Clarify suppress redundant in aggregated mode

[infn-ke]

Can you add some clarification how suppress_redundant shall work in conjunction with aggregation (e.g. allow_aggregation)? Section 3.5.1.5.2 describes the principles of the suppression, but is this valid only when updates are sent for a single leaf?

Assume a subscription to container node /a/b[name=b1]/c (example in section 2.3.1) with leafs d and e, what is the correct behavior for suppress_redundant?

• Leafs d and e are suppressed ---> container c is also suppressed? No data sent?
• Leaf d suppressed ---> An update sent with path elements for container c, but json_val contains only leaf e?