Comment by nilirl

> Truly robust engineering isn’t about what works for most; it’s about gracefully handling the edge cases.

How do you justify this when you factor in cost and time?