Your cover story, Critics Can't Find the Logic in Many of Today's CPM Schedules, paints a disheartening picture of the current state of CPM schedules (ENR 5/26 p. 30). It's only 46 years since Morgan Walker and I first worked out CPM for duPont and yet project people are still falling into some of the same scheduling traps warned against during CPM's childhood. The use of features like "leads and lags," "multiple calendars" and "assigned constraints" do provide some levels of schedule flexibility. In practice, their use too often leads to inconsistent schedules and misleading views of project condition.
In the early days of CPM, computing capability was at a premium. Rooting out inconsistencies in scheduling data had to be left completely to the planner. In practice, this meant deliberately limiting the use of the "flexibility" features. Today, the desktop computer I'm using to compose this letter has far more capability than the UNIVAC we used for our first CPM calculations. Thus, there is no reason why the computer cannot be programmed to tell me that my scheduling input is inconsistent and why.