Fifty Years Of Relational And Other Database Writings

Fifty years of relational. It’s hard to believe the relational model has been around now for over half a century! But it has—it was born on August 19th, 1969, when Codd’s first database paper was published. And Chris Date has been involved with it for almost the whole of that time, working closely with Codd for many years and publishing the very first, and definitive, book on the subject in 1975. In this book’s title essay, Chris offers his own unique perspective (two chapters) on those fifty years. No database professional can afford to miss this one of a kind history. But there’s more to this book than just a little personal history. Another unique feature is an extensive and in depth discussion (nine chapters) of a variety of frequently asked questions on relational matters, covering such topics as mathematics and the relational model; relational algebra; predicates; relation valued attributes; keys and normalization; missing information; and the SQL language. Another part of the book offers detailed responses to critics (four chapters). Finally, the book also contains the text of several recent interviews with Chris Date, covering such matters as RM/V2, XML, NoSQL, The Third Manifesto, and how SQL came to dominate the database landscape.

Fifty years of relational. It's hard to believe the relational model has been around now for over half a century! But it has-it was born on August 19th, 1969, when Codd's first database paper was published. And Chris Date has been involved with it for almost the whole of that time, working closely with Codd for many years and publishing the very first, and definitive, book on the subject in 1975. In this book's title essay, Chris offers his own unique perspective (two chapters) on those fifty years. No database professional can afford to miss this one of a kind history. But there's more to this book than just a little personal history. Another unique feature is an extensive and in depth discussion (nine chapters) of a variety of frequently asked questions on relational matters, covering such topics as mathematics and the relational model; relational algebra; predicates; relation valued attributes; keys and normalization; missing information; and the SQL language. Another part of the book offers detailed responses to critics (four chapters). Finally, the book also contains the text of several recent interviews with Chris Date, covering such matters as RM/V2, XML, NoSQL, The Third Manifesto, and how SQL came to dominate the database landscape. About Chris: Chris Date has a stature that is unique in the database industry. He is best known for his textbook An Introduction to Database Systems (Addison-Wesley), which has sold some 900,000 copies at the time of writing. He enjoys a reputation that is second to none for his ability to explain complex technical issues in a clear and understandable fashion. He was inducted into the Computing Industry Hall of Fame in 2004.

Some things seem so obvious that they don’t need to be spelled out in detail. Or do they? In computing, at least (and probably in any discipline where accuracy and precision are important), it can be quite dangerous just to assume that some given concept is “obvious,” and indeed universally understood. Serious mistakes can happen that way! The first part of this book discusses features of the database field—equality, assignment, naming—where just such an assumption seems to have been made, and it describes some of the unfortunate mistakes that have occurred as a consequence. It also explains how and why the features in question aren’t quite as obvious as they might seem, and it offers some advice on how to work around the problems caused by assumptions to the contrary. Other parts of the book also deal with database issues where devoting some preliminary effort to spelling out exactly what the issues in question entailed could have led to much better interfaces and much more carefully designed languages. The issues discussed include redundancy and indeterminacy; persistence, encapsulation, and decapsulation; the ACID properties of transactions; and types vs. units of measure. Finally, the book also contains a detailed deconstruction of, and response to, various recent pronouncements from the database literature, all of them having to do with relational technology. Once again, the opinions expressed in those pronouncements might seem “obvious” to some people (to the writers at least, presumably), but the fact remains that they’re misleading at best, and in most cases just flat out wrong.

Along with its companion volume (Database Dreaming Volume II), this book offers a collection of essays on the general topic of relational databases and relational database technology. Most of those essays, though not all, have been published before, but only in journals and magazines that are now hard to find or in books that are now out of print. Here’s a lightly edited excerpt from the preface (so this is the author speaking): I went back and reviewed all of those early essays, looking for ones that seemed worth reviving (or, rather, revising and reviving) at this time. Of course, some of them definitely weren’t! However, out of a total of around 130 original papers, I did find some 20 or so that seemed to me worth preserving and hadn’t already been incorporated in, or superseded by, more recent books of mine. So I tracked down the original versions of those 20 or so papers and set to work. When I was done, though, I found I had somewhere in excess of 600 pages on my hands—too much, in my view, for just one book, and so I split them across two separate volumes. Highlights of the present volume include a discussion of the difficulties involved in providing a relational interface to a nonrelational system; a tutorial on the quantifiers and what happens to them under three-valued logic; an examination of the effect of user defined types on optimization; some thoughts on normalization and database design tools; and caveats regarding certain important database operators, especially outer join and negation.

