In passato si basava lo sviluppo software sul credo che, operando in modo disciplinato, si potessero generare requisiti perfetti all’inizio del progetto…

“The old days of software development were grounded in a belief that discipline could generate perfect up-front requirements — a belief that perhaps emerged from the hope that such reasoning could eventually be fully automated, but that good human effort was an adequate substitute in the mean time. Methodology — another one of the truths that was a product of that culture — was held to be enough to achieve success. If we ever failed to deliver, we blamed ourselves […]. No one dared question the methodology itself.”

James Coplien, in Methodical Doubt, sostiene che ciò che assumiamo come verità deve essere dimostrato sul campo, e potrebbe cambiare nel corso del tempo. Anche nel mondo del software.