[.ca] Cocoon 2 Programming: Web Publishing with XML and Java (ISBN 0782141315)

Too little too late:
Six months ago, as no Cocoon book existed yet, this thin volume could have been interesting, but with the release of Moczar and Aston it is simply not worth the money anymore. It is often rather superficial and is packed with questionnable fillers à la Wrox: one really wonders why the authors and the publisher felt compelled to include useless 'introductions' to css, xhtml and xsl (and more) in a book on cocoon!

Dissapointing:
When I originally saw this book, I was intrigued by an alternative way of presenting this fairly complicated material and the seemingly relevant examples. I felt that it was generally incomplete and scattered. Take the section on custom Generators, for example. The example is derived from an example is a previous book written by another Sybex book on XML and JSP, so it seemed like a nice comparison with an alternative technology. The introduction to the example discusses the dynamics of a branching survey and why a custom Generator is sometimes necessary. Unfortunately, the example doesn't follow through on the promise or the lesson. The example code that I downloaded from the website turned out to be incomplete and inconsistent with the book. Although I eventually got the example running, the end result wasn't compelling enough for me to understand the purpose for writing the custom Generator in the first place. I was dissapointed and frustrated with this book. For those trying to learn about Cocoon I would suggest Zeigeler and Langham's New Riders book (Cocoon: Building XML Applications).

Desperately seeking editing:
What the authors of this book desperately needed was an editor. This book is a mess. There is no sense of flow. There is no attempt by the authors to explain a topic and then delve further into it building upon what they show us. Instead we get detail. We are told Cocoon is made up of these frameworks and these frameworks use these design patterns and are made up of these pieces which are made up of these pieces. And in the end we know no more than when we started. We are told Cocoon contains these Java classes and are given a sentence explaining each one. We are given lists of SAX classes with no clear explanation of why we should be interested. There are lots of tables and lists and charts that explain nothing. The book never gives a clear explanation of what Cocoon does or how Java fits into Cocoon. The chapter on LogicSheets is a perfect example. There are thirty three tables in this chapter going page after page with no clear explanation of what the tables mean. Then the example at the end of the chapter is so trivial as to be meaningless. In fact, the examples throughout the book are much too simplistic and don't demonstrate the power of Cocoon. There is no question that an enormous amount of research went into this book. It's a shame that the authors weren't able to present that research in a useful way.

... before disaster sets in.:
Deal with a thing while it is still nothing; keep a thing in order before disaster sets in. A quote from Lao Tzu on the cover of this book which the Sybex production team would have done well to have heeded. It does not help that the competition - Zeigeler and Langham's offering from New Riders - is both clear, structured and liberally sprinkled with examples. This only accentuates the contrast with this exhibit, which leaves the reader as confused after closing the covers as before he or she opened them. No, correction, make that more confused. There is no lack of substance here, but also no shape, no argument and no goal. Cocoon is a system where data flows naturally from generators, through transformers, and is dispatched on its way by a final serialiser component. Yet, here in chapter four, the authors announce unconvincingly that a explanation in reverse sequence is ... errr ... in order. Off they go explaining serialisers. Fine. Chapter five, bafflingly, skips transformers and discusses site maps. We finally reach generators in chapter eight. Is this the wrong end of the telescope or are we staring into the proverbial liquid filled boot? It goes on like this, avoiding any form of educative example and meandering though theory without ever fully explaining why and where. To use a generator I need to know what it generates, but I can search in vain for coverage of even a fraction of Cocoon's generators. The book almost hits its stride in a reasonable explanation of XSP but then blows it , with a chapter on logic sheets. How would you explain what a logic sheet is? Why, obvious! by listing twenty seven tables of unannotated data before offering any explanation how to put this information to use. Oh dear. I can imagine the work the writers put into this book. It is often very detailed and many long evenings must have been devoted to research. How sad that the editors at Sybex didn't insist that some shape should be battered into the manuscript before it hit the press.