Design Patterns: Elements of Reusable Object-Oriented Software (1994)

[I am only going to include 10 tech books, a represtative sample]

Compulsary reading. At the time, (when it first came out) this really affected me. I have always thought that IT is a backwards field, compared to making any other objects, and this is the first step in the right direction. ( )

A classic in the field of computer science, this book describes how object-oriented software can be used to develop meaningful (and reusable) classes. It is written by the so-called "gang of four" affiliated with a working group at the University of Illinois. It has achieved notoriety for its assistance in helping programmers write code more effectively in languages like C , SmallTalk, and (my current language) PHP. The book has started a large movement in identifying design principles that are inherent in all computer science problems.

The book reads like a catalog of class types. After an initial critique on the state of computer programming, it reads like a textbook and a reference manual to be used. It is academic, analytical, and brilliant. Obviously, the book is only for the serious programmer who aspires to master her/his trade. It is not for the casual programmer or even for the intermediate programmer. It does not hold the reader's metaphorical hand as the book is perused. Instead, it provides data and analysis like a research paper. Of course, in such a style, wisdom is distilled into compact sentences.

I'll be happy to put this book on my bookshelf at work and refer to it when designing software. ( )

It held my interest, but I find it hard to incorporate what seem like very abstract lessons like these into my actual programming practice. I guess I just don't think in OO. ( )

If there were required reading mandated for any degree involving computer software development, this would be it.

Gamma, et al (AKA The Gang of Four) have given us a comprehensive guide to various software patterns, and it's the least we could do to follow them.

It discourages me to see and hear about professional industries making lousy software that can't communicate with other tools, even those made by the same vendor. It bothers me when there are tools out there that make use of a network connection, but apparently don't play nicely with people using other operating systems or even other interfaces. Why can't I play Scrabble on my iPhone with someone using Facebook's Scrabble app? Why can't I play a co-op game of LittleBigPlanet on my PSP with another user playing on a PS3?

Well, the short answer is that you should be able to do these things, using interfacing, one of the fundamental patterns in this book. Just as the user (usually) doesn't care what's going on behind the curtain with their software applications, as long as they get the right feedback when the click the right buttons, the client-side application shouldn't care about what's going on in the server-side, as long as it gets the right feedback from the server when it sends it the right data. When you hear things like this for the first time, it's kind of eye-opening, but at the same time, when you think about it for a minute, it really makes sense. It's like when you're playing a puzzler game, and you need a hint about how to solve a puzzle; once you see the solution, you smack your head. "That's makes so much sense!" you say, "Why didn't I think of that." Well, consider this book the solution guide to software design.

No, it's not a recipe book (like the O'Reilly Cookbook series), but it does provide examples of the patterns you should be using, and unless you describe your coding style as "anarchist," it should really help you.

If you're a software engineer and get your hands dirty by mucking with software, and you haven't yet read this book, then take off those coding gloves and give yourself about a week to absorb the amazing knowledge this book presents. If you want to go into software engineering, ensure that you have a firm grasp of a particular programming language, and then read this book, preferably before you start interviewing for a job. Then, in your interview, when asked a technical question, you can start off your answer by saying, "Well, The Gang of Four said in their book on design patterns..." And then, you're working in some software engineering field (results may vary).

If you're truly devoted to making yourself a better software engineer, though, you simply must read this book. ( )

This book was really a game-changer. The art of creating a computer program had not been analyzed much at all up until that point - plenty of algorithms had been created by computer scientists, and systems designers had some ideas about API's, but there wasn't a whole lot of direction for people who wanted to engineer applications to do things that user wanted them to do. Some of the patterns in the book are in common use today; others have been relegated to the dustbin; but the larger point is that after this book came out, people started using a whole new language to describe the work they were doing. Today we have UI patterns, management patterns, data access patterns, antipatterns, patterns, patterns, patterns everywhere, and this was the book that really caused the concept to take off.

Is it worth reading today? That's a little tougher. There's no question to my mind that every developer should be familiar with the patterns in this book, but a more modern book on the same topic would probably display source code in a language that more people would more commonly be using. Still, there's no question that if you want to go back to the basics, you go back to Gamma. ( )