Date on Database: Writings 20002006

PART 1 - Some Preliminaries CHAPTER 1 Edgar F. Codd: A Tribute and Personal Memoir Background Database Contributions Other Contributions Personal Memories Concluding Remarks CHAPTER 2 An Interview with Chris Date CHAPTER 3 Good Writing Does Matter Preamble Exhibit A Restriction Conditions in DB2 A Suggested Replacement for Exhibit A Exhibit B Exhibit C Concluding Remarks PART 2 - And Now for Something Completely Different CHAPTER 4 On the Notion of Logical Difference Preamble Why Is It Important to Think Precisely? Logical Differences Are Big Differences Some Things Are Much the Same Object Orientation The Unified Modeling Language Miscellaneous Examples Concluding Remarks Acknowledgments Appendix A: Some Quotes from the Literature CHAPTER 5 On the Logical Difference Between Model and Implementation Terms and Definitions What Are Models For? Definitions from the Literature Examples of Confusion: Performance Issues Examples of Confusion: Many Different Data Structures Examples of Confusion: Objects Examples of Confusion: Miscellaneous Consequences of Such Confusions Concluding Remarks Acknowledgments CHAPTER 6 On the Logical Differences Between Types, Values, and Variables Types Are Fundamental More on Types Values and Variables Values and Variables Can Be Arbitrarily Complex Variables Are Updatable, Values Aren't Pseudovariables Variables Have Addresses, Values Don't Relation Values and Variables Examples of Confusion: Values vs. Variables Logical Sameness A Hypothesis Examples of Confusion: Types vs. Values and/or Variables Consequences of Such Confusions Concluding Remarks PART 3 - Relational Database Management CHAPTER 7 Why We Need Type BOOLEAN CHAPTER 8 What First Normal Form Really Means Some Preliminaries Data Value Atomicity Relation-Valued Attributes Domains Can Contain Anything! Relation-Valued Attributes Make Outer Join Unnecessary Relation-Valued Attributes in Base Tables? - The Bad News Relation-Valued Attributes in Base Tables? - The Good News A Remark on "NF^2 Relations" A Definition Concluding Remarks Acknowledgments Appendix A: Terminology Appendix B: The Information Principle Appendix C: Definitions from the Literature Appendix D: So What About "Multi-Value Systems"? Appendix E: Formal Definitions CHAPTER 9 A Sweet Disorder SQL Tables The "Select *" Problem Duplicate Column Names Missing Column Names Implications for the UNION Operator Type vs. Degree Which Union? Relation Types Predicates Revisited Where Did Column Ordering Come From? Concluding Remarks Acknowledgments Appendix A: Row Comparisons CHAPTER 10 Double Trouble, Double Trouble The Cat Food Example Expression Transformation Rows Represent Propositions The Conclusion So Far Rats and Ratlets The Pennies Example Bag Theory The Performance Issue Some More SQL Problems Concluding Remarks Appendix A: A Tuple-Bag Algebra? Appendix B: More on Counting Pennies CHAPTER 11 Multiple Assignment A Little History Assigning to Several Variables at Once A Multiple Assignment Example Why We Need Multiple Assignment Why Integrity Checking Must Be Immediate Assigning to a Variable More Than Once Semantics Syntax Assigning to Several Components at Once Concluding Remarks Appendix A: Multiple Assignment in SQL Overview The SET Statement: Single Assignment The SET Statement: Multiple Assignment The SQL UPDATE Statement Revisited CHAPTER 12 Data Redundancy and Database Design The Running Example Some Prerequisites Database Design Is Predicate Design A Brief Review of Normalization Normalization Is Not Enough Orthogonality (I) Another Kind of Decomposition Tuples vs. Propositions Orthogonality (II) What About Views? Violating Orthogonality But Isn't It All Just Common Sense? Other Kinds of Redundancy Managing Redundancy 1. Raw Design Only 2. Declare the Constraint 3. Use a View 4. Use a Snapshot Concluding Remarks CHAPTER 13 Data Redundancy and Database Design: Further Thoughts Number One More on Predicates Identity Decompositions More on Propositions More on Normalization More on Dependencies More on Dependency Preservation Generalizing Key Constraints Orthogonality Revisited Concluding Remarks CHAPTER 14 Tree-Structured Data Definitions A Tree-Based Sort Technique How Treesort Works Data Structures Relational Operators Implementing TREESORT Integrity Constraints