The Pragmatic Programmer: From Journeyman to Master [Paperback]

Programming is a rapidly changing craft. A machinist can learn to work a lathe or a milling machine, and expect that his knowledge will stand him in good stead for the rest of his working life. Not so for the craftsman programmer. Ours is a new craft. We are still learning how to do it. Having survived in the game for a decade or two, and having learned dozens of languages, operating systems, database management systems, transaction processing managers, editors, we come to the realisation that this is a hard game. Each of us learns skills that help us cope with all that change. We learn basic programming skills. We go on learning them. We learn to see what is coming, and move in anticipation. We learn what is important, and what is not. We watch those who are successful, and try to emulate them. We watch the unsuccessful with horrid fascination, and try to learn from them also. "There but for the grace of God go I!"

I don't know how to make an object oriented design. I can do design sketches. So I start from there. I build my system, dealing with the problems as they arise. I rely on my experience to keep me out of trouble. When I see commonality between two classes, I take the opportunity to refactor and eliminate the commonality. I am quite happy to rewrite any piece of code to make it better. You know what happens? I end up with a well-designed system despite myself. I am an opportunistic programmer.

I saw the title of this book, and thought: "That's me!"

So I bought it. What a disappointment! It is full of platitudes. It reads like a writer's style manual. It is good to do things this way. It is a bad idea to do it that way. It has no meat to it, no depth. If you want to know more about the topics they discuss, you won't find it in this book. You won't find much of it in the references either.

Let me quote a typical example from the section entitled "Text Manipulation". "Pragmatic Programmers manipulate text the same way woodworkers shape wood ... We need a general purpose text manipulation tool ... Text manipulation languages are to programming what routers are to woodworking. They are noisy, messy, and somewhat brute force. Make mistakes with them, and entire pieces can be ruined ... in the right hands, both routers and text manipulation languages can be incredibly powerful and versatile..." What rubbish! The analogy flows on, and is followed with the advice to learn a text manipulation language, and a list of things possible with such a language. There is not one practical example.

This continues for section after section. In Appendix A: Resources, the authors say "The only reason we were able to cover so much ground in this book is that we viewed many of our subjects from a high altitude. If we'd given them the in depth coverage they deserved, the book would have been ten times longer." All I can say is that they should have covered ten times fewer subjects, to the depth of coverage they deserved.

A journeyman programmer wanting to become a master is advised to study at the side of a master. Buy Kernighan and Pike's The Practice of Programming.

This is a decent but not spectacular book, which is written as a series of 46 five to ten page articles on various programming topics, such as "Orthogonality", "Design by Contract" and "The Requirements Pit". The segments are quite heavily cross-referenced (which I didn't find very useful).

The authors dole out a lot of solid advice, which is the book's strength. I found myself disagreeing with very little. There are memorable tips and some good stories. The writing style is also very accessible and conducive to diving in at any page.

The book seems a bit lightweight. The exercises are a little simple and I don't think you'll find yourself going back to this book a lot. Also, the typeface is annoyingly large.

I'd recommend this to someone who has done some programming and understands the syntax and fundamentals but hasn't gotten into programming larger pieces of software. I think that if you have programmed anything significant you won't find much new in here.