Just Enough Software Architecture: A Risk-Driven Approach [Hardcover]

George H. Fairbanks (Author)

Book Description

Publication Date: August 30, 2010

This is a practical guide for software developers, and different than other software architecture books. Here's why:

It teaches risk-driven architecting. There is no need for meticulous designs when risks are small, nor any excuse for sloppy designs when risks threaten your success. This book describes a way to do just enough architecture. It avoids the one-size-fits-all process tar pit with advice on how to tune your design effort based on the risks you face.

It democratizes architecture. This book seeks to make architecture relevant to all software developers. Developers need to understand how to use constraints as guiderails that ensure desired outcomes, and how seemingly small changes can affect a system's properties.

It cultivates declarative knowledge. There is a difference between being able to hit a ball and knowing why you are able to hit it, what psychologists refer to as procedural knowledge versus declarative knowledge. This book will make you more aware of what you have been doing and provide names for the concepts.

It emphasizes the engineering. This book focuses on the technical parts of software development and what developers do to ensure the system works not job titles or processes. It shows you how to build models and analyze architectures so that you can make principled design tradeoffs. It describes the techniques software designers use to reason about medium to large sized problems and points out where you can learn specialized techniques in more detail.

It provides practical advice. Software design decisions influence the architecture and vice versa. The approach in this book embraces drill-down/pop-up behavior by describing models that have various levels of abstraction, from architecture to data structure design.

Editorial Reviews

Review

The presentation is brisk, and the examples are frequent and interesting. This is a worthy book, a joy to read, and one I wish I had earlier in my career --- to help me avoid making so many mistakes! It is deserving of a place on your shelf if you want become a better software designer. --Dr. Timothy J. Halloran, Director of Engineering at SureLogic Inc.

If you're going to read only one book on software architecture, start with this one. --Michael Keeling, professional software engineer

This book reflects the author's rare mix of deep knowledge of software architecture concepts and extensive industry experience as a developer. If you're an architect, you will want the developers in your organization to read this book. If you're a developer, do read it. --Paulo Merson, Visiting Scientist, Software Engineering Institute

The Risk-Driven Model approach described in George Fairbanks' Just Enough Software Architecture has been applied to the eXtensible Information Modeler (XIM) project here at the NASA Johnson Space Center (JSC) with much success. It is a must for all members of the project, from project management to individual developers. In fact, it is a must for every developer's tool belt. The Code Model section and the anti-patterns alone are worth the cost of the book!--Christopher Dean, Chief Architect, XIM, Engineering Science Contract Group

This book shows how software architecture helps you build software instead of distracting from the job; the book lets you identify and address only those critical architectural concerns that would otherwise prevent you from writing code. --Dr. Kevin Bierhoff, professional software engineer

All in all, this book is easy to read, concise, yet rich with references -- a well-architected and finely-designed book! --Dr. Shang-Wen Cheng, flight software engineer

This book presents a unique view on software architecture that makes it both accessible and practical. The concepts of just enough architecture and risk-driven design are breakthrough ideas developed by Fairbanks. This book demonstrates how architectural principles can be effectively used to solve real world applications. I found it extremely useful and a must read for anyone working in software development. --Dr. Marcus Fontoura, Principal Research Scientist & Architect, Yahoo! Research.

This book [is] a uniquely practical and approachable contribution to the field of software architecture. For anyone who must create innovative software systems, for anyone who is faced with tough decisions about design tradeoffs, for anyone who must find an appropriate balance between agility and discipline -- in short, for almost any software engineer -- this is essential reading. --Dr. David Garlan, Professor, School of Computer Science, Carnegie Mellon University

From the Author

This is the book I wish I'd had when I started developing software. At the time, there were books on languages and books on object-oriented programming, but few books on design. Knowing the features of the Java or C++ language does not mean you can design a good object-oriented system, nor does knowing the Unified Modeling Language (UML) mean you can design a good system architecture.

