I'm still stuck on square one.
I'm not sure how to proceed with structuring this.
My current thought is to create a grid that is an entity/entity-archetype, which holds say 64^3 render mesh components, or blobassets (alive or dead). It also holds a regular matrix to track the state of the grid. Then I'll just change the approriate render mesh based on dead/alive.
But, if I want to have more grids, they must communicate with each other, and exchange information (fill in if there's alive neighbor cell)... How do you do that?
I want to eventually expand this one into fluid/pressure simulation, so I want to be able to do the neighbor check by matrix convolution.