Chris Date Seminars

Abstract:

Two recent trends: the plummeting cost of storage and the widespread adoption of data warehouse technology, have led to an increasing interest in temporal databases; indeed, the idea of maintaining and processing historical data has become not just a goal but a reality for many organizations. As a consequence, the ability to deal properly with the time dimension in databases has become an increasingly important practical problem. Yet today's DBMS products offer absolutely nothing to help with this important requirement. What's more, the database research community has largely failed in this regard as well; many approaches have been proposed, but they have all have proved deficient in one way or another. Help is on its way, however. This seminar describes a new approach to the problem that looks set to address the (surprisingly complicated!) issue of proper temporal support ─ an approach that, let it be said immediately, fits squarely into the classical relational tradition.

Note: Since no commercial products support that new approach as yet, this seminar might be regarded as somewhat theoretical. But the relational model too was once "just theory"! Learning about the new temporal approach now is like learning about the relational model before there were any relational products. And just as relational knowledge was helpful (with database design, for example) even before there were any relational products, so temporal knowledge can be helpful with similar matters now, even in the current absence of temporal products.

A more detailed breakdown of this seminar can be found here.

(return to top)

4. Go Faster! The TransRelational Approach to DBMS Implementation (1 day / 6 hours)

Extended Version

ABOUT THIS SEMINAR:

In the field of scientific endeavor, an idea emerges from time to time that is so startlingly novel, and so dramatically better than anything that went before, that it can truly be described as a breakthrough. The relational model provides the obvious example in the database world; almost everything done in that world since the relational model came along stands as testament to the radical nature and impact of that one brilliant idea. And now we are witnessing the birth of what looks set to be another major breakthrough: The TransRelational^tm Model. In this speaker's opinion, the TransRelational model is likely to prove the most significant advance in this field since Codd gave us the relational model, over 40 years ago.

So what is the TransRelational model? In essence, it is an implementation technology; it represents among other things a radically new and elegant approach to the question of implementing the relational model. When the relational model first appeared, skeptics accused it of being impossible to implement efficiently. Over the next few decades, therefore, vast sums of money were spent on attempts to prove the skeptics wrong. And those efforts did not prove in vain; today's products do perform fairly well, if not always spectacularly. But at what cost? The products have become quite notoriously unwieldy and difficult to manage, with their huge array of storage structures, access methods, optimization techniques, tuning knobs, analysis utilities, performance hints, installation options, and so on, and the job of the database administrator has become virtually impossible.

The TransRelational model has the potential to change all that. It provides a totally new approach to implementation, one that is dramatically different from those that have been tried in the past and found wanting (including all of the approaches encountered in today's SQL products). Such an implementation would be orders of magnitude faster than, and would deliver a far greater degree of data independence than, today's SQL products. It would also greatly simplify the job of the DBA.

Here are a few more specific features of the TransRelational approach:

The data is effectively stored in many different sort orders at the same time.
The data occupies a fraction of the space required for a conventional database today.
Indexes and other conventional access paths are completely unnecessary.
Optimization is much simpler than it is with conventional systems; often, there is just one obviously best way to implement any given relational operation. In particular, the need for cost-based optimizing is eliminated.
There is no need to compile database requests ahead of time for performance.
Join performance is linear!!─meaning, in effect, that the time it takes to join twenty relations is only twice the time it takes to join ten (loosely speaking). It also means that joining twenty relations, if necessary, is feasible in the first place; in other words, the system is scalable.
Logical design can be done properly (in particular, there is never any need to "denormalize for performance").
Physical database design can be almost completely automated.
Database reorganization as conventionally understood is completely unnecessary.
In general, far fewer human decisions are needed.

In a nutshell, the TransRelational model allows us to build DBMSs that ─ at last! ─ truly deliver on the full promise of the relational model.

A more detailed breakdown of this seminar can be found here.

