One of the difficulties I had when trying to learn Node.js was how to get started. The references that I found either dealed with quasi-academic topics such as datagrams and event emitters, or else myopically focused on a topic without regard for the big picture. This book takes you through the complete process of building an application in Node.js. It starts with the canonical “HelloWorld” example, and goes on to build a real-time web application capable of sending trading information to thousands of connected clients.

What make Node.js different? First, it provides a unified language between the back end and front end. This means that all your thinking can be in a single language, with no cognitive overhead when switching from front end to back. Furthermore, it allows for shared code libraries and templates. This opens up a lot of interesting possibilities, the surface of which is just beginning to be properly explored.

Second, it’s fast. One of the common complaints of interpreted languages such as PHP, Python, and Ruby is speed. Jason Hoffman, CTO of Joyent, has discussed how Node.js is at the point where its performance can break operating systems. A single core with less than 1GB of RAM is capable of handling 10GB of traffic and one million connected end points. Combining 24 of these into a single machine produces an overall level of throughput that exceeds the capacity of operating systems and TCP/IP stacks. In other words, with a properly designed application it’s not Node.js that’s the bottleneck—it’s your operating system.

Third, its nonblocking architecture is made for the real-time web. JavaScript was chosen as a language because it’s based on nonblocking callbacks and has a very small core API. This means it was possible to build the entire Node.js ecosystem around nonblocking packages, of which there are currently in excess of ten thousand. The end result is a platform and ecosystem that architecturally fits perfectly with the modern demands of the real-time web.