Overview:

polysemy is a library for writing high-power, low-boilerplate domain-specific languages (DSLs). It allows separation of business logic from implementation details, making implementation code reusable library code. It is similar to mtl, but offers better composition, requires less boilerplate, and avoids the O(n^2) instances problem. It is more powerful than freer-simple and has significantly less boilerplate than fused-effects. Unlike mtl, polysemy does not have functional dependencies, allowing multiple copies of the same effect to be used. This eliminates the need for hacks like classy lenses and theReaderT pattern, and solves the issue with typed errors. Type inference is well-supported in polysemy with the companion polysemy-plugin.

Features:

Higher-order effects: Allows easy writing of bracket and local as first-class effects.
Low-boilerplate effects: Creating new effects can be done in a few lines of code. Interpreters are simply functions and pattern matching.
Beginner-friendly tutorial by Raghu Kaippully.
Porting to Polysemy from transformers/MTL-style monads, explained in a post by Sandy Maguire.
Talk by Paweł Szulc on how to start thinking about polysemy.
Performance implementation talk by Sandy Maguire.
Additional blog posts by Sandy Maguire on other implementation details.

Summary:

polysemy is a powerful library for writing domain-specific languages with minimal boilerplate. It offers better composition and performance compared to similar libraries like mtl and freer-simple. The separation of business logic and implementation details allows for reusable code. The library comes with various tutorials and resources to help users get started, and its companion polysemy-plugin ensures good type inference. Overall, polysemy is a valuable tool for building efficient and maintainable DSLs.

Polysemy Research

Polysemy