Object-Oriented Multithreading Using C++

Rating:
Summary: In addition to my previous note
Review: I read the reviews on this page, and thought I have to add a couple of words to what I've already said. I think the negative reviews are somewhat unfair--there is nothing ungrammatical about the book (definitely not by today's standards of technical writing--go MS Press for the mainstream .) While the book contains much vapour and is wildly overpriced--imho--on the balance isn't bad. I wish it was half-the price and half the size though .

Rating:
Summary: Fits the pattern
Review: I've read previous books by the same author (Streams, for example.) There seems to be pattern here: a few very good ideas totally spoiled by abysmal writing, lack of editing, and unjustifiably high price. Not a useless book by any means, but purchase ONLY upon a personal perusement. Buying blind may result in disappointment.

Rating:
Summary: Absolutely the BEST!
Review: If you're doing Posix threads in OO and in C++, this book covers the intermediate concepts and builds to more advanced issues. Well thought out. Well designed. If you're doing (or going to do) a distributed, multiThreaded applications, you will thank God every day that you found this book before you started. Unless, of course, you wrote another book on Posix threads. I didn't so I did. :)

Rating:
Summary: Lots of Padding
Review: When I purchased this book, I was already familiar with the basics of multithreading and synchronization. My hope was a complete, authoritative manual to help me better develop technique and methodology. In the end I was satisfied that I had learned quite a bit from the authors, but this book is hardly a good value.

First of all, a reader with an extremely rudimentary understanding of multi-tasking will find nothing but review in the first five chapters (out of twelve) of the book. To be fair, the authors state in the preface: "If the reader is familiar with the fundamental C++ class types and the basic operation system concepts needed to understand threads, then the reader can skip Chapter 1 through 4." This is a bit of an understatement, in my opinion. If a reader understands at a basic level what a thread is and how it differs from a process, he or she will find nothing but (tedious) review in the first five chapters. Don't be fooled by the table of contents; any aparently interesting topics in the first five chapters will be covered in such shallow, abstract detail that you will probably get nothing from them. Additionally if you have done any experimentation with threads and understand basically what a mutex/semaphore/event/critical section is you can probably skip chapter six as well.

For me things did not really get interesting or informative until chapter seven at the earliest, maybe even chapter eight. From there on out I would consider the book to be pretty good. Some of it was a little out of place for a book on multithreading. For instance there is a long-winded CSci 101 explanation of what an interface class is at the beginning of chapter seven, which struck me as odd since this is well-covered territory in any OOP C++ book, and familiarity with C++ and OOP in general is a stated pre-condition for reading this book. As for the allegedly poor grammar, I can't honestly say mine is any better so I found it all strangely comforting.

In conclusion, much of this book will be review to the average reader. This fact combined with the high price tag does not make this book a good value. It is nice to have on the shelf because it is pretty comprehensive in content, but keep in mind that much of that content is so fundamentally basic to multithreading that you will find it in many of the more thorough general programming books. If you are looking for money well spent, this probably shouldn't be your choice.

<< 1 >>