Once we have this and #1 we can start bootstrapping.

At its core, yaml.yaml just wants to be a functor that maps between the yaml and lambda categories.

With yaml we get a human friendly data serialization standard for all programming languages. More specifically I see:

• specification
• syntax
• (de)serializers
• performance
• community
• advanced editors
• online editors

On the other hand, the yaml spec is pretty dense and we don't want to deal with everything from the beginning.

The mapping to lambda is not just for the math, but we actually want to run software in a computer. In #2 I took a lambda calculus interpreter written in Rust and I was able to map simple documents to lambda terms.

A different approach could map between languages (transpiling). Haskell seems to be the closest but I'm not familiar with the syntax or tooling.

I can think of two approaches:

1. A transpiler would take YAML files as text input and output a code project (e.g an sbt project). Clever use of Rust macros could also fit this category.
2. An interpreter would read YAML files at runtime. This could be well suited for Scala + Cats