Along with its companion volume (Database Dreaming Volume I), this book offers a collection of essays on the general topic of relational databases and relational database technology. Most of those essays, though not all, have been published before, but only in journals and magazines that are now hard to find or in books that are now out of print. Here’s a lightly edited excerpt from the preface (so this is the author speaking): I went back and reviewed all of those early essays, looking for ones that seemed worth reviving (or, rather, revising and reviving) at this time. Of course, some of them definitely weren’t! However, out of a total of around 130 original papers, I did find some 20 or so that seemed to me worth preserving and hadn’t already been incorporated in, or superseded by, more recent books of mine. So I tracked down the original versions of those 20 or so papers and set to work. When I was done, though, I found I had somewhere in excess of 600 pages on my hands—too much, in my view, for just one book, and so I split them across two separate volumes. Highlights of the present volume include a detailed explanation of the multiple assignment operator and why it’s so essential; an investigation into why object and database technologies are so much more different than they’re often made out to be; a critical examination of SQL’s support for pointers (“references”); a tutorial on the counterintuitive (but crucial) concept of tables with no columns; and an annotated and extended debate between the author and E. F. Codd, inventor of the relational model, on the subject of nulls and three-valued logic.

E. F. Codd’s relational model of data has been described as one of the three greatest inventions of all time (the other two being agriculture and the scientific method), and his receipt of the 1981 ACM Turing Award, the top award in computer science, for inventing it was thoroughly deserved. The papers in which Codd first described his model were staggering in their originality; they had, and continue to have, a huge impact on just about every aspect of the way we do business in the world today. And yet few people, even in the professional database community, are truly familiar with those papers. This book—a thorough overhaul and rewrite of an earlier book by the same name—is an attempt to remedy this sorry state of affairs. In it, well known author C. J. Date provides a detailed examination of all of Codd’s major database publications, explaining the nature of his contribution in depth, and in particular highlighting not only the many things he got right but also some of the things he got wrong. Database theory and practice have evolved considerably since Codd first defined his relational model, back in 1969. This book draws on decades of experience to present the most up to date treatment of the material possible. Anyone with a professional interest in databases can benefit from the insights it contains. The book is product independent.

C. J. Date is one of the founding fathers of the relational database field. Many of today’s seasoned database professionals "grew up" on Date’s writings. Those same professionals, along with other serious database students and practitioners, form the core audience for Date’s ongoing writing efforts. Date on Database: Writings 2000-2006 is a compilation of Date’s most significant articles and papers over the past seven years. It gives readers a one-stop place in which to find Date’s latest thinking on relational technology. Many papers are not easily found outside this book.

Introduced forty years ago, relational databases proved unusually succe- ful and durable. However, relational database systems were not designed for modern applications and computers. As a result, specialized database systems now proliferate trying to capture various pieces of the database market. Database research is pulled into di?erent directions, and speci- ized database conferences are created. Yet the current chaos in databases is likely only temporary because every technology, including databases, becomes standardized over time. The history of databases shows periods of chaos followed by periods of dominant technologies. For example, in the early days of computing, users stored their data in text ?les in any format and organization they wanted. These early days were followed by information retrieval systems, which required some structure for text documents, such as a title, authors, and a publisher. The information retrieval systems were followed by database systems, which added even more structure to the data and made querying easier. In the late 1990s, the emergence of the Internet brought a period of relative chaos and interest in unstructured and “semistructured data” as it wasenvisionedthateverywebpagewouldbelikeapageinabook.However, with the growing maturity of the Internet, the interest in structured data was regained because the most popular websites are, in fact, based on databases. The question is not whether future data stores need structure but what structure they need.