Currently if a function depends on configured values, we need to pass that data to that function from the initial graph construction. This often touches many layers of code and causes large unit test churn as we update parameters throughout.
We should look to see if there are better ways to handle dependency injection in clojure or re-structure the code to reduce the amount of this passthrough.
Currently if a function depends on configured values, we need to pass that data to that function from the initial graph construction. This often touches many layers of code and causes large unit test churn as we update parameters throughout.
We should look to see if there are better ways to handle dependency injection in clojure or re-structure the code to reduce the amount of this passthrough.