Robert L. Glass is an old-timer in the field of software engineering; this book is founded on decades of experience. Moreover, his discussion of any given topic is based not only on his own experiences but also on any relevant studies that have been done in the area (if any). As a software engineer myself, I found his observations a refreshing change from the usual sort of thing one hears: "If you'll just look at things my way, and follow my process, then all of your software engineering problems will go away!" I found much of what he had to say to be useful and timely.
The book isn't perfect. In a number of places he makes observations and then doesn't follow up on their obvious corollaries; he has a touching faith in ten and fifteen-year-old studies that have never been replicated; and his attitude toward the Unix programming community is almost patronizing at times, which is annoying.
On the whole, though, the book serves as a useful reality check, especially for those who want to elevate the process over the people involved.Posted by Will Duquette at January 20, 2004 08:15 PM