Customer Reviews

Object-Oriented Software Metrics

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

 

 

I liked the book's pragmatic approach to metrics. The author
presents a set of simple, code-based metrics and suggests what they may mean. There is an emphasis on finding your own way, using the book as a starting point. The book includes a "project experience database" that gives the reader an idea of what can be expected.

The book's approach is a lot less intense compared to other metrics books like those written by Capers Jones. The focus is much more on measuring object-oriented aspects of the code rather than on measuring the functionality as with Function Points.

Defines a set of 39 metrics classified into 9 project metrics and 30 design metrics. The author explains how design metrics can be used to detect quality problems and offers recommended thresholds. Advice is provided on how to correct problems with the design and source code. A project experience database displays metrics for real projects in Smalltalk and C++. The book is relatively short at 146 pages.

The metrics covered in the text are number of scenario scripts, number of key classes, number of support classes, average number of support classes per key class, number of subsystems, average person-days per class, average number of classes per developer, number of major iterations, number of contracts completed, number of message sends, number of statements, lines of code, average method size, method complexity, number of public instance methods in a class, number of instance methods in a class, average number of instance methods per class, number of instance variables in a class, average number of instance variables per class, number of class methods in a class, number of class variables in a class, class hierarchy nesting level, number of abstract classes, use of multiple inheritance, number of methods overridden by a subclass, number of methods inherited by a subclass, number of methods added by a subclass, class cohesion, global usage, average number of parameters per method, use of friend functions, percentage function-oriented code, average number of comment lines per method, average number of commented methods, number of problem reports per class or contract, class coupling, number of times a class is reused, and number of classes/methods thrown away.

There are few software metrics books available specifically targeted at OO technology, so it is nice to have this one. As another reviewer stated, this book is best used as a starting point for "finding your own way." It has some good ideas and justifications, some of which are backed up with numbers from actual projects, and some which are just ideas with no backup statistics.

However, at only 146 pages, with 52 of *those* being the Appendix and Index, this book is ridiculously overpriced. Buy a used copy. Or read the other review of this book where the person basically typed in the Table of Contents. That'll give you the list of suggested metrics, which are generally self-explanatory, so all you'll miss from the book is the narrative providing justification and background explanation of each metric.