Ironically, Cabal’s “dependency hell” issue that this project is intended to solve prevents it from building from time to time, so it would be wise to reduce the scope and solve that problem first.

The plan is to have a few intermediate versions of the tool to bootstrap (or a few intermediate versions of the project model for that matter). The first feature to be implemented is dependency resolution supporting multiple versions; after that test support can be added, then other features.

This incremental approach would allow the project to move forward without having to implement a significant subset of functionality before any results can be demonstrated.