The Simple Essence of Overloading - Making Ad-Hoc Polymorphism More Algebraic with Flow-Based Variational Type-Checking

Watch this 15-minute conference presentation from OOPSLA 2025 that introduces a novel approach to overload resolution in programming languages by combining variational type checking with algebraic subtyping. Learn how researchers Jiří Beneš and Jonathan Immanuel Brachthäuser from the University of Tübingen address the long-standing challenge of implementing type-directed overload resolution in Hindley-Milner-style type systems without relying on backtracking. Discover their flow-based variational framework that represents overloads as choices, effectively separating constraint collection, constraint solving, and overload resolution into distinct phases. Explore how this approach enables programmers to reuse the same name while offloading disambiguation to the type checker, potentially leading to more modular and efficient type checker implementations. Understand the theoretical foundations that could pave the way for improved error messages and serve as a simpler mental model for overload resolution. The presentation covers the essence of ad-hoc polymorphism and demonstrates how making it more algebraic can solve compatibility issues with traditional type inference systems.