TCP/IP: Architecture, Protocols, and Implementation

Just a short note to express my appreciation for this wonderful book. I have attempted several times to read books on this protocol, but have never been able to return to them. Until now!The author has taken the very intelligent approach of weaving protocol structure, protocol mechanics and protocol behavior into an integrated description that is useful to the network engineer as well as to the programmer.All features of the protocols are explained in clear unambigous language. The reasons for each protocol construct are clearly explained. The book is entirely free of typos (to the best of my knowledge), and is very well produced with excellent diagrams and generous margins.

This book explains IPv4 quite well while providing me with current, practical information about IPv6. It has very little fluff and gets straight to the point. I also found the detailed coverage of all the upper layer protocols and the Winsock programming interface quite helpful.

I have both of Dr. Feit's previous editions. They have grown in maturity and breadth with each new release.I am a TCP/IP instructor. I commend this book to all my students for several reasons. First, it's as readable a text on a dry subject as you will find. Her writing style is easy to follow, even engaging. Equally important, she never leaves you hanging. If there are three possibilities, she explains all three. Many other technical books leave you wondering about that third possibility. She is a stickler for technical accuracy. If you read it in her book, you can take it to the bank.When I have a question, I turn to her book first. Most of the time, the answer is right there.I do on occasion have a very detailed question, and for that I refer to Stevens if necessary. However, his book is too detailed for most people and sometimes lacks perspective. In particular, his cursory treatment of BGP was not up to his usual high standard.In Dr. Feit's book, there are subnetting explanations, exercises (and answers!) in the back which I use all the time in class. They are well thought out and each one has a point to make. I have read Doug Comer's books. Students often bring his books to class to show me. As a net-head, I don't care for them. Way too detailed for net people, but great for TCP/IP programmers writing stacks, internals or apps requiring detailed API knowledge. Not great for people who run and maintain TCP/IP networks. I think this explains why the disparity in some of the reviews. In sum, if you want only one book, I recommend hers!

As with any book, its review depends on the readers needs and expectations. I had a solid ground in programming and computer skills, but little knowledge of client/server interaction and how messages got from A to B. The book provided everything I needed, plus more (if i need it as reference, good, if not, its thick so it will lok good on the shelf). It did seem to containing some repetative and uninformative examples, but that is likely due to my inability to understand deep technical knowledge on the subject at this stage. This brings me to another point. The book isnt a SAMS teach yourself before lunchtime (ie simple, although thats not fair to all SAMS books) and nor is it a desktop reference, it is both. It is introductory and it also contains solid, usefull technical information in depth. This is overall a really great book.