Here's the idea:
>The game world is a grid-based cellular atomata type structure, with one "alive" generation 0 cell at the beginning
>at the beginning of each cycle, all living cells (with any change in their neigbors) will look at all 8 neighbors, and read whether they are alive and if so it will read their generation number, as well as fetching their own generation number
>at the end of the cycle cells will use this information to choose which of their dead neighbors to populate with cells of their gen number +1.
the goal is to design an algorithm to populate a goal image. my original idea was to populate a circle (as in one you create in mspaint, based on distance formula calculations).
what are the consequences in higher dimensions?
can a generalized form be written that takes the goal image as a string of data and creates the appropriate cellular instruction set?
if we added the ability for cells to suicide, how does this change the problem?