Voxel Model

[DigitalLibrarian]

For things like organisms and other components that need to be composed of one or more chunks, with a set of bone transforms, we will need a custom model.

There will probably be several shaders for this, but they will all follow a familiar bone transform flow, like a model of meshes. We want to have lots of voxel models, so the bone transforms won't be full matrices.

There will be tightly packed (byte/bit level) parameters for a generic transform system, that supports dynamic transform as well. For example, the first half byte might select from a set of sizes for a chunk of uniformly sized voxels. The second bit might tell us if it is just an empty cell.

The idea is to get each chunks byte size as close to zero as possible, so we can pump more into the GPU.

Another optimization is would be to allow for a way for the user to provide occlusion clues. Organs and bones don't need to be considered for lots of things, like rendering.

[DigitalLibrarian]

This will open up development for rendering relationships like biting and mating etc.

Adding a new body part will consist of adding a set of connected chunks to the organism and treating that set of chunks as a single body part for insertion purposes. (sounds like a BodyPart derivative class)

[DigitalLibrarian]

From #21

Might experiment with creating a file format with precalculated transforms as poses. The model can then be put into these poses.

Might be possible to interpolate between these poses for mixable animation verbs. If so including feedback from the physics system would be really cool.