Customer Reviews

The Business Rule Book: Classifying, Defining and Modeling Rules, Version 4.0

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

 

 

This is not a book one simply reads. It does for business rules what Ted Codd's 'Relational Model Version 2' did for data management: details an incredibly complete and robust foundation for an entirely new discipline, but possibly little else. Ross does a fair job of suggesting why the new discipline is important, but doesn't prepare the reader for (what I found to be) an entirely new and foreign language. With over 10 years of data modeling experience, I am still unable to decipher his definitions of data type and entity. And in spite of a half-dozen commonly used notations for ERDs, I think Ross manages to introduce yet another one.

By page 16 I felt completely overwhelmed, with no sense of what problem was being solved with, for example, the Table of Atomic Rule Types (even though it's clearly pretty cool), and no sense of how rules would be used after they were so carefully captured, identified, classified, and normalized. My mental picture was the specter of pouring 1000s of labor hours into building a taxonomically-correct butterfly collection, and proclaiming, "Here are the rules of your business. I know all their names and I've organized them properly." With a data admin background, I have less of a problem doing the very same thing with data...but I know conceptually how my work can be implemented and utility achieved. Not so with this material.

My biggest disappointment: the book doesn't provide a conceptual architectural implementation. Without some foundation portrayed in architectural components (e.g., a rulebase and its possible structure, a rulebase management system, rule interpreter or rule object broker, etc.) there's little motivation to move forward.

Again, like Codd's RM Version 2, Ross's work seems to be very 'important' and very intellectually appealing. But, it too leaves me with the feeling of, "So what"? On the other hand, Ross's own presentations and those of other business rule proponents leave me with no doubt about the soundness and potential of the overall approach--it could be a huge IT breakthrough. Unfortunately, this book did nothing at all to help me understand how that breakthrough might be possible, or how the effort to capture rules could be practically applied to solve business problems.

CASE tools with code generators have attacked the problem of requirements specification in a piecemeal manner. This has led to significant
(and unpredictable) gaps in the ability to generate useable code from the specifications.

Ronald Ross presents what could best be described as a unified field theory to explain how to define business rules. His generic way to define precise and unambiguous rules (and in a graphical manner!) lays the cornerstone for truly significant CASE tool breakthroughs in capability.

Like Das Kapital by Karl Marx, this is a long book, it's full of powerful insights that can change the face of analysis as we know it; and it can best be described as tediously brilliant. If you are looking for an analytical razor to hone your specification skills, this is well worth slogging through!

As designs for middleware software layers increase with the arrival of Web Services, a crucial question arises about how to enforce server side business rules while executing well-formed transactions. In "What, Not How," C.J. Date gives examples of how to support transactions declaratively.

Unfortunately, most commercial off the shelf software does not allow for such possibilities, which necessitates the need for procedural enforcement of business rules through the use of properly designed triggers and stored procedures. The tome from Donald Ross gives a formal methodology for the enforcement of such business rules in the design of a software application.

It lists eight categories of business rules which can be created in an Object/Relational environment. These can be used to provide an integrated approach to transactions for the application to be developed. He provides a formal graphic set of design tools which can be used to address these procedurally on data.

The author provides hundreds of examples of business rules in the designated categories. It should prove a valuable reference for integrated application design for all concerned architects, developers, and subject matter experts.