The Mythical Man-Month

Probably most of our readers have either already heard of this classical book by Frederick Brooks or even read it. However, since the goal of this blog never has been to provide up-to-date information but rather to provide a comprehensive list of important and influential literature and multimedia resources about Software Engineering in general, I want to use the time to present briefly this master-piece of SE to our readers who haven’t heard about it yet.

Even though the initial version of this book dates back until 1975 and is therefore already 35 years old, it is still worth throwing a glance at it. The Mythical Man-Month has been republished once in 1995 [1] with additional four chapters, among others, his famous No Silver-Bullet essay and a critical review at his initial statements he made 20 years ago.

In the 2nd chapter that carries the same title as the book, Brooks argues that the man-month is a myth and that man-power cannot be directly interchanged for time as many managers in that time (and still today) assume. He thus states Brook’s Law, that adding manpower to a late software project makes it later.

In the following chapters Brooks further discusses the importance of the human factor in SW teams, the significance of good SW design and design reviews, the cost/effort of SW and the importance of good tools.

Astonishingly, many of his previsions, opinions and advices can be interpreted as some of the common (agile) practices and methodologies which represent nowadays quasi de-facto standards in the SE community. To give some examples:

  • His advocacy for incremental development (in the No Silver Bullet chapter) fits very well into modern agile approaches such as e.g. SCRUM.
  • Brooks also outlines (although not invented by himself and only introduced in the new chapters of the 1995 version) the use of continuous integration, which is nowadays a standard methodology of SE.
  • The advocation of a good design and early design reviews adresses the consideration of fault costs and concepts like fault slip-through [2] which is very important in organizations that create critical SW.
  • The chapter about project documentation reads like a manual on how to use a modern wiki for documentation in SW project management (which is very common nowadays).
  • Similar to this, his chapter about bug tracking previews already some of the features of modern bug trackers.

Although some parts of the book are already out-dated (such as his debugging advices) due to the advances in technology and SE experience, the book is still worth reading and amazingly up-to-date in the rest of its chapters. Furthermore, as we always can learn from history, the book suits for learning the reasons that brought up some of today’s widely accepted SE standards. Additionally, it contains a lot of references to other interesting, classical SE literature.

Link to the book: F. P. Brooks, The Mythical Man-Month : Essays on Software Engineering, 2nd ed. Addison-Wesley Pub. Co., August 1995. [Online]. Available:http://www.pearsonhighered.com/educator/product/Mythical-ManMonth-The-Essays-on-Software-Engineering-Anniversary-Edition/9780201835953.page

References: [1] http://www.cs.unc.edu/~brooks/ [2] L. O. Damm and L. Lundberg, “Using fault slippage measurement for monitoring software process quality during development,” in WoSQ ‘06: Proceedings of the 2006 international workshop on Software quality.    New York, NY, USA: ACM, 2006, pp. 15-20. [Online]. Available:http://dx.doi.org/10.1145/1137702.1137707

Written on October 17, 2010