Ah, so your point (ba-dum-ts) is that equality should just be type-invariant. That's a reasonable take, but it's a bit orthogonal to the larger conversation. I

Ah, so your point (ba-dum-ts) is that equality should just be type-invariant. That's a reasonable take, but it's a bit orthogonal to the larger conversation. In your invariant world, is the integer 2 equal to the floating-point number 2.0. Or is 2.0 actually a composition of an integer part and a fractional part, while 2 and 2.0 are not directly comparable?