On average things get more broken, not less, because a small company can only break so much before it simply stops working altogether, while a big one can absor

On average things get more broken, not less, because a small company can only break so much before it simply stops working altogether, while a big one can absorb enormous blows of failure and continue on inertia alone, while simultaneously taking many more minds and bodies along for the ride. Nearly everything in software is based upon a mad scramble to create value without collapsing. When software manages to not collapse for decades, we look upon it with 20/20 hindsight, thinking, "wow they did great engineering," but really, it's the same as thinking the Romans were great architects because of the things they built that by chance stood for a few thousand years.