[toonalfrink]

I find that function point estimation, potentially adjusted with COCOMO factors, is pretty decent. I find it strange that no one is talking about these methods