Customer Reviews

Agile and Iterative Development: A Manager's Guide

5 star:		(34)
4 star:		(10)
3 star:		(5)
2 star:		(1)
1 star:		(1)

 

 

If ever there is a book that should be part of a college-level software engineering curriculum as well as carefully read by software engineering development and project managers this is it. Every major iterative development methodology is covered in complete detail, with an emphasis on Agile methods, and a solid business and technical case is provided for the general approach.

Why make a case for? As difficult as it may be to believe, the waterfall method is still prevalent despite the large body of literature on rapid, iterative development SDLCs. Indeed, I have worked in environments that claimed to embrace the RUP as the enterprise methodology in principle, yet in practice projects were planned and managed using the waterfall SDLC. Why the disconnect? Managers were set in their ways and had no true understanding of the mechanics or value of Agile and iterative development methods.

This book can change that because each major approach is carefully described using the following format for easy comparison and to clearly show strengths and weaknesses:
Method Overview
Lifecycle
Workproducts, Roles, and Practices
Values
Common Mistakes and Misunderstandings
Sample Projects
Process Mixtures
Adoption Strategies
Fact versus Fantasy
Strengths versus "Other"

More importantly, these approaches are placed in the context of the benefits of incremental delivery, with clearly presented evidence of the benefits, which is provided in Chapter 6.

Regardless of biases or preferences, any objective reader will come away with a clear sense of the meaning of 'Agile' and the power and value of iterative development. You will also come away with a good frame of reference with which to compare your own organization's approach to development and delivery, and how to improve it.

I was expecting a lot from this book, having read and enjoyed Larman's prior work. On the other hand, I expected it to be somewhat simplistic as the title implied the target group being managers, which I am not. One of these expectations was correct.

Larman's latest presents a wonderful introduction into what iterative and evolutionary development is about. The word "agile" in the title seems a bit displaced as the text mostly discusses about "iterative" and "evolutionary" rather than "agile", but that really is no big deal because what's inside the covers is pure gold for any one.

After a thorough introduction to the theory, Larman drops a bomb on the table; the chapter titled "Evidence" is worth the salt alone. Larman has collected an impressive list of references to early, large projects employing iterative and evolutionary development. He also reminds us how the creators of predictive planning based methods have themselves preferred an iterative approach from day one.

The book also packs nice descriptions of four iterative and evolutionary processes, namely XP, Scrum, UP, and Evo. The descriptions are clear but, to some degree, repetitive.

Although the chapter on evidence is definitely the gold chip, the last 70 pages proved to be a very pleasant surprise. Larman presents a list of practical tips and tricks for adopting and running iterative processes, as well as answers the toughest questions in a Q/A section.

Highly recommended. Have your boss read it as well.

Yes, indeed, Finally. Abundant proof in one book that the traditional waterfall approach is a terrible way to manage software projects, and is therefore slowly being displaced by agile and iterative approaches. Larman does a devastatingly thorough job of debunking waterfall once and for all.

The book cogently and painstakingly explains how several of waterfall's practices have been conclusively linked to project failures, and how, on the other hand, the practices of Agile and iterative methods like Scrum and XP reduce project risk. Larman summarizes research findings encompassing thousands of projects, and quotes the supporting opinions of standards bodies and industry thought-leaders. The net effect is compelling, to say the least.

If you are an Agile skeptic, this book may rattle your conviction. If you are fence-sitter, it may convince you. And if you already have Agile fire in the belly, then certainly this book will stoke that fire. After reading it, I am left wondering how intelligent, experienced software development management can justify the continued use of a process that has wasted so much money and caused so much pain.

Craig's book begins with an excellent presentation of the fundamental concepts behind agile development and follows with a strong "Evidence" chapter.

Next is the clear, easy-to-read comparison between the leading agile methodologies (XP, Scrum and UP) and Evo that illustrates their commonalities and differences.

The 'icing on the cake' is the "Practice Tips" chapter that contains many practical insights that I learned the hard way. I think it will be especially useful for project managers new to agile.

As a manager with over 3 years experience managing XP projects, my opinion is that this book is a must-have for any manager interested in agile and iterative development.

The following is a letter I wrote to someone about this book:

Carlton, You should get hold of Craig Larman's book "Iterative and Agile Management". It has some of the best information about the failings of up-front requirements that I have seen. He quotes from dozens of different peer-reviewed research studies that date back to the 70's and 80's showing that the vast majority of software project failure can be traced to up-front requirements and waterfall mentality.

This information is so significant that I can't believe it's not more widely known. These research studies were done by some of the best people in the industry, including people like Fred Brooks, Capers Jones, and Tom Gilb.

There are 50 pages of such information, all very easy to read. Show it to your managers. Show them the report that the U.S. Department of Defense lost nearly half their major software projects in the 70's and 80's because of up-front requirements analysis.

