Ilya Grigorik

This book introduces readers to many technologies - old and new (ish). It doesn't go very deep into any one particular technology, it is a great book that can act as a refresher, crash-course reference.

It discusses on how to optimize website's performance. I also suggest High Performance Web Sites and Even Faster Web Sites both by Steve Souders.

TCP, UDP, TLS, WiFi, and mobile networks sections / chapters have good amount of detail and some of it was new to me. The chapter on HTTP 2.0 was quite interesting and found it in-depth as well.

Though the language is simple and effective for the complex topics covered in this book, some parts are repetitive. For example, 'Latency, not bandwidth, is the performance bottleneck for most websites.', is mentioned in many times in the book.… (mere)

The first two parts of this book describe the four lower layers of the OSI stack with an emphasis on performance and wireless technology. The third part treats HTTP (including HTTP 2.0), also with a performance emphasis. The fourth part is on Javascript APIs for Networking in the Browser(XHR, SSE, WebSockets, and WebRTC); this part only hints at performance as an alibi.

The book starts weak, but then gradually improves in quality, until the last chapter (about WebRTC) is then again only mediocre. This means that the part about HTTP and the chapters on XHR, SSE, and WebSockets are quite readable.

If one is only interested in delivering performance improvements for a website this whole book may be a great help; especially if one needs to do so under time-pressure.

However, I have some serious objections against the book:

First, it is overly repetitive and long-winded. Key-phrases that the author seems to enjoy are sometimes repeated once every page. The whole content could be shortened to maybe 60% of the length.

Second, especially in the beginning, I had the feeling that this book was put together in a rush, which is a problem that I already observed with other web-related books from O'Reilly. Also, the book contains to many elaborations on history for my taste. These "brief histories of ..." only consist of shallow enumerations of Standards, which leads to the third and most significant objection:

Too often, I did not find that the book describes the underlying principles (as promised in the foreword). Often the explanations left me in an unclear state and I found them shallow. I would by no means be able to implement, say, a WebSocket client. Granted, this might not be necessary to setup a website. But it shows that the underlying principles are not conveyed... Well, I guess it depends on what your definition of "understanding the underlying principles" is. Everything is relative.… (mere)