(return to top)

5. Go Faster! The TransRelational Approach, An Introduction (1 - 1.5 hours)

ABSTRACT

This presentation describes a radically new and elegant technology for implementing computerized data storage and retrieval systems—which is to say, primarily, database management systems (DBMSs). In the speaker's considered opinion, the technology in question is one of the most significant advances (quite possibly the most significant advance) in the data management field since E. F. Codd first invented the relational model, over 40 years ago. In particular, it represents among other things a highly effective way to implement the relational model: a way that's dramatically different from ways that have been tried in the past and found wanting (including all of the ways encountered in today's mainstream SQL products). Such an implementation would be orders of magnitude faster than, and would deliver a far greater degree of data independence than, today's SQL products. In fact, the technology should allow us to build DBMSs that—at last!--truly deliver on the full promise of the relational model.

Topic outline:
· Preliminaries
· The historical context
· Three levels of abstraction
· Core concepts
· Major-to-minor orderings
· Condensed columns
· Merged columns
· Relational operators
· Review and analysis

Duration: 1½ hours (approx.).

Who should attend: Anyone interested in the field of data management who wants to learn about a truly dramatic new development in that field should benefit from this presentation. The only prerequisites are an elementary appreciation of data management concepts, techniques, and issues, including some knowledge of the relational model in particular.

(Download this seminar description in PDF for printing here)

(return to top)

6. Type Inheritance (1 day / 6 hours)

an indepth technical seminar

ABOUT THIS SEMINAR

The basic ideas of type inheritance have been around for several years; indeed, languages and products supporting such ideas, in one form or another, have also been available for some time. However, those existing languages and products are typically quite ad hoc, and they all exhibit surprising—not to say undesirable—behavior on occasion. Why? Because there has been, until now, no consensus on any well defined (i.e., formal, abstract, rigorous, and robust) type inheritance model. As a result, every language and every product has gone its own way, with predictable consequences. Moreover, (a) the concepts of type inheritance and related matters are much more complex than they might appear at first sight, and (b) much of the existing literature on this subject is very confusing (not to say confused). For such reasons, the authors of The Third Manifesto, Chris Date and Hugh Darwen, have been more or less forced to develop an inheritance model of their own. This seminar describes that model in detail.

Please understand immediately that the model in question is not meant to be just an academic exercise. Rather, it’s proposed as a serious contender for filling the gap alluded to above (i.e., as a candidate for the role of an inheritance model that is “formal, abstract, rigorous, and robust” and can be generally agreed upon by the community at large). It is offered here in that spirit.

A more detailed breakdown of this seminar can be found here.

(return to top)

7. The Third Manifesto: Databases, Types and the Relational Model (2-3 hours)

a technical presentation

"The Third Manifesto", by Chris Date and Hugh Darwen, is a detailed proposal for a solid foundation for the future of data and database management systems (DBMSs). Like Codd's original papers, it can be seen as an abstract blueprint for the design of a DBMS and the language interface to such a DBMS. Among other things, it lays the foundation for a logically correct approach to integrating object and relational technologies, or in other words to building what are sometimes called object/relational (O/R) DBMSs. However, it is not limited to O/R databases as such but is applicable to other kinds of databases also: for example, temporal databases, spatial databases, and databases used in connection with the World Wide Web. It is also applicable to the design of rule engines, also known as business logic servers, which some regard as the next generation of general-purpose DBMS products.

Note: It goes without saying that the Manifesto rests squarely in the classical relational tradition. In other words, it is in no way intended to supersede the relational model, nor does it do so; rather, it uses the relational model as a base on which to build. The relational model is still highly relevant to database theory and practice and will remain so for as far out as anyone can see. The Manifesto is thus very much in the spirit of Codd's original work and continues along the path he originally laid down. We are interested in evolution, not revolution.

DURATION: 1.5-3 hours.

PREREQUISITES: Attendees will be expected to be professionally interested in database technology.