Traversing the Nodes of a Tree Bill-of-Materials Processing Another Kind of Tree What About Relation-Valued Attributes? Concluding Remarks Acknowledgments CHAPTER 15 Twelve Rules for Business Rules Assumptions A Note on Terminology Further Preliminary Remarks Prescription 1: Executable Rules Prescription 2: Declarative Rules Prescription 3: Kinds of Rules Prescription 4: Declaration Sequence vs. Execution Sequence Prescription 5: The Rule Engine Is a DBMS Prescription 6: The Rule Engine Is a Relational DBMS Prescription 7: Rule Management Prescription 8: Kinds of Constraints Prescription 9: Extensibility Prescription 10: Platform Independence Prescription 11: No Subversion Prescription 12: Full Automation Summary Acknowledgments References PART 4 - SQL Database Management CHAPTER 16 Two Remarks on SQL's UNION UNION in the Relational Model Data Type Conversion Duplicate Rows Concluding Remarks CHAPTER 17 A Cure for Madness The Syntax and Semantics of SELECT - FROM - WHERE Model vs. Implementation Nested Functions Further Thoughts A Final Observation CHAPTER 18 Why Three- and Four-Valued Logic Don't Work Some Basic Concepts Tautologies, Identities, and Contradictions De Morgan's Laws Codd's Three-Valued Logic The Unknown Truth Value and A-marks How Many 3VLs Are There? Codd's Four-Valued Logics 4VL Number 1 4VL Number 2 4VL Number 3 Some Questions of Intuition PART 5 - Further Relational Misconceptions CHAPTER 19 There's Only One Relational Model Celko's Preamble Chris Date= No Duplicates, No NULLs E. F. Codd, RM Version I E. F. Codd, RM Version II SQL-92= Duplicates, One NULL Duplicates, One NULL, Non-1NF Tables Rick Snodgrass= Temporal SQL How Many Relational Models? Concluding Remarks Technical Correspondence CHAPTER 20 The Relational Model Is Very Much Alive! Some Specific Issues Concluding Remarks Technical Correspondence CHAPTER 21 What Do You Mean, "Post-Relational"? What About Object Technology? Further Points Arising What About Object/Relational Technology? Technical Correspondence CHAPTER 22 A Database Disconnect The SMART Paper The Version 8 Paper Conclusion PART 6 - Subtyping and Inheritance CHAPTER 23 Is a Circle an Ellipse? Preamble A Little Background The Main Argument Software Engineering vs. Conceptual Modeling? What About "Colored Circles"? A Type Inheritance Proposal Technical Correspondence CHAPTER 24 What Does Substitutability Really Mean? What Is LSP? The Liskov Substitution Principle A Behavioral Notion of Subtyping The Liskov/Wing Paper: Assumptions The Liskov/Wing Paper: Objectives The Liskov/Wing Inheritance Model The Running Example Preconditions and Postconditions Covariance and Contravariance Extending the Example Extension and Constraint Subtypes Substitutability Defined PART 7 - Relational vs. Nonrelational Systems CHAPTER 25 Models, Models, Everywhere, Nor Any Time to Think Modeling in the UML What Does "Model" Mean? Miscellaneous Observations Concluding Remarks Postscript CHAPTER 26 Basic Concepts in UML: A Request for Clarification Some Immediate Discoveries The OCL Book The UML Book Types, Classes, and Domains Further Puzzles Concluding Remarks CHAPTER 27 A Comparison Between ODMG and The Third Manifesto Preamble Overview Objects Atomic Objects Structured Objects Collection Objects Literals Types, Classes, and Interfaces Inheritance Object Definition Language Object Interchange Format Object Query Language Summary RM Prescriptions RM Proscriptions OO Prescriptions OO Proscriptions RM Very Strong Suggestions OO Very Strong Suggestions IM Prescriptions References CHAPTER 28 An Overview and Analysis of Proposals Based on the TSQL2 Approach TSQL2 and the SQL Standard TSQL2 Tables The Central Idea Temporal Upward Compatibility Current, Sequenced, and Nonsequenced Operations Current Operations Sequenced Operations Nonsequenced Operations Accessing the Hidden Columns Data Definition Statements Valid-Time Tables Transaction-Time Tables Bitemporal Tables Statement Modifiers Are Flawed Consequences of Hidden Columns Lack of Generality Imprecise Specification Concluding Remarks References APPENDIX The Role of the Trade Press in Educating the Professional Community: A Case Study Overview The Saga Continues Conclusions Exhibit A Exhibit B Exhibit C Exhibit D Exhibit E Exhibit F Exhibit G Exhibit H Exhibit I Exhibit J Postscript INDEX