ValueType

A valuetype is a category of types in programming where instances store their data directly, rather than as references to separate heap-allocated objects. In languages that distinguish value types from reference types, operations on valuetypes typically copy the entire value, giving each variable its own independent data.

In the .NET ecosystem, value types are implemented as structs or enums and derive from System.ValueType. They are stored inline in their containing storage (on the stack or embedded within other objects) rather than on the managed heap. Assigning or passing a value type copies its data, implementing value semantics. If a value type needs to be treated as an object, it can be boxed into a reference type; unboxing retrieves the original value.

Common examples include primitive numeric types such as int and double, booleans, and user-defined types declared as structs or enums. Other languages adopt value-type semantics in various ways: Swift and Go treat structs as value types; C++ commonly uses value types for objects allocated on the stack or embedded in other objects; Java separates primitive types from objects, with primitives behaving as value-like types even though they are not classes.

Advantages of valuetypes include predictable memory usage and copying semantics, which can reduce the overhead of garbage collection and avoid shared mutable state. Disadvantages include the potential cost of copying large valuetypes and the need for boxing when passing to APIs that expect a reference type. Practical guidance is to use valuetypes for small, immutable, or insignificantly sized data, and to prefer reference types for large, polymorphic, or mutable data structures.

In metadata terms, the Common Language Infrastructure signals value types with the valuetype designation in type signatures, distinguishing them from classes that are reference types.