They say it's hard to compute an isogeny between any two elliptic curves, and yet they spend their time computing them. Isogeny people have played us for absolute fools! What does "compute" even mean for an isogeny, anyway? If you think you know, think twice. I, for one, change definition every other day. The truth is that as we kept discovering more and more algorithms / protocols / attacks, our understanding of what it means to compute an isogeny has changed, and some protocols we believed were secure were lost in the process, while others were created. In this talk I will explain how our understanding of isogeny computations has changed over time and what it means for cryptography.