Very cool

If you gave a slight speed advantage to one side, they would "win" eventually.

maybe. should be easy enough to check.

Mesmerizing! Is there any randomness introduced in it, or is it totally deterministic? I watched it run once for a long time, and then for a shorter time, and couldn't see any obvious changes in the second run.

it is currently fully deterministic. it should be easy to introduce randomness in the starting directions of each "ball".

After some time both of the circles were on the same side:

yup, happened to me too.. thats a bug. will attempt to fix tonight.

(or we can see it as a metaphor maybe)

It's great. A speed-up option would be nice to have access to some long term behavior.

the suspense is like 80% of the fun!!

I want to know if an opposition between "center" and "encircle" strategies will always emerge, or whether a flat half-and-half division is equally probable

My intuition is that a flat line is unstable since negative curvature tends to cause less bumps (it's hard for the ball to squeeze in). Somewhat similar to DLA in spirit but different in the details en.m.wikipedia.org/wiki/Diffusi...

francisduvivier.github.io/eternal-stru...

Thanks! Interestingly my first hyper-fast forward led to a very early situation where both balls were on the same side ... strange! Quickly the whole area was black. But now I can't recreate it.

not early - but i did get both balls on the same side after a day+, a bug on my part which should be fixed at some point...

Amazing. If I play it for an hour, does the shape remain similar?

wdym by "similar"? the outer circle cannot change. but it is quite diverse within that constraint.

I dunno, it remains a mushy yin yang-y shape.

it kinda looked like a pokeball at some point.

Got it now!

I'd one can prove something about the steady state. Very interesting.

😂

This is amazing

Now I want to know if a) there's a phase transition if the magnitude of the dent increases and b) what happens if the dent magnitude is a random (say) exponentially distributed variable

interesting. i'd say "take the code and try it out", but i suspect it is too brittle and will just trash the hacky collision mechanism and get you some very weird results... (so, "yes", i guess, but for the wrong reasons)