has anyone seen a better answer for day 20 (colliding particles) rather than it's just stabilized at this number? The first part can be solved because there is a particle with a minimum acceleration, but if not for this it would be quite a bit more difficult. Similarly, the end of collisions seems to happen after a few iterations but there is no guarantee of this.

@dpsutton a friend solved it using quadratic equations to detect collisions, but in elixir: https://gist.github.com/sasa1977/a35e87540669f32c28e3bd2b7698ca7e