Understanding the New SQL: A Complete Guide

Rating:
Summary: Has its good points, but look for something better.
Review: Given the first author's position as editor of the SQL92 standard and the association with Jim Gray, this book has been a surprising disappointment. The back cover of my edition describes it as "a valuable reference", states that it will "clarify the differences between SQL89 and SQL92", and targets it to both newcomers to SQL and experienced programmers. On the positive side, it is attractively typeset, the paragraphs individually are readable, and the railroad-track syntax diagrams are attractive. Taken as a whole however, "Understanding the new SQL" falls down.

As a reference, it frustrates. A good reference needs a good index, but the attractive railroad-track diagrams are not set off from other entries in the index, greatly slowing searches. And while there is a complete grammar for SQL92 in an appendix, its organization defies use as a reference.

The index is generally inadequate as well. For example, most operator char! acters are not indexed at all, and the word "operators" only indexes a peripheral subject. I have yet to determine the stance of the standard on case-sensitivity.

Comparison with SQL89 appears only occasionally in the body of the book. Most of it is in a laundry list in an appendix and not useful for reference. The discussion is sometimes less than rigorous or uncritical, for example dismissing the implementation-defined precision of INTEGER in a single sentence. As another example, only in the discussion of LIKE is treatment of trailing blanks by the related '=' operator discussed. The book is organized generally by language features and syntax, and is not as strong in moving from problem to solution.

Rating:
Summary: Has its good points, but look for something better.
Review: Given the first author's position as editor of the SQL92 standard and the association with Jim Gray, this book has been a surprising disappointment. The back cover of my edition describes it as "a valuable reference", states that it will "clarify the differences between SQL89 and SQL92", and targets it to both newcomers to SQL and experienced programmers. On the positive side, it is attractively typeset, the paragraphs individually are readable, and the railroad-track syntax diagrams are attractive. Taken as a whole however, "Understanding the new SQL" falls down.

As a reference, it frustrates. A good reference needs a good index, but the attractive railroad-track diagrams are not set off from other entries in the index, greatly slowing searches. And while there is a complete grammar for SQL92 in an appendix, its organization defies use as a reference.

The index is generally inadequate as well. For example, most operator char! acters are not indexed at all, and the word "operators" only indexes a peripheral subject. I have yet to determine the stance of the standard on case-sensitivity.

Comparison with SQL89 appears only occasionally in the body of the book. Most of it is in a laundry list in an appendix and not useful for reference. The discussion is sometimes less than rigorous or uncritical, for example dismissing the implementation-defined precision of INTEGER in a single sentence. As another example, only in the discussion of LIKE is treatment of trailing blanks by the related '=' operator discussed. The book is organized generally by language features and syntax, and is not as strong in moving from problem to solution.

Rating:
Summary: Let's wait for the second edition
Review: I confess I was a bit disappointed with the book. I hope its second edition will be more 'Complete' than the first one. 'An Introduction to SQL' would be a more appropriate title.

Rating:
Summary: A very useful source of information
Review: This is a very useful book, and every professional user of SQL should read it. The authors demonstrate a clear understanding of the subject matter, and their writing style, with good examples, down-to-earth explanations, and the occasional touch of humor, is simply enjoyable. A particularly helpful feature of this book, is that syntax is shown in easy-to-read railroad diagrams.

It would have been helpful if the index was organized in such a way that one could immediately locate the railroad diagram for a given SQL statement. A separate list of railroad diagrams would also be nice. And although I consider heavy reliance on defaults to be bad programming practice, I would have appreciated it if just a quick look at a railroad diagram told me what defaults, if any, applied for the statement in question.

Any author of books or articles must have his or her target audience in mind. For example, what may be an excellent article for a trade magazine, would almost certainly be rejected by a scientific journal. Thus, when judging this book one must keep in mind the perspective of its authors. If you are looking for an academic textbook with emphasis on scientific rigor, then this book will disappoint you. It is pretty obvious that is not what this book was meant to be. If, on the other hand, you want a practically oriented book that will help you do your job in a better way, then this is a great place to start.

Having worked with databases and related technologies for 18 years, I have seen so many books where the number of pages outweigh the knowledge of their authors by a factor of 2 or more (and just in case there are any publishers out there listening; computer professionals are sick and tired of having to read through hundreds of pages of blah-blah-blah in order to find the information we are looking for). This book is different. It contains 394 pages of ordinary chapters, followed by 133 pages of appendices, and the signal-to-noise ratio in all these pages is excellent.

If you are an application programer, a database administrator, or some other kind of database practitioner, and you are serious about your work, then there are some books I would consider more or less mandatory reading. This is one of them.