Like many books, this is a book about software architecture, but this one is for developers.  In it, you'll find not only advice on how to think about design, but also hard-won advice from real projects.

This is not a book specifically about Agile Architecture, but it is fully compatible with Agile processes.  It shows you how to build up your conceptual model of architecture rather than shelves of documentation.

I hope you enjoy it. --George Fairbanks

See all Editorial Reviews

Product Details

• Hardcover: 376 pages
• Publisher: Marshall & Brainerd; 1 edition (August 30, 2010)
• Language: English
• ISBN-10: 0984618104
• ISBN-13: 978-0984618101
• Product Dimensions: 1.1 x 7.4 x 9.2 inches
• Shipping Weight: 2.1 pounds (View shipping rates and policies)
• Average Customer Review: 4.0 out of 5 stars  See all reviews (19 customer reviews)
• Amazon Best Sellers Rank: #282,173 in Books (See Top 100 in Books)

More About the Author

Dr. George Fairbanks has been teaching software architecture and object-oriented design for over a decade for companies including Kinetium, Valtech, and Platinum Technology. In the Spring of 2008 he was the co-instructor for the graduate software architecture course at Carnegie Mellon University.

He holds a Ph.D. in Software Engineering from Carnegie Mellon University, advised by David Garlan and Bill Scherlis. His dissertation introduced design fragments, a new way to specify and assure the correct use of frameworks through static analysis. He has publications on frameworks and software architecture in selective academic conferences, including OOPSLA and ICSE.

He has written production code for telephone switches, plugins for the Eclipse IDE, and everything for his own web dot-com startup. He maintains a network of Linux servers in his spare time.

George has been a program committee member for the Working International Conference on Software Architecture (WICSA), the International Conference on Software Maintenance (ICSM), and the European Conference on Software Architecture (ECSA). He has been a referee for IEEE Transactions on Software Engineering (TSE) and IEEE Software.

Most Helpful Customer Reviews

16 of 17 people found the following review helpful

5.0 out of 5 stars Highly recommended for novice and master October 18, 2010

By Olaf Zimmermann

Format:Hardcover

This a broad and deep book on all things software architecture except the architecting process (for process, one recommended read is "Process Software Architecting" by Eeles/Cripps). Several readerships will benefit from "Just Enough Software Architecture":
- As an experienced IT architect, I do not necessarily agree with everything in the book (this does not come as a surprise, as architects have opinions). That said, I certainly learned a lot that I can apply immediately on my projects and some of the more provocative statements challenge me to leave my comfort zone (or at least consider doing so).
- Junior architects can use the book both as a tutorial and as a reference when/while growing in their profession.
- Developers with a "who needs architects" mindset (hopefully) will understand architects and modelers much better after having read this book, and appreciate the value of archtecture.

Things I liked in particular:
- Overall vision and message of pragmatism sent
- The risk-driven approach increases chances to get accepted both in agile development and in more traditional architecture communities
- There is a lot of practical advice e.g. in Chapters 10, 11 and 15
- The author is in command of a large body of relevant related work (both industra and academia) and puts them in perspective adequately
- Editorial quality: structure, figures, command of the English language (some words and expressions a bit be hard to comprehend for non-native speaker)

Some room for improvements (2nd edition?):
- Not all metaphors and analogies work internationally, e.g.... not all IT people are sports fans that follow baseball or know what a rookie is
- The connection between parts 1 and 2 could be a bit stronger, even if loose coupling generally is a good thing; e.g., some more backward references

In summary, I'd say Just Enough Software Architecture is a highly recommended read for every architect in touch with development reality and every developer with a desire to build complex systems that will stand the test of time. Read more ›

9 of 10 people found the following review helpful

5.0 out of 5 stars Architecture as a risk reduction agent October 31, 2010

By Kenneth C. LaToza

Format:Hardcover

George Fairbanks book on "Just Enough Software Architecture" signals a directional shift away from architecture as a separate entity and movement toward viewing software architecture as holistic component of normal software development. Dr. Fairbanks outlines this shift by developing a thesis that software architecture is primarily a risk reduction agent and he contends that stakeholders should look for architectural views that highlight the risky parts of a software intensive system and understand how the architectural design reduces that risk.