(Download this seminar description in PDF format for printing from here)

(return to top)

8. Database Graffiti - Scribbles from the Askew Wall (1-1.5 hours)

an extended presentation,
based in part on an article that appeared in
Database Programming & Design 10, No. 3, March 1997
(the 10th anniversary issue)

ABSTRACT:

This session is based in part on one of Chris Date's regular columns in Database Programming & Design (the tenth anniversary issue), but includes much additional material. It consists of a series of quotations, aphorisms, and anecdotes - seasoned with a fair degree of personal commentary - that are (mostly) relevant to the general subject of database management. The session is not technically deep, but several serious messages do lie not too far below the surface. The aim is partly to edify, partly just to amuse.

TOPIC OUTLINE:

The prehistoric era

Objects and objections

Normalization, networks, and nulls

The role of simplicity

The joy of self-reference

Relational misconceptions

Some good quotes

Books and book reviews

Miscellany

DURATION: 1-1.5 hours.

PREREQUISITES: Attendees will be expected to have a broad familiarity with database technology in general.

(Download this seminar description in PDF format for printing from here)

(return to top)

9. The History of the Relational Database, A Personal Perspective (1 hour)

a keynote presentation, originally prepared for the 25th anniversary of the relational model

ABSTRACT:

Chris Date offers his own personal view of some of the most significant events in the first 25 years of relational database history.

TOPIC OUTLINE:

The foundation
Importance of Codd's original work
Other major contributions
A multibillion dollar industry
The historical context
The battles for acceptance
Major milestones
The future

DURATION: 1 hour

(Download this seminar description in PDF format for printing from here)

(return to top)

11. Nullology (1 hour)

The Zen of Database?

ABSTRACT:

Nullology is the study of the empty set. Sets per se crop up all over the place in the relational world; the question is—and it’s a crucial one—what happens if the set under consideration happens to be empty? For example, a relation contains a set of tuples: What about the possibility of a relation containing no tuples at all? Now, this particular possibility is reasonably familiar; a relation with no tuples is much like an empty file. But even in this case, there are some interesting aspects to consider that are probably unfamiliar to most database professionals. And there are many other places in the relational world where the empty set rears its head, several of them both unfamiliar and of fundamental importance. Such matters are the subject of this presentation.

TOPIC OUTLINE:

Relations with no tuples
Relations with no attributes
Empty lists of operands
Empty heading partitions
Empty determinants and empty keys
Empty types (domains)

DURATION: One hour (approx.).

PREREQUISITES: Attendees will be expected to have a broad familiarity with the relational model.

(Download this seminar description in PDF format for printing from here)

(return to top)

12. Dropping Acid (1 - 1.5 hours)

a new look at databases and transactions

ABSTRACT:

Here are some questions that any database professional should be able to answer almost without thinking:

1. What exactly is a database?
2. What is the most important property a database should possess?
3. When should integrity constraints be checked?
4. What are the ACID properties of a transaction?

It is the speaker's contention, however, that there's more to these questions - or less, depending on your point of view! - than meets the eye. In fact, empirical evidence, gathered over many years, strongly suggests that a majority of database professionals are not able to give good answers to Questions 1, 2, or 3, even though most will happily give quite a lengthy response to Question 4. It's therefore interesting to observe that finding the "right" answer to Question 4 relies on finding the "right" answers to Questions 1, 2, and 3. What's more, the "right" answer to Question 4 (in the speaker's opinion) is not the one usually given; in fact, that "right" answer is somewhat at odds with conventional wisdom.

The purpose of this presentation is to explore the foregoing issues in some detail. It offers possibly novel answers to Questions 1, 2, and 3. It then goes on to examine the ACID properties (Question 4) in the light of those answers, and concludes that, while ACID might be a nice acronym, the concepts it represents don't really stand up to close examination. To be more specific:

The "A" property (atomicity) might be pragmatically useful, but it's not logically required.

If the "C" property is consistency, it's trivial; if it's correctness, it's unenforceable. In fact, the "C" property isn't really a property at all but merely a desideratum.

The "I" property (isolation) is likewise merely a desideratum; it's unenforceable, even if all transactions execute at the maximum level of isolation.

