8 | 2002

In this month's CBR, we take on a classic issue: software estimation. It's a classic because it looks, on the surface, like something we ought to have figured out by now. There's a "way it's supposed to work" that looks plausible. That way -- the "objective estimation" version -- goes something like this: