> I despise writing a bunch of dao/repo code, so I've resorted to codegen-ing my daos/repos.

Spring Data does this very well, in a fashion partly inspired by ActiveRecord. You define a Repository interface with whatever search methods suit your case; a concrete class is built and injected for you. The same interface can back onto JPA databases, Mongo, Redis and I forget what else.

If you add Spring Data REST, you get a complete RESTful API -- complete with HATEOAS, paging, sorting, searching and HAL -- automagically.

Use Spring Boot and you won't need to do any more than include the dependencies in your Maven POM or build.gradle file.

I also highly recommend Aurelia. It's a fantastic and future proof front-end library.

Out of curiosity, why Spark over Scalatra? (http://scalatra.org/)

What makes you say it is future proof?

I assume he's been to the future, that's the only way one could know.

By future-proof I didn't actually mean forever. It heavily makes use of es6/es7, so it was designed with "forward-thinking" design choices (forward thinking is what I intended when I wrote future proof). Right now it relies on babel, but as es6/es7 gains adoption it'll grow out of requiring it. Similar to Angular 2.0, which is also being developed with a forward thinking mindset.

I used Scalatra on one other project, and I found it to be difficult to deploy. Whether it's true or not, my impression is that Spark is a little simpler to use and deploy than Scalatra. Scalatra is my second choice though.