I've edited my comment to add more context to that quote. He absolutely advocated for the most minimal of function lengths, beyond what is reasonable.