Customer Reviews

Why Programs Fail: A Guide to Systematic Debugging

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

 

 

"Why programs fail" attempts to provide a systematic approach to finding, reproducing, and fixing programming errors, with a strong focus on the automation of many debugging techniques. Zeller covers the whole debugging process:

- Problem-tracking systems are discussed, not only as tools for tracking and managing problem reports, but also as valuable idea repositories and requirements management systems.

- You will also find advice on how to set up automated tests that support debugging tasks.

- Apparently straightforward, reproducing problems can be harder that it seems, as "heisenbugs" testify (i.e. when debugging tools interfere with the problem so that it disappears when it is being observed).

- Delta debugging, an interesting application of the classical divide-and-conquer strategy, provides an automated method to simplify test cases (and focus on the truly relevant part of the problem).

- Applying the scientific method is the right way to debug (i.e. reasoning about programs to create hypotheses and performing experiments to validate or discard those hypotheses). Here, the use of a debugging logbook helps to make debugging explicit by writing down all hypotheses and observations.

- Plenty of techniques for creating hypothesis and determining the failure cause of an observed problem are covered, from static analysis tools and introducing assertions, to experimental techniques that try to make debugging more efficient.

"Why programs fail" is outstanding. Many interesting (and practical) ideas are explored. If you would like to improve your detective skills, this book is highly recommended.

This book is indeed a solid reference on a whole new approach to debugging, and I recommend it to every professional programmer.

I just wish they did a better revision work before publishing it, as scarcely past the third chapter you will have already ran into half a dozen bizarre errors. Misplaced or missing words are just the beginning; wait until you run into the misplaced paragraphs...

None of the errors take away this book's value, but thet are sort of a black eye in an otherwise beautiful figure.

Andreas Zeller created the GNU Data Display Debugger. That
fact set my expectations for this book. I was not
disappointed. Being a developer for over 15 years, I was
pleasantly surprised at the advances in debugging presented
in this book.

The great benefit of this book is that it uses the
scientific method to create a formal discipline for
debugging. This discipline can be automated in ways that
were unthought-of until recently. One example of this is
the DDCHANGE plug-in for Eclipse that automatically
identifies which of multiple code changes has introduced a
given bug.

I found no major faults in this book. The author's style
of writing is very enjoyable. The only thing I'd change is
to drop the second chapter as it contains material on defect
tracking that is covered elsewhere (unlike the rest of the
material where this book is pretty much the sole source of
information).

This is a fabulous book that any serious developer should
read.

That's the best book on debugging I have ever read. Zeller has put together a reference on the practice of debugging that reveals aspects of it that had not been throughly investigated thus far.

This book contains perhaps a bit too much general background. After all, there are already many shelfloads of books out there that cover things like design and project management.

Otherwise, there are lots of things that sound quite interesting. Plenty of references to real products (both commercial and freeware). Clearly the author has extensively researched the field (being one of the authors of DDD probably helped).

Delta debugging, parallel debugging and omniscient debugging all look very interesting. There seem to be more tools available for Java than for other languages. This is probably a reflection of the ease of debugging the JVM. As a C and C++ programmer, I was a bit envious of this situation.

As a practionner I am delighted to see a systematic method applied to software. Too often is writing software compared to art or magic, while more often that not it is a matter of rigour, analysis and rational thinking (intuition does not hurt of course). The author attacks the problem of dealing with software defects with method and his analysis is sound. This book is a very welcome help to developers but can also help the quality assurance departement.

I think this should be the title of this book, because "Andreas Zeller" not only talking about dry methods of managing and debugging software code but he teaches how to draw a full picture of your software project "from the design till delivering fixes and updates", this book helps you solve future software problems by building software that is fixable and manageable. "Andreas Zeller" is a good text book writer he move by subject from the bottom up smoothly to support deep understand, so, I ask him to write a book about "Software Design Patterns and System Resources Management" using same way followed in writing his book and with cooperation with the Author of "Design Patterns" "Christopher G. Lasater" he is a good writer too.

This book is great for both software testing engineers and software developers. I definitely recommend it for anyone interested in the field of developing or testing softwares - being beginners, intermediate, advanced or experts in the field.