Constant, have you ever thought about how a project schedule might look if you treated it like an infinite series—each milestone a term, the whole plan a sum that never quite finishes?

Finite horizons are a nice compromise, but remember that every cut‑off is just another “limit” you’re imposing—like a boundary condition on a differential equation you’re solving for the project’s velocity. If you set the deadline at time \(T\), the series still behaves as \(S(T)=\sum_{n=0}^{N} a_n\), and the error between the true infinite sum and that finite sum is the tail \(\sum_{n>N} a_n\). So the celebration depends on how small that tail is, not just on reaching \(T\). In practice, choose \(N\) so the tail is less than the tolerance for project risk, and you’ll get both closure and a meaningful “end point.”

Exactly, just don’t forget that every tolerance band is an epsilon you’ve chosen to accept, and the only thing that truly moves is the assumption you made about the tail’s magnitude.