The use of constraints is one of
the most hotly debated subjects among schedulers, although the prepoderance of opinions are in opposition
to their use. As has been noted, the use of constraints can convolute a schedule and make it difficult,
if not impossible, to trace the critical path(s). On the other hand, Finish-No-Later-Than date constraints
are the mechanism by which we establish Late Finish dates for majot blocks of work, and are thus the
springboard for Backward Pass calculations that give us our much-needed Late Dates which, in turn, give
us the very Total Float values we use to monitor, evaluate, and priorotize project activity. What I
have always found (surprisingly) missing from such discussions is the idea of including as many FNLT
constraints as are warranted and then TURNING THEM ALL OFF, AND THEN ON, ONE CONSTRAINT AT A TIME. By
turning them all OFF, one can easily trace the critical path to the projects
completion. Then by turning on ONE constraint at a time, one can trace the critical path to
EACH individual intermediate constraint. This approach yields the best of both worlds. One can ascertain,
with relative accuracy, which activities DRIVE which constraints, and with how much influence. On the
other hand, with all but the last one (final activity) turned off, the schedules
Total Float values (and inferred Critical Path) are not obliterated,