Generic programming

Generic programming is a paradigm in which algorithms and data structures are written in terms of abstract types that are specified only at the point of use, enabling a single implementation to operate correctly across many concrete types. The archetype is the generic list or map: written once, instantiated for integers, strings, user-defined objects, or any combination thereof, with the compiler ensuring type safety at each instantiation. Generic programming trades the duplication of type-specific implementations for the abstraction of type-parameterized ones.

The concept originated in C++ templates and the Standard Template Library (STL), where it was pushed to extremes that blurred the boundary between generic programming and compile-time computation — template metaprogramming. Later languages, including Java, C Sharp, and TypeScript, adopted more restrained forms of generics that sacrifice some expressive power for predictability and clearer error messages. The tension between these poles — C++'s unconstrained template explosion versus Java's erasure-based bounded generics — remains unresolved, with newer languages like Rust attempting synthesis through trait-based polymorphism.

At its best, generic programming is parametric polymorphism made practical: the promise that a function's behavior is uniform across all types, with no hidden type-specific special cases. At its worst, it is abstraction for abstraction's sake, producing error messages so long that programmers learn to ignore them. The discipline lies in knowing when a generic abstraction genuinely captures a pattern and when it merely obscures what the code actually does.