The "D" property (durability) isn't an ironclad guarantee, either.

These conclusions are clearly somewhat counter to orthodox opinion! Please note clearly, therefore, that they're not meant as an attack on the vast amount of excellent research that has been done on transaction management over the past 30 years or so, nor on the many elegant and useful results that have been obtained from that research. Rather, they're offered in an attempt to improve our understanding of some of the issues that lie at the very foundations of our field.

DURATION: 1-1.5 hours.

PREREQUISITES: Attendees will be expected to be professionally interested in database technology.

(Download this seminar description in PDF format for printing from here)

(return to top)

13. The Problem of Missing Information (1 - 1.5 hours)

a technical presentation

ABSTRACT:

The speaker will describe the three-valued logic approach to missing information ("nulls") in depth. Both the underlying theory and SQL's attempt to implement that theory will be described. In particular, the speaker will show why the approach is a disastrously bad one and will sketch an alternative and logically sounder approach.

TOPIC OUTLINE:

Types of null
UNK, unk, and three-valued logic (3VL)
Consequences of the 3VL approach
SQL's implementation of 3VL
Outer join
Further thoughts
So what can we do?

DURATION: One to one and a half hours.

PREREQUISITES: Attendees will be expected to have at least a basic familiarity with the ideas of the relational model and/or SQL.

(Download this seminar description in PDF format for printing from here)

(return to top)

14. Foundation Matters (1 - 3 hours)

a keynote presentation originally prepared for VLDB'02 Hong Kong, August 2002

ABSTRACT:

This session is meant as a wake-up call ... The foundation of the database field is, of course, the relational model. Sad to say, however, there are some in the database community, and certainly in the industry, and to some extent in academia also, who don't seem to be as familiar with that model as they ought to be; there are others who seem to think it isn't very interesting or relevant to the day-to-day business of earning a living; and there are still others who seem to think all of the foundation-level problems have been solved. Indeed, there seems to be a widespread feeling that "the world has moved on," so to speak, and the relational model as such is somehow passé. In my opinion, nothing could be further from the truth! In this session, I want to sketch the results of some of my own investigations into database foundations over the past twenty years or so; my aim is to convey some of the excitement and abiding interest that is still to be found in those investigations, with a view - I hope - to inspiring others in the field to become involved in such activities.

Acknowledgments: Most of the work I'll be reporting on was done in conjunction with my friend and colleague Hugh Darwen of IBM in the UK. Other collaborators include David McGoveran of Alternative Technologies in California and Nikos Lorentzos of the Agricultural University in Athens, Greece.

First of all, almost all of the ideas I'll be covering either are part of, or else build on top of, The Third Manifesto [1]. The Third Manifesto is a detailed proposal for the future direction of data and DBMSs. Like Codd's original papers on the relational model, it can be seen as an abstract blueprint for the design of a DBMS and the language interface to such a DBMS. Among many other things:

It shows that the relational model universal servers - and I do mean the relational model, not SQL ─ is a necessary and sufficient foundation on which to build "object/relational" DBMSs (sometimes called universal servers).

It also points out certain blunders that can unfortunately be observed in some of today's products (not to mention the SQL:1999 standard).

And it explores in depth the idea that a relational database, along with the relational operators, is really a logical system and shows how that idea leads to a solution to the view updating problem, among other things.

Note: The foregoing interpretation - i.e., of what a database really is - is directly relevant to the process of logical database design (and I will mention some recent results in this connection). It is also directly relevant to what the commercial world calls business rules [2].

Reference [1] also complements the relational model by introducing a detailed proposal for a theory of types. In particular, that theory includes a novel approach to the vexing issue of type inheritance, an approach in which the answer to the famous (or infamous) question "Is a circle an ellipse?" is - pace much of the object literature on the subject - a resounding yes. In fact, I'll explain why I believe objects and a "good" approach to type inheritance are fundamentally incompatible.

