Subdivision and Nesting

Download Audio

Fledgeling is super huge. Everything from the firmament to the flagella is in there. So how do you handle all of that stuff?

Well, I don’t know about you, but I tend to have a bunch of small ideas clumped together to make bigger ones. The “bigger” ideas aren’t actually more complicated or difficult to handle, they just happen to have a lot of other ideas “inside” them. Ideas are made out of other ideas. I call this nesting; things inside other things.

And (since it makes it simple) things (in my mind) are mostly made of smaller things like themselves. Mountains (big rocks) are made of a bunch of boulders (smaller rocks). If the boulders are sandstone then they are also made of a bunch of grains of sand (even smaller rocks). Rocks are rocks all the way down. This is called subdivision; nesting of like things.

Nesting and subdivision are crucial to our Fledgeling implementation plan. Instead of telling the computer everything that exists, we’re just going to tell it how to build things, and let it go from there.

Unsurprisingly, this is not an original idea. Self-similarity and nested complexity are defining features of fractals, which have been around for a while. Even before that there was Genesis, with “each bearing seed after their kind.” And, if we’re right, the world was designed this way even before it was written down.

Of course, all this is small bones without meaningful interactions. Pretty fractals are well and good, but this is a game! With all these levels of recursion, things could get really complicated really fast. Fortunately, meaningful interactions take place mostly at similar levels of nesting. People mostly interact with people sized things. Even so, we’re going to need some serious simplification and generalization to deal with the implied complexity of all these nested structures. Not honestly sure how we’re going to handle that, but we’re working on it.

Games so far rarely employ these concepts. Automatic subdivision and nesting is one of the few ways to model realistic complexity without things blowing up in your face. Unfortunately, when elements are hand-crafted it’s just too labor intensive. But now that procedural techniques are coming back into favor, we should be seeing fractal hierarchies more and more. It’s about time.

Leave a Reply

Your email address will not be published. Required fields are marked *