This book, written in 1975, offers a fascinating insight into the software engineering process used at that time. The author draws from his experience as project manager for the IBM System/360 and OS/360, and in doing so also sheds light on how they were put together. The 1995 edition, reviewed here, is particularly good as it presents the same 1975 text with only typographical alterations, followed by a couple of extra essays and reflection on the previous essays after 20 years.
The basic premise of the book is that Men and Months are not interchangeable on a software project, and that their design and management can be dramatically improved, but will still remain the most complex part of the process. The book demonstrates that this premise is true and then offers some advice on how to successfully improve the management of software projects. The advice provided is really helpful and well explained. Despite technology's rapid advance, the information in this book is still relevant.
The computer world had changed massively since that time. Throughout the essays you can see glimpses into that world as computers were moving from off-line preparation and batch processing towards a more interactive experience. So many things have changed, and it is fascinating to look back to when programmers had to make space and cost considerations which seem so different from today. As an example, the book mentions renting memory on an IBM Model 165 at $12 per kilobyte per month! At those prices you were sure to keep you code tight!
This work is well worth reading, whether for its retro value or for the insights that it can still offer today.