Programmer Musings

Review of Joel on Software

Joel on Software
Joel Spolsky
Apress, 2004

This book collects some of the best entries from the Joel on Software website. Joel has an interesting view of the software industry based on three years at Microsoft, time spent at Juno, and now with his own company. This is a guy with quite a bit of experience and some interesting insights.

The essays in this book cover a large portion of the fuzzy part of developing software. Joel doesn't spend much time on algorithms or methodologies. Instead, he focuses on the real process of developing software. If you don't believe programs need a functional specification, you need to read what Joel has to say. If you are not sure how to estimate projects, read on. If you would like to know how to hire great programmers (or would like to know why you should), Joel has some answers.

In reading these essays, I often found myself surprised by Joel's opinions. His explanations are clear, and his ideas are interesting. Although I do not always agree with his conclusions or interpretations, many times I feel he is right on target. Some authors make you feel that they are the authority on a subject. Joel sounds like a guy who has been in the trenches and has his own views of how things work. Even when I don't agree with him on a particular subject, the essay gives me another viewpoint to compare to my opinions.

Sometimes he seems to be thumbing his nose at conventional wisdom just to stir up an argument. In some of his essays, Joel takes a position against rewriting bad code, for the not invented here syndrome, against incentive pay, and for bloatware. Each of these essays is bound to tick someone off. But, in each case, his arguments are lucid and force you to re-examine the question in a new light. In the end, you may still not agree with him, or he may have changed your mind. Either way, he will have made you think about the issue more than you probably have before.

I would recommend this book to anyone who has been in software development for a while. I think new programmers are less likely to really get what Joel is trying to say. If you need to manage programmers, you also need to read this book. For those who have been here a while, it's an interesting and entertaining read.