Is it true that every large task can (usefully) be divided into quasi-modular subtasks which can be completed in a relatively short time? This notion is certainly useful for tracking progress (and spotting a lack thereof so corrective action may be taken)—it's the "sprint" model (of project organization) currently in favor in industry. Might we, however, be making an unexamined choice of trade-offs by exclusively using this model? Might the (previous? older?) model of "mileposts"—regarding a task as a marathon rather than as concatenated sprints—be more productive and useful in some circumstances?