More recently, Hugh Darwen and I, along with Nikos Lorentzos, have been building on Lorentzos's original work and the ideas presented in reference [1] - including the type inheritance ideas - to investigate the question of support for temporal data [3]. Again, it is our belief that the relational model is a necessary and sufficient foundation on which to build such support. It is true that we have defined a large number of new relational operators (with a view to raising the level of abstraction and simplifying implementation), but all of those operators are, in the final analysis, nothing but shorthand. We have also, among other things, defined a new ("sixth") normal form and proposed a temporal database design methodology.

REFERENCES:

1. C. J. Date and Hugh Darwen: Foundation for Future Database Systems: The Third Manifesto, 2nd edition (Addison-Wesley, 2000). A detailed study of the impact of type theory on the relational model of data, including a comprehensive model of type inheritance.

2. C. J. Date: WHAT Not HOW: The Business Rules Approach to Application Development (Addison-Wesley, 2000).

3. C. J. Date, Hugh Darwen, and Nikos A. Lorentzos: Temporal Data and the Relational Model (Morgan Kaufmann, to appear 2003). A detailed investigation into the application of interval and relation theory to the problem of temporal database management.

DURATION: 1- 1.5 hours.

PREREQUISITES: Attendees will be expected to be professionally interested in database technology.

(Download this seminar description in PDF format for printing from here)

(return to top)

16. The Closed World Assumption (1 - 1.5 hours)

a technical presentation

ABSTRACT:

The Closed World Assumption (CWA) is an important concept in the database world, despite the fact that it isn't usually spelled out explicitly. Basically what it says is this: Everything stated by the database, either explicitly or implicitly, is true; everything else is false. This presentation explains the CWA in detail and shows why it is preferred over its rival, the Open World Assumption (OWA). In particular, it examines the claims that are sometimes heard to the effect that the database community operates under the CWA while ther semantic web community operates under the OWA. It also explains how “missing information” can be handled without any need for nulls or three-valued logic.

TOPICS:

Background review: what a database really is (?)
Relation values vs. relation variables
Relvar predicates and constraints
The CWA and OWA defined
Why the CWA is preferred
Relvar predicates revisitied
Unknown and uncertain information
Negation and disjunction

DURATION: 1 - 1.5 hours

PREREQUISITES: Attendees will be expected to be familiar with basic relational database concepts. (However, the presentation will begin by briefly reviewing such matters.)

(Download this seminar description in PDF format for printing from here)

(return to top)

17. Don't Mix Pointers and Relations!

An Attack on (among other things) the SQL Standard

ABSTRACT:

The idea that databases should be allowed to include pointers to data as well as data per se has been around for a long time. Certainly it was a sine qua non in the old pre-relational (IMS and CODASYL) world, and in the shape of "object IDs" it pervades the object world as well. Despite the fact that Codd very deliberately excluded pointers from the relational model when he first defined it, the same idea also rears its head from time to time in the relational world (the most recent manifestation of the idea in a "relational" context occurring, not surprisingly, in connection with extensons to the SQL standard to include support for objects). This presentation argues strongly that mixing pointers and relations in this way is a bad idea.

TOPIC OUTLINE:

Introduction
Values and variables
Relation values and variables
Referencing and dereferencing
Mixing pointers and relations: the basic idea
Mixing pointers and relations: a closer look
Can the idea be rescued?
Further criticisms
Concluding remarks

DURATION: One hour (approx.)

PREREQUISITES: Attendees will be expected to have a basic familiarity with either the relational model or SQL

(Download this seminar description in PDF format for printing from here)

(return to top)

18. Rethinking Foreign Keys (1 - 1.5 hours)

a technical presentation

ABSTRACT:

Foreign keys as classically understood are much too limited - but they can be generalised, cleanly, to become really useful! This session takes this no doubt rather heretical position as its jumping-off point. However, it quickly backs up its claims with several realistic examples - examples that don't just demonstrate serious defects in foreign keys as historically defined, but also suggest a series of compatible extensions to that historical definition that have the effect of extending, dramatically, the range of applicability and usefulness of the foreign key concept. Overall, the session makes a strong case for including the proposed functionality within existing database products.

