What is NULL

Indicates "value does not exist" or "value is unknown."

Cannot be compared like a value because it is not a value.

It is not an empty set (a set with 0 elements) but is considered a non-existent set.

NULL Contrary to Relational Model

The relational model is based on the closed world assumption (assuming that anything not proven true is false), and NULL contradicts this.

The relational model is based on binary logic, so three-valued logic, which includes something other than true and false, is unacceptable.

NULL remains NULL even after operations or string manipulations
Possibility of unintended SELECT results
- Queries change depending on how NULL is interpreted
Introduces three-valued logic (TRUE, FALSE, Unknown)
- SQL becomes more complex
Adverse effects on the optimizer
- The presence of NULL affects the optimizer's calculations
  - Negatively impacts rewriting queries for optimal performance and estimating query costs

Table Normalization
- By advancing normalization, columns can be set to NOT NULL
  - Avoid using default values that are synonymous with NULL while defining NOT NULL
Use of COALESCE function
- A function that allows setting a default value when a specified column is NULL
  - IFNULL also exists, but IFNULL is not SQL standard
- Can be effectively used in patterns where SQL evaluation results in NULL (e.g., results of aggregate functions like SUM or AVG, results of OUTER JOIN, evaluation results of NULLIF, etc.)