Extreme Reaction Counter Commentary

Extreme Reaction: Counter Commentary by HakanErdogmus, National Research Council

Reprinted from WOW: ICSE's Window on the World, Volume 8, Issue 2 - Thurday, 23 May, 2002

Jason Hallstrom's perceptions of Extreme Programming are plain misguided. XP and other agile processes my very well fade into oblivion if proven ineffective through practice. In the meantime, the jury is still out.

Hallstrom is ignoring compelling anecdotes coming from the trenches by latching onto mantras that he hopes will not resonate well with the ICSE audience. The research community in particular should be open to new ideas, as well as the reincarnation of old ideas whose time may have arrived. We should resist the temptation to dismiss them just because they threaten deeply rooted beliefs and appear to nullify our past efforts.

XP and other agile processes, besides their emphasis on the social context and team dynamics (not dissimilar to what have long been advocated by the DeMarco, Lister, Constantine, and their likes), encourage technical practices that merit at least further investigation. Test-first and refactoring come to mind. It is unfortunate that Hallstrom tries hard to give a negative spin to refactoring -- a practice that aims to improve the maintainability of code -- in the name of trashing a philosophy with which he disagrees.

XP may not be a silver bullet, but it is not an undisciplined approach for hackers, as Hallstrom claims. If you are really curious, do not take Hallstrom's, my, or KentBeck's word for it. Try it, and decide for yourself. Who knows, you might be surprised!