Preface

Languages shape the way we think. How we approach problems and formulate solutions for them depends on the concepts we can express using language. This is also true for programming languages. Given a problem, the programs written to solve it may differ from one language to another. This book is about writing programs by expressing concurrent algorithms in the Go language, and about understanding how these programs behave.

Go differs from many popular languages by its emphasis on comprehensibility. This is not the same as readability. Many programs written in easy-to-read languages are not understandable. In the past, I also fell into the trap of writing well-organized programs using frameworks that make programming easy. The problem with that approach is that once writing is over, the program starts a life of its own, and others take over its maintenance. The tribal knowledge that evolved during the development phase is lost, and the team is left with a program that they cannot understand without the help of the last person remaining from the original development team. Developing a program is not that much different from writing a novel. A novel is written so that it can be read by others. So are the programs. If nobody can understand your program, it will not age well.

This book will attempt to explain how to think in the Go language using concurrency constructs so you can understand how the program will behave when you are given a piece of code, and others can understand what you produce. It starts with a high-level overview of concurrency and Go’s treatment of it. It will then work on several data processing problems using concurrent algorithms. After all, programs are written to deal with data. I hope that seeing how concurrency patterns develop organically while solving real-life problems can help you acquire the skills to use the language efficiently and effectively. Later chapters will work on more examples involving timing, periodic tasks, server programming, streaming, and practical uses of atomics. The last chapter will talk about troubleshooting, debugging, and additional instrumentation useful for scalability.

It is impossible to cover all topics related to concurrency in a single book. There are many areas left unexplored. However, I am confident that once you work through the examples, you will have more confidence in solving problems using concurrency. Everybody says concurrency is hard. Using the language correctly makes it easier to produce correct programs. The rule of thumb you should always remember is that correctness comes before performance. So, make it work right first, then you can make it work faster.