Watts S. Humphrey (1927–2010)

Perhaps it says more about me than about the calibration of this book, but while working through it on my own I got to program 2A (1/6th of the way through the book), said "Writing a *good* Lines-of-Code counter seems hard to do in any language I know", and dropped the book. Has a time-and-motion-like focus on quantifying the fine-grained details of the student's effectiveness at writing software, under the theory that we can then vary the processes used and *measure* whether the new show more processes are actually any better. show less

The book is, as someone put it, a precursor to the CMM model of the early 2000s. A fascinating book with core principles still valid. Just not sure how old technologies fit in. But barring that, this book will hold at least until some radically different method of developing software comes up.
Boring in some parts (it is a techie book), but one can just glaze over those pages and move on.