TOPICS:

Foreign keys, referential integrity and related terminology
Inclusion dependencies
What do key constraints really mean?
Generalising the target
Generalising the source
More than one target
Compensatory actions
Implementing correctness
Recommendations
Appendix: A little history

DURATION: 1 - 1.5 hours (approx)

PREREQUISITES: Attendees will be expected to be professionally interested in database technology.

(Download this seminar description in PDF format for printing from here)

(return to top)

19. SQL and Relational Theory: How to Write Accurate SQL Code

Revised Version

ABOUT THIS SEMINAR

This seminar shows you how to write SQL code that's logically correct; how to avoid various SQL traps and pitfalls; and, more generally, how to use SQL as if it were a true relational language.

SQL is ubiquitous. But SQL is complicated, difficult, and error prone (much more so than SQL advocates would have you believe), and testing can never be exhaustive. So to have any hope of writing correct SQL, you must follow some discipline. What discipline? Answer: The discipline of using SQL relationally. But what does this mean? Isn't SQL relational anyway?

Well, of course SQL is the standard language for use with relational databases──but that doesn't make it relational! The sad truth is, SQL departs from relational theory in all too many ways; duplicate rows and nulls provide two obvious examples, but they're not the only ones. Thus, systems based on SQL give you rope to hang yourself, as it were. So if you don't want to hang yourself, you need to understand relational theory (what it is and why); you need to know about SQL's departures from that theory; and you need to know how to avoid the problems they can cause. In a word, you need to use SQL relationally. Then you can behave as if SQL truly were relational, and you can enjoy the benefits of working with what is, in effect, a truly relational system.

Of course, a seminar like this wouldn't be needed if everyone already used SQL relationally──but they don't. On the contrary, there's a huge amount of bad practice to be observed in current SQL usage. Such practice is even recommended in textbooks and other publications, by writers who really ought to know better; in fact, a review of the literature in this regard is a pretty dispiriting exercise. The relational model first saw the light of day in 1969──yet here we are, almost 40 years later, and it still doesn't seem to be very well understood by the database community at large. Partly for such reasons, this seminar uses the relational model itself as an organizing principle; it discusses various features of the model in depth, and shows in every case how best to use SQL to implement the feature in question. Note: Classroom exercises are an integral part of the seminar, and attendee discussion and interaction are encouraged.

ABOUT THIS REVISED VERSION

This seminar is a revised version of an earlier seminar with the title How to Write Correct SQL and Know It: A Relational Approach to SQL. The overall objective remains the same, of course─using SQL relationally is still the emphasis─but the seminar has been revised throughout to reflect, among other things, experience gained from teaching the earlier version. New material has been added (including, importantly, a discussion of how to deal with missing information without using nulls); examples have been expanded and improved; and the treatment of SQL has been upgraded to cover recent changes to the SQL standard.

A more detailed breakdown of this seminar can be found here

(return to top)

20. View Updating - How to Make It Work

a technical presentation

ABSTRACT:

Ever since the relational view concept was first invented, view updating has been a contentious issue. Support in today's SQL products is ad hoc and meager at best. The SQL standard is even more impenetrable in this area than it usually is. Even the research literature is weak on this topic; numerous approaches have been proposed and found wanting over the years. In this presentation, by contrast, a way is described of looking at the problem that (a) appears to be logically correct, (b) is certainly not ad hoc, and (c) works for all kinds of views. The overall message is: Views in general are just as updatable as "base tables" are! Attend this presentation and see why this isn't as extravagant a claim as it might seem.

Note: This presentation is essentially a brief technical introduction to the material covered in much greater depth in #25 ('The View Updating Problem and a Proposed Solution'). However, it is self contained and does stand alone as a presentation in its own right.

DURATION: 1 to 1.5 hours..

PREREQUISITES: Attendees will be expected to be professionally interested in the theory and practice of database management.

