This book was not what I expected. Humphrey uses mostly anecdotal examples to illustrate his many points. Though this helps get his points across, it does not really prove his assertions.
Most of his advice is not practical, or even possible in the employment situations I've seen (and heard about) over the last ten years or so. I found a few interesting parts, much like I would find it interesting to listen to the tales of any old-timer about the 'good old days', and some of his insights about people in general are quite keen.
Some parts really hurt my will to read on. For example, he seems to believe that if a manager can get his team members to work lots of overtime, that higher productivity will automatically follow. Someone who has written books about the use of careful measurements during software development should know better. The evidence I've seen and read (in other books) indicates that regular overtime is a 'bad smell' of deeper problems, and a perfect recipe for low quality and ultimately failed projects.
He even claims that the manager's job is to put schedule pressure on the engineers, otherwise they'll take forever and never get anything done. Again, he includes a little anecdotal example. However, with very few exceptions all of the engineers I've worked with hold themselves to certain standards of quality and productivity. Usually management pressure (especially the old time-crunch game) just hurts more than it helps.
Overall, much of his advice doesn't fit with the reality I've been experiencing lately.
I recommend comparing and contrasting Humphrey's advice with that found in "Peopleware" (2nd ed.) by DeMarco & Lister.
Also, for even better book full of 'management tips' see "201 Principles of Software Development" by Davis.
