Book review by Ted Felix
Another classic. Portions are a bit dated, but the general theme is timeless. Software estimates are always optimistic, and Brooks explains why, and what can be done about it. I like his discussion of how conceptual integrity requires that there be only one system architect. However, I find his definition of the term architect a bit unusual. By architect, Brooks means one who defines the user interface and feature set, not the person who decides how the software is broken into modules. Brooks advocates small, meaningful documentation as opposed to an ISO 9001 paper (or digital) avalanche. His discussion of how flow-charts were useless from day one was intriguing to me. Apparently the only reason so many were made was because the government required them for certain projects. The "No Silver Bullet" essay is fascinating in that it places the blame for software costs on the inherent complexity of software itself. Overall, an interesting glimpse at the past, present, and future of software development. Short and easy to read, but worth the time if it sounds interesting to you.
I would recommend reading Software Project Secrets (Stepanek 2005) prior to reading this book. Stepanek's book is more modern and better written.
Read from 9/21/2005 to 10/20/2005.
<- Back to my software books page.
Copyright ©2005-2006, Ted Felix. Disclaimer