Search in Book Reviews

The first thing that springs to mind on seeing the title is to ask if the author can possibly be serious. I think the answer is 'yes but misguided.' The concept of the '... in 10 minutes.' books is that in our crowded lives with a constant demand to be productive we do not have time to wade through long chapters when we only need an answer to a single question. If you are already familiar with the principles of and application (such as word processing) such a mechanism can work and work well. However anyone who honestly believes that they can learn something as complicated as C++ in 27 10-minute nuggets has to be unbelievably naive.

What I find disturbing is that an author with the experience of Jesse Liberty should lend his name to a book such as this one. Sadly, despite his high profile and his endorsement by the editors of C++ Report, which publishes a regular column by him, he is far from faultless. The book's back cover states that it covers ANSI/ISO Standard C++. This is simply untrue. Not only has he effectively ignored the whole of the C++ Standard Library apart from a rudimentary use of classic

iostream

s (pre ANSI/ISO) but he completely ignores

namespace

(in particular namespace

std

) and leaves exceptions till the chapter 27.

By the way, he makes the elementary error of confusing class definition with class declaration (

class A;

is a declaration while

class A { /* 
member declarations */ };

is a definition) Authors often defend such errors on the grounds that the distinction is unimportant to the reader. It may be when they start, but 'learning wrong' makes long term learning harder than it need be. In addition such errors are good indicators of the depth of an author's own understanding.

If you have any doubt that this book is superficial and completely inadequate even for the beginner at whom it is targeted, look up

new

in the index and read what the author has to say about it. Ignore the fact that he does not tell you what happens if there is insufficient memory (novices are unlikely to be bitten by this) but note that he has nothing to say on the subject of

new[]

(he does not even mention it.) and that is serious.