How I Fixed 1,000 Bugs In Six Months Without Stopping Delivery
And how deceived management to secretly refactor the code, so I could achieve the objective.
This story takes place around ten years ago. The situation was the following: we had an application in which, release after release, the number of defects were growing. For any new functionality, three or four more bugs appeared. The experience was sloppy, and nothing seemed to reverse the progression. At that rate, eventually the application would become a failure and thrown away. It seemed completely out of control, but one day, I realized that every problem raised from one big mistake. That helped me to make a methodology change that reverted the tendency. Without stopping the addition of new features.
How the problem appeared
The application was from one of my customers, and it was being built by another contractor. It was a good company, with good references, a solid contract, and with good tools and good technology.
When I joined the project, the application was already started. I was involved in later stages when my customer saw that something was off. During the first months, the application was too small, and they did not notice any problem. But time passed, and while the development of…