To develop that thesis "Just Enough Software Architecture" covers architectural modeling, classification, styles and uses. Most of the chapters reinforce the risk reduction thesis either through examples or models where these principles are clarified. Dr. Fairbanks also uses a slight variation of the standard UML 2.0 notation, simplifying that notation in an attempt to show how even straightforward models can highlight the mechanics of the architecture.

One concept that "Just Enough Software Architecture" introduces is the concept of "architecture focused design", which is a deliberately chosen architecture to achieve a collection of acknowledged goals. The author espouses the concept that these goals are driven by the constraints of the system. These constraints should act as "guide rails" and they ultimately can be used to influence and direct the system or systems being developed. Many systems have hidden constraints (for example performance or security) and this architectural approach can help influence system implementers in making architecturally aligned development decisions....

Overall, I enjoyed George Fairbank's book and believe it forms an interesting discussion starting point for any organization performing software architectural work. I would hope that a companion volume is added that does a set of more detailed case studies of how actual risk-based architecture fares in practice. Read more ›

2 of 2 people found the following review helpful

4.0 out of 5 stars Just Enough is Plenty September 3, 2012

By Grown Up Geek

Format:Hardcover|Amazon Verified Purchase

Designing good software architecture is, in my opinion, undervalued. "Just Enough Software Architecture" gives specific and valuable evidence on the need and on the value of having appropriate architecture. Rather than the blind "document it all" attitude, it gives insight into how to decide what to architect and what to ignore.

It's a very thoughtful and thought provoking look at the design process.

2 of 2 people found the following review helpful

5.0 out of 5 stars Review of Just enough architecture February 20, 2012

By justenough_reviewer1

Format:Hardcover|Amazon Verified Purchase

So far I have read the first 7 chapters of the book. By what I have seen so far the book contents are interesting and relevant for architecture development. The book contains examples or case studies that makes it easier to understand the topics and see architecture development concepts in practice.

The book has hardcover and seems to be a book that will last long if well taken care of. This book will definitely reside on my bookshelf of books of reference.

1 of 1 people found the following review helpful

5.0 out of 5 stars Fantastic. Talking sense to anxiety. March 20, 2013

By Charles E Rider

Format:Hardcover|Amazon Verified Purchase

Pragmatism, thy name is software. I really appreciate the non-dogmatic approach to architecture. The simple premise that the more danger you are in the more tightly one wants to predict the outcome is simple. To those who build software every day it is a lovely antidote to those who value process over results and a lovely reality check to wave at those who value "organic" over everything else.
Thanks

1 of 1 people found the following review helpful

4.0 out of 5 stars An excellent overview and guide on practical software architecture January 10, 2013

By Theodore M Wong

Format:Hardcover

Having suffered through a project in which specifying architecture ad infinitum in part led to the death of the project, I found myself wishing that I had had a copy "Just Enough Architecture" at hand to restore sanity and get everyone to focus on the riskiest, least-understood parts of the system. The book is divided into two parts: "Risk-driven software architecture" and "Architecture Modeling". The first part discusses the process of creating software architecture, and emphasizes reducing "risk" (be it ability to deliver a finished product, ability to meet some quality-of-service metrics, etc.) as the driver of the process - once one has drawn out the architecture is sufficient detail to reduce risk to an acceptable level, one can move on to development. The second part discusses best-practice methods of representing software architecture. Critically, the book does not mandate a particular design and implementation process, but instead argues that the risk-driven architecture process is as applicable to traditional (outdated, in my opinion) waterfall design and implementation, as well as to more contemporary agile / XP processes.

Even a casual application of the risk-driven approach would do much to improve and simplify the development and maintenance of any software system, and I therefore recommend this book to any software engineer that aspires to lead a development project.