I'm thinking about scrapping the work I've done on this. Besides it being a bit antiquated, it really hasn't simplified the code much. The exception handling in Curator isn't that great yet and leaves me wanting. The recipes are nice, but the locks do not behave in a way that are interchangeable with what we're doing. I may contribute some to the project to bring features into it that would benefit us, but for now I'm skeptical about the gains from switching to Curator.
I was initially hoping with all of this we could abolish the ZooReader, ZooReaderWriter, and ZooCache, but ultimately they turned into management interfaces for Curator that, while less complex then the straight Zookeeper, were marginally less code. I was hoping for a big chunk of code being offloaded, but I wasn't quite seeing it. Unless anyone had another strong reason for this migration, I'm thinking we may want to push off for Curator to become a bit more fitting for our purposes.