Generic programming allows you to write type-independent code, which was introduced in Go 1.18 with type parameters, type sets, and type inference.
Type parameters, declared in square brackets, allow a single function to operate on multiple numeric types, eliminating code duplication.
Go provides built-in constraints like `constraints.Integer` and `constraints.Float` to easily define what types a generic function can accept.
Interfaces now define a set of types, not just methods, and the tilde (~) operator can be used to include a type's underlying types in a constraint.
Type constraints can be restricted by other constraints, and the `any` keyword serves as a convenient alias for an empty interface to simplify declarations.
The compiler can infer type arguments from function parameters, but they must be specified explicitly for types used only in return values.
A generic function can load YAML data into any struct, but using the `any` constraint sacrifices compile-time type safety.
A marker interface with a unique method can be used as a type constraint to ensure only intended structs are passed to a generic function.
Generics in Go 1.18 use type parameters and constraints to create flexible functions, supported by type inference to simplify calls.
Using generics without proper constraints can lead to a loss of type safety, but the benefit is often less and more readable code.
Jobs that call for the skills explored in this talk.