Multi-node sync [Experimental]

[rmehta]

To make this app multi-user we will follow a multi-node strategy. Each node will be a fully self-sufficient (offline-first) unit. Data will be migrated to and from "super node" real time. This will work similar to the data migration tool in frappe-py

• Allow any doctype to be synced
• Designate a node as a super node
• Establish connection to the super node (via WebRTC / WebSockets)
• Inbound sync with each table with the super node (super node to maintain a write log - similar to a binlog)
• Sync local documents with supernode
• Add global id in documents so that we know if they have been synced or not