(Download this seminar description in PDF format for printing from here)

(return to top)

21. What is a Relational Database? Shedding Some Light

a technical presentation

ABSTRACT:

Over 40 years after the relational model was first invented, there's still a widespread lack of understanding of what a relational DBMS really is, or should be. Of course, SQL is a serious impediment here; SQL departs in so many ways from the relational model as such that it's a considerable stretch to think of SQL DBMSs as relational at all. This presentation explains what it means for a DBMS to be truly relational. It also offers a careful definition of the relational model and explains why (in the speaker's opinion) databases really "must" be relational. As an appendix (if time permits), it summarizes and illustrates some of the major differences between the relational model and SQL.

TOPIC OUTLINE:

"Data looks relational"

"Relational operators available"

Why databases must be relational

The relational model vs. others

The relational model defined

Appendix: SQL departures from the relational model

DURATION: l'/2 hours.

PREREQUISITES: Attendees will be expected to be professionally interested in database technology

(Download this seminar description in PDF format for printing from here)

(return to top)

22. Database Fundamentals: Sets, Logic, and Related Matters

a Chris Date Master Class

ABOUT THIS SEMINAR

This seminar consists of a detailed examination of certain foundational topics introduced, but not really covered in depth, in the companion seminar 'SQL and Relational Theory: How to Write Accurate SQL Code'. The topics in question are:

Set theory

Nulls, three-valued logic, and missing information

Nullology

Why pointers are considered harmful

How to make foreign keys really useful

Further details about this seminar are available here.

(return to top)

23. Normal Forms and All That Jazz: A Database Professional's Guide to the Theory of Database Design

a Chris Date Master Class

ABOUT THIS SEMINAR

How many of these questions can you answer?

1. What's the difference between 3NF and BCNF?
2. Is it true that if a table has just two columns, then it's in 4NF?
3. Is it true that if a table has just one key and just one other column, then it's in 5NF?
4. Is it true that if a table is in BCNF but not 5NF, then it must be all key?
5. Is it true that 5NF tables are redundancy free?
6. What precisely is denormalization?

As you can see, these questions all have to do with normalization and normal forms. Normal forms are important, of course, but there's much more to database design theory than just normal forms as such.

Further details about this seminar are available here:

(return to top)

24. Relational Database for Computer Professionals

a Chris Date Master Class

Chris Date is the world’s best known specialist in relational databases. In this seminar, he explains exactly what a relational database is, and describes the underlying theory of such databases (“the relational model”) in considerable depth. He also discusses the standard database language SQL.

ABOUT THIS SEMINAR

The relational model is one of the great technical inventions of the last hundred years. It’s the foundation of everything we do in the database field; in fact, the relational model is what turned database management into a truly scientific discipline instead of (as it was formerly) just an ad hoc collection of tricks and techniques. Thus, everyone technically involved in database management, even to the smallest degree, owes it to himself or herself to acquire a reasonable knowledge and understanding of the relational model─for without it, fully effective and productive job performance is impossible.

Further details about this seminar are available here:

(return to top)

25. View Updating and Relational Theory

solving the view update problem

ABOUT THIS SEMINAR

Views are “virtual tables.” That means they should be updatable, just as “real” or base tables are. In fact, view updatability isn’t just desirable, it’s crucial, for both practical and theoretical reasons. But view updating has always been a controversial topic; in fact, ever since the relational model first appeared, there has been widespread skepticism as to whether view updating is even possible, at least in the general case. And today’s SQL products certainly don’t help in this regard, since they don’t let most views be updated at all (what’s more, even when they do, they often don’t do it right).

In stark contrast to this conventional wisdom, this seminar shows how views can always be updated (just so long as those updates don’t violate any integrity constraints, of course). More generally, it shows how updating ought always to work, regardless of whether the target is a base table or a view. In other words, it describes a theory of updating—a theory that’s 100% consistent with the relational model but is, sadly, rather different from the way updating typically works in today’s SQL products.

