meow :3

figuring out this metric took a while, I figured out pretty quickly it had to do with rate.

eventually ended up putting different metric scores for different rates in desmos and fitter an quadratic function through them, and saw that t was zero in 2.285... which is about 16/7.

I decided to just make a rate 2 machine that occasionally pauses, because special casing every fourth input just seems more expensive.

I also tried a period 6 solution with duplication, but the 20G cost of duplication makes this seem unviable.

because for the rate 2 machine, there needs to be something  disposed every cycle, there need to be 4 arms that can reach disposal.

Having 2 tri-arms that both debond and output makes me think that it's probably not possible to get any cheaper than this, but then again, some people are just insane.

meow meow meow meow :3