Agile methods

The so-called "agile" methods that appeared under this name in the 2000s, however, have a slightly longer history. Unfortunately, as is often the case in IT, terminology is changed rather than concepts. The only certain result of this deplorable practice is confusion in the minds of decision-makers, who have ended up losing all confidence in what their CIOs and/or IT experts tell them. Skepticism and mistrust reign.

At the end of the 1980s, the prolific author James Martin published his Information Engineering in three volumes with Prentice Hall, which a few years later gave rise to RAD (Rapid Application Development), quickly known as "Quick and Dirty" and completely discredited as a result of the bad practices it spawned. There were many good things in Information Engineering, but as we all remember, it was a time of technological disruption, with the dazzling development of distributed architectures, client-server and its various variants (fat client, semi-thick client, thin client, etc.). Older languages such as Cobol, even when revamped with the application generators and L4Gs that flourished towards the end of the 1980s, were no longer of interest to many people (even though there were still billions of lines of Cobol code in banks, insurance companies, government agencies, etc.). The fashion was now for object-oriented languages (C++, Java) and object-oriented design methods, which in the 1990s gave rise to the UML language. We're still in the euphoria of artificial intelligence, whose media hype has succeeded in making many decision-makers believe that we're going to be able to mechanize the production of programs (by getting rid of programmers, those annoying people who make mistakes and are incapable of expressing themselves in a language that everyone can understand) and finally experience the best of all computer worlds, where errors have magically disappeared. The awakening, which takes on the appearance of a crash, will be particularly brutal (cf. the Standish Group's Chaos report [Doc. H 3 200] ).

This was also the period of all the excesses resulting from a misunderstood quality approach, which quickly turned into a bureaucracy, whose objective was to meet often inept standards (the infamous DOD 2167 standard is still remembered by some), invented by "experts" who had never built a system in their lives. As we all know, the result will be failure, and a general discrediting of quality, the name of which is no...