Review By: Carolyn Rodda Lincoln
12/03/2002

This book addresses the issue of what to do with all the systems in production that no longer meet business needs but that house valuable information about the business. The answer is legacy architecture transformation, which is the “process of modifying the form, design, and/or function of one or more legacy applications and/or data structures.” The book is divided into three parts with four chapters each.

Part I includes the background (why legacy systems need to be transformed), current IT technologies (such as the Unified Modeling Language and agile methodologies), and some architectural strategies. The strategies include 1) application analysis and re-documentation; 2) selective application of enterprise application integration; 3) program-level and system-wide improvements in legacy systems; 4) language migration; 5) data consolidation, migration, and cleanup; 6) migrating legacy systems to modern platforms; 7) applying modern distributed front ends to legacy applications; 8) redesign and reuse of legacy business logic; 9) package software/legacy application hybrid deployment; and (10) augmentation of new development efforts through legacy analysis and reuse.

Part II is how to create the infrastructure for legacy transformation and then how to plan for it. Part of the planning is justifying a legacy transformation project and assessing the current systems against the proposed architecture.

Part III is how to implement the transformation. The book also includes a list of tools by category (business modeling, source program analyzers, etc.) and information about the vendors of the tools.

This book is very good for its intended purpose—to assist planners in transforming legacy systems. The audience is IT management and system architects rather than project teams. It provides ideas on how to plan and implement transformation projects in a logical way without suggesting any one “right” method. The author emphasizes that all projects, including transformation ones, must provide business value. The only method that is not recommended is a “big bang” approach where a company tries to transform everything at once. It is clear in many studies that the “big bang” approach never works, but only wastes significant time and resources. Using an incremental approach, however, requires more planning, so that most projects not only accomplish the business purpose but also improve the infrastructure.

The author provides a wealth of information on possible strategies, but they may quickly become outdated. The book includes descriptions of technologies through 2002 when it was published. It is an excellent resource for a topic that is too often ignored while the team is trying to quickly implement the next crucial business capability.