xiii software is very troublesome because it is among the most labor intensive and error-prone technologies
of all history.
xiv Why some software projects succeed and others of the same size and nominal difficulty fail or run amok
is what this book is all about.
p.169 Dr. Barry Boehm has stated in Software Engineering Economics (Prentice Hall, 1981) that "a good
estimator should be able to come within ten percent of actual results 90 percent of the time." That remains a very good general
concept. However, an additional caveat might be added: "a good estimator should never deviate by more than 20 percent."
p.279 There is nothing more important to successful software than the availability of capable teams guided
by capable managers.
