Summary and Conclusion

Why and when you should use HRR in a live system

[Describe what use scenario HRRs handles best; outline the benefits:

type safety of queries defined in Haskell and type checking of queries against the DB schema; avoiding boilerplate]

Drawbacks you might be in for

[HRRs limitation to the base cases, as per decision of the authors; sometimes clumsy

syntax; need for a complete and running DB backend to build Haskell project; while HRR entry-level docs are quite good, lack of mid-level documentation and reference examples becomes obvious later on]

Integration

[To live up to the promise of type-safety, a change in the underlying DB schema

must be reflected when (re)building a Haskell project which uses HRR. Say something about how HRR can be integrated with a build system like stack so that data types get derived anew in their respective Haskell modules when necessary]