Customer Reviews

Joe Celko's Data, Measurements and Standards in SQL (Morgan Kaufmann Series in Data Management Systems)

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

 

 

Part 1 covers the need, principles, and variety of measurement standards and Part 2 provides examples of actual standards across many industries such as language codes and phone numbers. This book is extremely informative and easy-to-read (I like Joe's writing style), and I found myself laughing out loud at several points in the book most notably around beer and Mad Magazine's metric system. As a data analyst and modeler, one of my tasks is to work with the business to understand and document existing domains as well as new and improved enterprise/industry domains. I will use this book as a reference in this endeavor.

The marketing speak in the description of this book is a bit confusing, and I worry that it leaves potential buyers unaware of what the book is really about. The book's title doesn't help much in this regard, either.

Database store data -- everyone knows that. There is some art in properly modeling and typing data so that it is represented consistently, flexibly, and efficiently in a database. This book endeavors to catalog different commonly used types of data, suggest an appropriate representation for that data, and describe the implementation of that representation as applied in a SQL database management server.

The first half of the book discusses different approaches that, in general, can help with the storage and representation of data. What are different approaches to help us handle data that's obvious bad or out of bounds? How can we best encode different data? What are good ways to represent values of vastly different scale? How can we efficiently validate certain types of data?

The advice presented in these first chapters makes it easy to extend the ideas of validation, storage, and efficient reference to any data type the user might encounter.

The balance of the book is divided into short chapters that examine different types of data. The book describes standards that influence the data stored; for example, for storing postal codes, the author examines Canadian, US, and UK schemes and discusses how they can be best represented and validated in a database. Clothing sizes are discussed by visiting the different dimensions and industry standards which govern the measurement and specification of these dimensions. All in all, more than 35 different types of data, ranging from spatial information to personal attributes, are examined.

The book is invaluable to database practitioners as they are often experts in the field of database work, but not experts in the domain modeled by the database. Most database engineers are not postal carriers or tailors, for example. The book facilitates the analysis and implementation of systems which involve domain- or industry-specific data, and I don't think there are many resources like it available.