits before.

This is already a C++ library / system! You should be able to run it on your linux repo of choice. I am also trying to make a "generalized" terrain generator. Once it's done, I'll probably voxelize and add dwarf fortress style agents like here.