Search in Book Reviews

Even given the constraints of novelty, this is still a mishmash. Some of the papers are so wide of the mark as to be almost comical. Emily A Vander Veer supplies a paper on JavaScript, which says, pretty much, 'JavaScript is jolly useful for making web pages'. Young Park and Ping Bai write about a technique for retrieving a function from a library by specifying by example what the function is required to do, which is very interesting but hardly seems widely applicable. Frank Buschmann and Peter Sommerlad tell us that Patterns are a Good Thing.

More on the mark is the first of two papers by the pairing of Cuno Pfister and Clemens Szyperski, which actually deals with the crux of Component-Based SE and what differentiates it from a simple OO approach. The second paper is more of a technical sell of their product, Oberon/F. (Szyperski's 'Component Software: Beyond Object-Oriented Programming' is reckoned worth looking at, but I haven't got my hands on that yet). Generally, the papers are too short to have any depth; there are 18 papers in 152 pages. Since there is some introductory material in each one, such as what CORBA is or what OO offers, there is actually very little meat at all. Depending on your interests, you are likely to find no more than a third of it worth reading, perhaps 50 pages at most, for the thirty pounds it costs. Given that it is all two years old as well (and consequently some of the URLs quoted are broken), not a very attractive package.

However, there is a very worthwhile paper in it; the first one, by David Smyth of NASA Jet Propulsion Laboratory. He manages to be both witty and insightful in a short (5-page) paper. It is a shame this one paper isn't available in some cheaper form. In summary, don't bother with this book.