The Mature Optimization Handbook

I started reading The Mature Optimization Handbook last night. It’s pretty great. For example:

There is a rich & vibrant oral tradition about how to write fast
programs, and almost all of it is horseshit…

All of this define-the-problem-then-measure stuff isn’t like a set of training wheels you cast off once you’ve learned how to ride on your own. You really do have to approach every act
of optimization as an experiment, starting over from scratch,
making every assumption explicit and testable.