a5de131711
I haven't checked this yet, but it seems like it also reduces the complexity (runtime) from ≈O(n^2) to ≈O(k^2 + n) (n = number of jump curves) (k = number of unique destinations) The worst case, that is k = n, is still O(n^2) but generally it's much more (scalably) performant. Thanks Arkanic for suggesting this! https://discord.com/channels/391020510269669376/663854113418641429/1256161815503704104 Co-authored-by: Arkanic <50847107+Arkanic@users.noreply.github.com>