Highlights of the seminar include:
1   The key concept of information equivalence
2   The role of good database design in avoiding ambiguity
3   The crucial role of constraints and compensatory actions
4   Why updates are always updates to the database as such
5   Why multiple assignment is so important

The overall message is: Views in general are just as updatable as base tables are. Attend this seminar and see why this isn’t as extravagant a claim as it might seem.

Further details about this seminar are available here:

(return to top)

26. An Introduction to Logic

a tutorial for database practitioners

ABSTRACT

Everyone knows that the relational model is founded on logic and set theory, and moreover that it derives much of its strength, rigor, and robustness from those solid foundations. Few database professionals can claim to be familiar with logic or set theory, however, even though an elementary knowledge of those disciplines is critical to successful use of a relational DBMS. This presentation is offered (along with its companion, “An Introduction to Set Theory”) as an attempt to rectify the situation. It explains certain key concepts from predicate logic and shows their direct relevance to a variety of database issues. A specific goal of the presentation is to enable attendees to use logic effectively in their day-to-day database activities. To this end, attendees will be asked to perform several pencil-and-paper exercises in class, and answers to those exercises will be the subject of group discussion.

TOPICS:

· The problem of ambiguity
· Propositions
· Connectives and truth tables
· Tautologies, contradictions, and inference
· Truth functional completeness
· Predicates and instantiation
· Quantification
· Free and bound variables
· Sorted vs. unsorted logic

DURATION: 2½-3 hours (approx.).

PREREQUISITES: Attendees will be expected to be broadly familiar with the relational model (or at least with SQL). No prior knowledge of logic will be assumed.

(Download this seminar description in PDF format for printing from here)

(return to top)

27. An Introduction to Set Theory

a tutorial for database practitioners

ABSTRACT

Everyone knows that the relational model is founded on logic and set theory, and moreover that it derives much of its strength, rigor, and robustness from those solid foundations. Few database professionals can claim to be familiar with logic or set theory, however, even though an elementary knowledge of those disciplines is critical to successful use of a relational DBMS. This presentation is offered (along with its companion, “An Introduction to Logic”) as an attempt to rectify the situation. It explains the basic concepts of set theory, showing both (a) their relationship to predicate logic and (b) their direct applicability to a variety of database issues. A specific goal of the presentation is to help attendees understand the relevance of set theory concepts to SQL query formulation and related matters. To this end, attendees will be asked to perform several pencil-and-paper exercises in class, and answers to those exercises will be the subject of group discussion.

TOPICS:

Sets and their properties

Set inclusion; power sets

Union and intersection; closure

Difference (absolute, relative, symmetric)

Commutativity, associativity, distributivity

Some important identities

The Principle of Duality

The algebra of sets

Ordered pairs and cartesian product

Relations and functions

DURATION: 1½-2 hours (approx).

PREREQUISITES: Attendees will be expected to be broadly familiar with the relational model (or at least with SQL). No prior knowledge of set theory will be assumed.

(Download this seminar description in PDF format for printing from here)

(return to top)

28. Why Three-Valued Logic Doesn’t Work

a technical presentation

ABSTRACT

This presentation is a kind of appendix to the presentation “An Introduction to Logic” (prior attendance at which is strongly advised). The logic on which the relational model is based is two-valued (2VL). In its attempt to deal with the so called “missing information” problem, however, SQL is based on a three-valued logic (3VL) instead. This presentation explains in detail why any such attempt is doomed to failure. More specifically, it shows why 3VL (a) doesn’t solve the problem, (b) isn’t useful, and (c) can actually be dangerous.

TOPICS:

What is 3VL?

Nulls and 3VL: the basic idea

Effect on the connectives

3VL gives wrong answers!

What about 4VL?

Many-valued logics in general

So what do we do?

DURATION: 1 hour.

PREREQUISITES: Attendees will be expected to be broadly familiar with the relational model (or at least with SQL).

(Download this seminar description in PDF format for printing from here)

(return to top)

Peter Robson,
Relational Database Events
Edinburgh
Scotland.
(first published January 2007 and last updated November 2012)