Why Software Gets In Trouble

Quality Software #1.2

by Gerald M. Weinberg

Many books have described How Software Is Built. Indeed, that's the first title in my Quality Software Series. But why do we need an entire book to explain Why Software Gets In Trouble? Why not just say "people make mistakes"? Why not? Because there are reasons people make mistakes, and make them repeatedly, and fail to discover and correct them. That's what this book is about.

Genres: Nonfiction

Nook

First published October 2, 2010

About the author

Gerald Marvin Weinberg (October 27, 1933 – August 7, 2018) was an American computer scientist, author and teacher of the psychology and anthropology of computer software development.

Reviews

[Mikhail Filatov · Rating 4 out of 5]

Overall, quite an interesting book, taking an approach of systems dynamics to Software Engineering.
E.g. "Controller fallacy" -that if a controller - e.g., an additional code review- is not finding a lot of issues it's useless. The goal of such a controller is to control quality and it can be useful as a stage by itself, making people being better prepared, etc.
He also is right that the approach to bugs as "moral issue" is not productive, as well as the goal of "zero bugs". This issue should be considered from systems dynamics perspective: which actions would create negative feedback to bugs creation so in the long term the number of bugs is getting smaller (but never reaches 0).

Cons:
Graphs are scanned pictures of low quality, it's difficult to understand sometimes what is there.
Anecdotes are really stacked against "software managers" vs. being objective. E.g., one software manager told the author that he considers an engineer asking a lot of questions "the worst", while the author believes this engineer is "the best". In reality, without more context - what questions, how said engineer used the answers, etc. it's difficult to make any objective conclusions.

[Johnny · Rating 3 out of 5]

Gerald M. Weinberg explains in this book the basic knowledge you should have of how failures happens when you write software and what you can do to reduce their effect. The explanation on why it is impossible to reach a 0 fault level is really good and should be read by everyone who is involved in any capacity in the creation of software. Many discussions could be avoided with that knowledge.

However, many parts are not new and splitting the series in 9 books gives you some continuity errors. This book is part 2 and when you didn’t read part 1 you will not know what he means with type 1, 2 and 3 companies. Repeating that would make the reading much simpler, but so you have to get part one or ignore it whenever it is mentioned.

[Ilya · Rating 5 out of 5]

True coal mine, indeed.