Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Universal database management
Saracco C., Morgan Kaufmann Publishers Inc., San Francisco, CA, 1998. Type: Book (9781558605190)
Date Reviewed: Oct 1 1998

The universal database (UDB) has recently received a lot of attention in the database software market. The name comes from extending relational database management systems (RDBMSs) with object-oriented features, not all of which are consistent with the original model or intention of the RDBMS. Specifically, at the top of the list is support of new, complex data types such as collections, hierarchies, BLOBs (binary large objects capable of holding images, sound, and video), and user-defined data types, which are useful for representing real-world business objects and components. Add user-definable functions, including those purchased from a growing industry of vendors providing data widgets of all kinds, and the merger of the object-oriented and relational models looks like a real possibility. Finally, the UDB is defined to include those features from traditional relational databases--triggers, stored procedures, and built-in functions--that simulate the encapsulation and subordination of data to method, characteristic of object-oriented technology’s message-like invoking of methods to access and manipulate data.

Saracco’s outstanding book provides a high-level, product-neutral briefing on the ins and outs of current UDB science and practice. Chapter 1 motivates the UDB imperative as a way of dealing with complexity in both the business and technology worlds. The promise of object-oriented technology (OOT)--enhanced productivity, visibility, and order through reuse--is still a strong selling point. It is effectively marshalled in an overview of OOT that accompanies the author’s explanation of OOT basics.

Chapters 2, 3, 4, and 6 are a briefing on simple and complex data types, including columns that contain structured data. I was fascinated to read about table hierarchies, in which a table type is defined and a table instance is inherited from that type, and about embedding what amounts to a table-like array with separately named buckets in a column of a table. These are innovations or requirements that are currently in the lab and will be appearing in future releases of UDBs.

When new data types are introduced, class libraries take on new importance. Functions are needed to manipulate these data types, and unless a firm is prepared to undertake development from scratch, the purchase of such components makes good business sense. These components are treated in chapter 8, which includes spatial, time series, text, image, and video data types. Links from within relational technology to external files and data stores deserve a complete chapter of their own. This is not just because most of the data on the planet still resides in nonrelational formats, but because of that ultimate proof of concept of distributed computing, the Web. Making the Web look like a vast relational database requires a link to a URL from within the RDBMS. Likewise, the externalization of Web content requires links in the opposite direction. Although this part is still a work in progress, it is interesting to consider the direction of development. Subsequent chapters consider alternatives to UDB technology--traditional object-oriented databases and variations on them--a short set of business case considerations, and high-level product evaluation guidelines.

The strengths of this book include its generality, in the best sense of the term, as it deals with concepts rather than specific implementations; its abstraction from particular UDB products; and its discussion of features of UDBs that are still in the software laboratory. It also provides the beginning developer or database administrator with an overview of both object-oriented and relational technology. At the same time, the sophisticated technology aficionado will find material on complex data types and functions. More technically oriented managers will also find much that is useful to them.

The book’s weakness, if it has one, is likewise its generality: perhaps unfairly, the book inherits people’s high expectations of the underlying UDB products. Saying that anything is “universal” in this day and age seems to reflect a pride that goes before the fall, but that may be the vendors’ problem, not the author’s.

The text is nicely supplemented with a ten-page glossary; a five-page bibliography usefully divided into commercial, research, general, and object-oriented sections; a sampling of URLs; and an index. As a physical artifact, the text is superbly produced without error and with wide margins and bold captions that are likely to be helpful to busy professional readers. It will be helpful to those who need to know what universal databases are and to acquaint themselves with their essential and most promising features.

Reviewer:  Lou Agosta Review #: CR121997 (9810-0779)
Bookmark and Share
 
Db2 (H.2.3 ... )
 
 
Systems (H.2.4 )
 
Would you recommend this review?
yes
no
Other reviews under "Db2": Date
DB2
Ranade J., Sehgal M., Elkind P., Grossman J., McGraw-Hill, Inc., New York, NY, 1991. Type: Book (9780070512658)
Nov 1 1991
DB2 performance and development guide
Geller J., Van Nostrand Reinhold Co., New York, NY, 1991. Type: Book (9780442005269)
Nov 1 1992
QMF
Sayles J., Gargiulo G., QED Information Sciences, Inc., Wellesley, MA, 1991. Type: Book (9780894353765)
Oct 1 1992
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy