Search in Book Reviews

This is a book that I'm going to love to hate. On the one hand the cowboy coder in me just wants to shoot from the hip. And on the other hand I really would like to have some of that 1 defect per 1000 lines of code delivered on time stuff. Having read the book, I wonder if "DRFS" might not have been an alternative title - "Development Reduced to Following Scripts". As far as I can tell, this is all about "getting it right". Whose right though? If you're trying to do things that have never been done before, then I'm not sure that a mechanistic approach is the answer. Such method would help with implementing the solution, but finding it? If you are far from the bleeding edge and your "getting it right" is synonymous with "not getting it wrong" then this may well be for you.

The other big problem that I had was with the examples. Clearly to get the best out of this book you need to do the examples. They seem to be designed to fit in with a 1 week course, lectures in the mornings and lab sessions in the afternoons. This means 3-4 hours for each example. I found that hard to fit into 'spare time', (which for me is usually in 1/2 hour to 1 hour chunks). That all assumes that you can even get to use the examples. You can download them from the Carnegie-Mellon SEI site. They are all based on Windows and Access. If you don't use Windows or don't have Access, then forget it. I didn't have Access when I started reading the book, but then I got a cheap copy through my employer. Not being familiar with it (and there are no explanations for complete neophytes) it was a bit of a struggle. No problem if you have an instructor at hand.