Customer Reviews

Concurrent and Distributed Computing in Java

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

 

 

I bought this book a couple of years ago now. Although it has some interesting discussion of distributed algorithms its big let-down is the abysmal Java code that it contains. I cannot believe for one moment that any of this compiles, let alone works - it all generally looks like pseduo-code written for the purposes of illustration rather being of any practical value.

Likewise I didn't find the material particularly helpfully structured or consistent in its level of detail. I would have liked to see more quantitive analysis of the suitability of different algorithms in practical applications and broader coverage of well-known distributed algorithms (for instance paxos and virtual synchrony are conspicuous in their absence).

I bought this book because I was interested in implementing
distributed algorithms. It has a good overview of the fundamental
algorithms and I found it helpful to see actual code. It's the only
book I know of that actually provides implementations rather than
pseudo code.

On the other hand, the algorithm descriptions were often condensed, and I
expected a more formal approach. I often found myself reading eagerly
to learn about an interesting variant of an algorithm only to find a
reference to a journal paper. It would be useful to cover a few more
variants more deeply.

Also, I was hoping to get an idea of an algorithm's usefulness and
performance in practice. Or is a particular algorithm more of
theoretical interest? The answer was not clear to me, although in
fairness, this was not the author's main purpose.

This is almost verbatim copy of other book by the same author, but with pseudocode replaced by Java. Rather poor Java, nonprofessional and hard to read Java. Using Java instead pseudocode doesn't convert the original theoretical book into one that is easier for practitioners. There is nothing directly practical in this book.

If you want to study distributed systems, don't buy this book. Buy other book by the same autor - this other book (Elements of Distributed Computing) is highly recommended

I give this book 2 stars out of respect for the great work of a dear colleague who's authored it, but the book really should have been provided as a supplement to his earlier (and very excellent book) on this topic.

I'm afraid the Java sample code in this book is horrendously bad (lack of adequate comments, unclear and esoteric syntax, inefficient implementations, etc.). There have been many other books on concurrent and distributed computing on Java before and after this book came out, so I suggest the reader buy this book for the theory, and use a more practical book for application. (Beginners in Java programming should also avoid looking at the code, as they will only pick up poor coding habits that will get them into trouble in a professional work environment).

I got this book because I have been a fan of Garg's Elements of Distributed Computing ever since I first read it. Concurrent and Distributed Computing in Java seems to be really great buy. It has a very strong theoretical foundation, but the emphasis seems to be on writing distributed programs and getting them to work. If you are a theory lover I wouldnt recommend ths book, but if you like implementing stuff then you will definitely like this book. The range of topics covered is quite comprehensive given the size of the book.