Note re possible measurement error

internal-latency.rkt

@@ -30,6 +30,9 @@
 		(transition s (send-ping src dst))]
 	       [(message (ping (== dst) (== src) start-time) _ _)
 		(define stop-time (current-inexact-milliseconds))
+		;; TODO: is there a way of reducing the measurement error here,
+		;; perhaps by recording against run-start-time instead of start-time, somehow?
+		;; TODO: first, characterize the measurement error
 		(define delta (- stop-time start-time))
 		(set! total-latency (+ total-latency delta))
 		(set! total-roundtrips (+ total-roundtrips 1))