Customer Reviews

Objects Unencapsulated: Java, Eiffel, and C++ (Object and Component Technology)

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

 

 

I'm midway through this book, and thoroughly enjoying it. Joyner has a clear and accessible writing style, and the content is obviously well-researched and thorough. That, however, is my main problem with the book -- its core grows out of a paper of Joyner's on weaknesses in C++ and the resulting book, while still excellent, follows from that mindset. The book sometimes feels not so much like a comparision between C++, Java, and Eiffel, but rather an exhortation to avoid C++'s misfeatures (and, to a lesser extent, Java's), and instead use Eiffel. While the author makes an admirable case for Eiffel, this isn't why I bought the book, and the non-critical treatment that Eiffel receives undercuts the overall strength of the book -- no programming language is perfect, and if what Joyner really wants is better language tools and better language design, he owes it to himself and his audience to treat the objects of his comparison equally.

All that said, the book is well-organized, even the most loaded statements Joyner makes are thought-provoking, and its approach is unique. If you're willing to keep your critical thinking filters on and stay engaged with the book instead of being a passive reader, blindly trusting the author, you'll get a lot out of it.

Coming from a Unix/C background, I was at first a little wary of such a scathing criticism of C++. However, after letting go of my initial caution, I was amazed at how many problematic areas of the language existed. This book really does get down to what's really important in an OO programming language, and discusses the +ves and -ves of each language in great depth.

Even though a great deal of positive attention is given to Eiffel, I do not think this comes from bias: I think that it's a natural reaction to studying the language with an open mind. After reading up on Eiffel on the net, and wallowing in Bertrand Meyer's incredibly lucid book 'Object Oriented Software Construction', I was totally blown away with the design of Eiffel. It really does seem (IMHO) to be far and away the most comprehensive OO language available today. While no language is perfect, it is a quantum leap ahead of both Java and C++ in too many areas to mention.

If you're willing to come to this book without emotional attachment to any particular language, you'll find yourself on a wonderful adventure exploring an amazing (relatively) new OO language which is designed from scratch without compromise to be a bullet proof way of constructing top quality reliable, reusable, maintainable, functional software, from initial analysis and design, all the way to implementation and subsequent maintenance and extension.

I bought this book to help designing a C like language. But I was disappointed by the content because it does not bear too much useful information.

Most of the critics and comparisions are already well known. Nothing novel.

The book is more like a handy memo of the language features and good/bad tags than a serious study of the trade-offs and concerns (theoretical and practical)

Some comparisions are biased or even superfluous. The author tried to put everybody's feet into Eiffel's designer's shoes. Many of the practical concerns of C/C++/Java were intentionally or unintentionally ignored. That, in my opinion, undermines the book's objectiveness a lot.

Overall, this book is not good for beginners who want to learn C,C++,Java or Eiffel¡£It is just not written as a tutorial or textbook.

On the other hand, it is not of much value for serious language researchers either, in either academic or practical sense.

The places where Joyner discuss each programming language in terms of how they implement a certain OOP feature are very good and even important. But then he starts to compare languages...

The problem, as I suspected, is that Joyner goes on with a great number of paragraphs critisizing C++ features that was never intended to be an OOP paradigm feature. But when an (even important) Eiffel glitch is found, he simply states that a fix is around the corner. NOT FAIR!

You should also note that the editor of the series is Betrand Meyer (author of Eiffel.)

In general, it is not a good idea to make a comparison between C++, Java and Eiffel in the first place because their goals are so different. C++ is a multiparadign programming language for large system-level applications, but Joyner does not seem to reckognize this. All these programming languages have pros and cons, but you don't settle on one of them for all purposes.

That said, the good parts of this books are so good that I recommend it.

You may feel impressed with this book if you are disappointed with other books about object-oriented. This book is rather unusual. He mentioned points that typical OOP books don't cover, for example, a dilemma between flexibility and correctness.

Put simply, this book is not informative but argumentative. The author believes Eiffel is simply better than C++ and Java. He could seem to try to convince readers.

Unfortunately, this book is not well-organized. This book is filled with unique opinions about C++, Java, Effel unlike so-called textbook. Hence, you may feel unconfortable when you just want to understand a certain topic such as exception.

If you just want to learn (not study) OOP, another typical and common book is more helpful. But, this book certainly offers something you cannot know from typical books.