Customer Reviews

The Elements of Networking Style: And Other Essays & Animadversions on the Art of Intercomputer Networking

5 star:		(2)
4 star:		(1)
3 star:		(1)
2 star:		(0)
1 star:		(0)

 

 

When ``Elements'' came out in 1985, I loved it. I referred to it as my favourite book critiquing networking in 1995. And I have mourned the fact that Prentice Hall let it go out of print. Padlipsky's essays are pungent and acerbic; he is opinionated and hilarious. If you want to know what's wrong with X.25 or why OSI failed, Padlipsky will let you know. A wonderful book!

I bought this book almost 11 years ago. I was on the Internet for about 2 years by that time. I was always intrigued by how the designers of the protocol suite designed the protocols for the Internet. I thought the book would tell me something about that.

Unfortunately, I did not understand most of the book. Not because the book isn't any good, but because I did not know enough to take advantage of the book. Still, I always remembered the author's opposition to X.25 and OSI.

The funny thing is, I went on to program X.25, OSI, and TCP/IP for a large telecom equipment company. I have to admit I totally agreed with the author by the end of my experience in that company.

I picked up the book again yesterday, thinking that maybe I would learn something more this time. Sure enough, every page and every paragraph is such a treat. I did not know it would take me 8 years of being in the networking industry to understand the book fully.

So, I heartily recommend this book to anybody who is even remotely connected to network programming. You don't have to agree with the author. But he does have a few points worth knowing.

I bought this book in 1985 when I was hip-deep in implementing the very protocols he writes against. It had a major impact on my thinking about protocols for the next 15 years and helped lead me out of the morass. Even when I did work on OSI protocols I was able to use the ideas to, hopefully, write a little cleaner and better than I would have otherwise.

The current architecture of computer internetworking is philosophically based on the ideas he presents. Without these ideas being so widely propagated we would live in a world dominated by X.25, X.400, and large vendors proprietary implementations of politically written "standards".

Thanks to MAP for writing it.

"The Elements of Networking Style" is an interesting classic in the area of networking. Originally published in 1985, the book contains a collection of papers which were published in several conferences. All papers relate to networking and most of them related to the Arpanet Reference Model (ARM) vs the OSI model. The book is not too easy to read, but incredibly funny and thoughtful though.

The book contains 11 chapters and 3 difference prefaces :) The first chapter discusses different styles in networks. It mentions prescriptive and descriptive styles and how OSI is a prescriptive style wheras ARM is more descriptive where the protocols got written to describe the successful implementation. Chapter 2 introduces the OSI model (really quick) and chapter 3 is the chapter after which the book got named, which discusses virtualization vs emulation and describes a how the telnet protocol has gone for a virtualization approach. Chapter 5 discusses the ARM model (in contrast with the OSI model). Chapter 6 and 11 are about the only chapters that don't attack the OSI model but bring different ideas: that of H-FE Protocol (an idea that doesn't seem very relevant anymore) and the discussion about a secure packet network.

The style in which this book is written is hilarious and probably different than any other technical book you read. The author himself calls it "constructively snotty". The author is very direct and not very sensitive ;P in his opinion about the OSI model, though he does come with good arguments (though, repeats the same arguments a bit too often). The book is especially interesting in historical perspective as it describes a world in which TCP/IP was 'officially' threatened by the OSI model and the X.25 protocol. In a historical perspective, the OSI model is rarely talked about and TCP/IP is the standard protocol on the internet, so the author has been proven correct by history.

I was thinking about rating the book 4 stars, but decided to go with 3. Why? Well, first of all, this book is definitively not for people who want to know about networking. It is an advanced book and it will be very difficult for people who are not familiar with OSI model and TCP/IP. Second, the book is good historically, but parts of the book are simply out-of-date (such as H-FE Protocol). Third, the book is not an easy read. The English used is not particularly easy (the author has a degree in English :P). It is written well, but not easy. Because of all that, if you are interested in history of networking then this is a must read! If you simply want to know more about networking, this is not the book for you. 4 stars originally, 3 stars due to partly out-of-date content.