5 comments

  • markstock 6 hours ago

    The tweaks that SW included are pretty common to similar simulations:

    1) the attraction-repulsion pair is typical of "boids" simulations, and also features in the Lennard-Jones forces acting in atoms (albeit with different powers on the decay rate);

    2) making the force drop to zero for particle-particle distances approaching zero is effectively treating each particle as a "cloud" and not a singular point, it's called "regularization" in vortex particle methods and "Plummer softening" in gravitational methods; it's primary benefit is with SIMD and parallelization by removing the need for a (i != j) conditional; and

    3) slow contraction (multiplying each particles' position by (1-epsilon)) is a technique used by generative digital artists to ensure that visual activity does not stray too far from a directed point.

    The unique component of this simulation is the (computationally-efficient) dependence of a very small number (N=2 here) of neighbor particles. The typical O(N^2) that limits real-time simulations to 10k-50k particles becomes O(N).

    • yehoshuapw 4 hours ago

      I find it fascinating how (in the new version) the further away clumps stay separate, until something small from the larger swarm reaches them, then they all merge back to it

      • Honali 5 hours ago

        What I like here is that the interesting behavior survives after throwing away one of the main assumptions behind boids: locality

        • palad1n 4 hours ago

          Brings me back to when there were boids applets.

          • Avicebron 10 hours ago

            (2012)