Various suggestions from Alex

[joepio]

@AlexMikhalev has a big ontology built that he updates and wants to run tests on before deploying to production. If he'd update the ontology directly, it would also change production stuff, so that's not good.

Exporting JSON without stopping server

Right now, his aim is to run atomic-server export but that requires a lock, which means his production server has to go down, which leads to unhappy customer. The issue becomes: allow Atomic-Server to export all JSON without stopping. Couple of ways to do this:

Stream all changes + new resources (drive level)

E.g. make the websockets API more powerful. Allow the SUBSCRIBE to have two options: 1) include full resources (instead of just commits) 2) include all children.

This would mean we'd need to either add a new function SYNC or something, or we'd allow WS messages to have arguments (`SUBSCRIBE (children,resources)

Get resource + all nested children recursively

We add /endpoints/getTree (or something) which returns a JSON-AD containing all children.