Roadmap

[jcrugzz]

1. Schema refactor
   • as this project evolved, the schema transitioned from consuming a json object and gave helpers for various conditions. This needs to be straightened out so that the operations associated with the various filtering are more easily mapped. Command pattern possibly.
   • normalize the lookup objects to be namespaced rather than specific camelCased named.
2. Engine agnostic
   • something to think about in this schema refactor is how to allow for different engines to be pluggable. The joi extension needs to expose a consistent interface. The data structure that the schema in datastar consumes from the joi extension can't be the same because of the nature of schemas.
   • another question: how to decouple the models themselves from statements. The engine itself needs to manage that aspect of the abstraction. We can only provide a thin layer around joi, a common set of types that get inherited.
   • Ultimately it seems like the engine needs to expose a schema implementation.
3. LevelDB engine.
   • use this as an exercise to see what elements can be re-used from the current implementation.