My background is as a developer. I see things from an engineering point of view. These days I am a consultant specializing in processes for software quality. I come into contact with managers whose concerns have to do with financing, staffing, reporting, and so on. Sometimes this leads to clashes of perspectives; most often it's an opportunity for learning.

In every company I've worked in, there exists at least one "expert user" who often yields more potential harm than benefit when it comes to systems and process improvement. This may sound strange; wouldn't you think the expert user would be your biggest ally -- your strongest resource for systems and process improvement? Not always, and here's why.

Projects begin with high hopes and dreams, which gradually fade away as the quality of the software decays. This process cannot be inevitable. Where, how, when, and by whom should testing happen so that we can feel better about a system after a year -- better yet, after a decade -- rather than after a mere month?

One of the great problems of science is forgetting what we've learned. Most students of the history of science know that after the fall of Rome and the rise of Islam, much of ancient science was lost to Europe until the 13th century, when ancient Greek works reentered the scholarly world. In fact, much of what we know today as the Renaissance was, in fact, a rediscovery of earlier knowledge.