This is like everything else in life: there is the initial cost, there are maintenance and support costs, and there are replacement costs. With roads, it's fai

This is like everything else in life: there is the initial cost, there are maintenance and support costs, and there are replacement costs. With roads, it's fairly obvious when maintenance or replacement is required, so people responsible for financial expenditures cannot ignore them. But with software, the "invisible" nature of it seems to allow those responsible to ignore reality and kick the problem down the road until hopefully it is someone else's problem. (Incidentally, this is how a lot of modern public companies with with respect to 5-year CEOs and quarterly earnings and actual responsibility to shareholders, employees, and public). Anyone who has technical experience knows that it's a tough sell to convince upper management to spend money on problems that they can't even see. The big question is, when is the technical debt (or outright catastrophic failure of a decayed system) worth more loss in revenue than its replacement would cost?