Thanks for this! So the one found in OCaml and most languages is T-division (for truncating) and the one I called more useful is E-division (for Euclidean) which the paper argues for. It says that T-division is found in Ada, that Lisp has two modulo operators, one that does T-division and one that does F-division (halfway between T and E, and works for the circular buffer case), and that Algol and Pascal break the relation between div and mod by doing T-division for div and E-division for mod (if I got it right). Interesting stuff.