Interesting phenomena from simple rule

[譲葉瑠珈]

Do not try to implement full-fledged simulator, because:

1. it's too much work
2. it requires infinite resource
3. we don't have enough ground truth to compare with

Instead, try to be next L-system and boid; that is, minimize rule and make interesting thing happen that doesn't depend on parameters.

Temporal L-system is already interesting, just add bunch of events (death, replication) and make evolution happen w/o considering physics or biology.

[譲葉瑠珈]

We focus on light. Just use Z-, directional light w/o global illumination. Don't make days or seasons; we don't have data for simulation.

All interesting coupling here is light - growth. Just dispense seeds when plant get's certain size, and die. Limit plant growth by moving energy to seeds.

Create a "gene".

[譲葉瑠珈]

This issue is too conceptual. Only remaining task is #16.