The Error at the Foundation: When the flaw is the foundation, not a bug

I have built that kind of error with my own hands. In the welding workshop we once cut a fixture, the jig every bracket gets clamped to, off a reference that was wrong. Nobody caught it, because every part after that was cut to match the jig, so the parts fit each other beautifully. They just did not fit the drawing. Here is the trap: the better my guys got at making parts to that jig, the more expensive it became to admit the jig was the problem. Skill was piling up on top of the mistake. Competence was voting for the error.

That is what separates an error at the foundation from a surface bug. A surface bug is local, and you patch it where it sits. An error at the foundation is a premise everything later was built to honor, so you cannot pull it out without disturbing all of it. Ward Cunningham named the softer version in 1992, technical debt, where every shortcut charges interest on the work that comes after. Paul David showed in 1985 how the QWERTY keyboard, shaped by the jams of early typewriters, stayed locked in long after the machines that justified it were gone. And G. K. Chesterton left the warning back in 1929: do not tear down a fence until you can say why someone put it there.

So the real decision is rarely how to patch. It is whether the thing is a bug or a foundation, and if it is a foundation, whether you will pay to rebuild rather than pay interest forever. Joel Spolsky warned in 2000 that a from-scratch rewrite is the fastest way to kill a company, so the honest path usually runs between the two: replacing a wrong foundation piece by piece while the thing still runs, the way you take down a load-bearing wall without dropping the roof.

This is a short book about the difference between an error at the foundation and a surface bug, technical debt and the interest it charges, path dependence and lock-in, sunk premises and load-bearing mistakes, Chesterton's fence, and when the honest move is to rebuild rather than patch.

This book is written from an Indonesian operator's context, but a mistake that everything competent was built to honor shows up the same way in other emerging-market and small-business settings.