You need this book if:

- You need to show or convince your bosses, your team, your customers about the virtues of agile and iterative development.

- You need to stop the estimation games in your company by showing the need to use an agile method.

You don't need this book so much if:

- You, your team, your boss, your customers are already convinced about the needs of agile development (if you already convinved all of them, please tell me how you did this so easily :-) !!!)

If you need a more practical approach I also recommend the book Crystal Clear by Cockburn.

During the spring semester 2004, I am teaching a course in software engineering. As a major class project, we are developing an application that will scan C/C++ code looking for potential security problems. In my opinion, there is only one way that a class of this type can develop a project of any significance. That is using an agile/iterative development model, where there is a little design, a little coding, a little testing and then go back to design. When I taught software engineering last spring, we used the same model, but were not as agile. Our iterations were longer and we pushed some of the more difficult tasks to the end. As the students noted, "we coded carefully at the start, but then just wanted to get it done at the end."
While this scenario might seem to be a problem, I found it gratifying, because it is just like the real world. The authors of this book are also firmly set in the world of software development. While reading it, I was constantly saying to myself, "It is about time." The reason for this singular conversation was that they completely disrespect the waterfall model of software development. In retrospect the use of the waterfall model is similar to the strict use of the word engineering in software development. Namely, the beliefs that the practice of building software development is just like building a bridge or a building. By thinking that all of the parameters can first be determined and then you build the software, an enormous amount of time, effort and expense had been wasted. Software development is a very dynamic process, one where circumstances are in a constant state of oscillation that gets damped down to a limit point as the project nears completion.
The waterfall model is one that is implicitly taught in school as well, but the only way we get away with it is because most of the programs that students write are small, well within the bounds of having hard parameters. Therefore, it is possible to completely design the program before coding it. In my experience with students fresh out of college, the two concepts they have the most difficulty with in their first job is the constantly changing requirements and the fact that they will know only a small part of the complete application they are building. And so, all educators must place more emphasis on dealing with changing requirements, and this book is an excellent place to start.
Fortunately, the movement towards object-oriented programming and encapsulation has made the change to iterative development easier. A programmer no longer has to be as concerned about possible data and method interactions/conflicts as they had to be when everything was visible to all.
I was sold on the iterative method of software development over a decade ago, when I started a job as a software developer. We were building a new product and received changing requirements on a weekly and sometimes almost daily basis. Quite frankly, we had no choice but to adopt an agile development style. I wish I had had this book with me at that time, it would have saved us a lot of stumbling around as we tried to deal with everything.

Waterfall style projects do not work for software development.

We need to be responsive to a dynamic marketplace - embrace change, don't fight it. If you've struggled with rigid processes imposed on your software development team, or at the other extreme, if you've never known how some process can help your software team - then this is a great book for you.

Craig provides the evidence to support my view that software development is not a repeatable, conveyor belt production line. It is an exploratory endeavour. Sure there are constraints imposed on a software team - budget, resources, time. But they should be the boundaries that your team operates within. You should not try to be specific about the tasks your team will do, just be specific about their boundaries - what they won't do.

If you struggle to understand this concept, or struggle to make others understand, point them to Craig's chapter on evidence. It seems the software industry project managers have hoodwinked us into believing every small step in a project can and must be planned out, otherwise the team will fail. My experience, and Craig's research, supports the fact that this is incorrect. Projects can be run in an agile manner and will be successful.

I've been able to use this text to address other people's concerns, to support my observations and to help move a very rigid team to be more adaptive. In fact, I find it hard to locate the book now as it is being passed from one PM to the next! Hmmm, about time they bought their own copy...

This book is written in the same agile fashion as what its subject is. It is concise, easy to understand, extremly well illustrated with examples and research results demonstrating practical value of the iterative development.

I found its greatest strength to be:

A. It compares all the prominent iterative methods using practical examples as you would do it in real world.

B. Your manager can read it,(As the title suggest) and maybe even beleive it.

Adopting the methods presented in this book may be the best thing ever that happened to your projects.

"Agile and Iterative Development" is a 'must read' for anyone involved in leading, managing, or doing software development. I have been using iterative methodologies for many years. Despite this, not only did I learn many things I did not know about the process, but it also gave me very useful information and statistics on the benefits of using it that I found useful when defending the approach against more traditional waterfall methodologies. Most interesting is discussions on practical applications when dealing with customers/clients that expect waterfall - a problem I have been dealing with for a long time. I bought several copies to distribute to my staff and I even refer to it in my resume where I discuss my ideal development approach.
This book was read and re-read with enthusiasm - it sits deservedly on my classics shelf next to Brook's Mythical Man Month. Get it!