NFD

This requirement comes from #3027 note-5, and the backwards compatibility needs to stay for one release only.

There is a circular dependency:

• The backwards compatibility needs to stay for one release, which means a release must happen before #3081 changes can be reverted.
• This task is not implementable after #2107, because management dispatcher forbids overlapping handlers. This implies this task must be implemented before #2107 merges, aka " #3081 revert" blocks #2107.
• No release can happen before #3049 is merged (see #3026 note-14).
• #3049 is to be implemented together with #3029 (see #3049 note-4).
• #2107 blocks #3029.

Thus, " #3081 revert" blocks #2107, which blocks #3029+#3049, which blocks any release, which blocks " #3081 revert".

Even though Dispatcher doesn't allow overlapping handlers, it still can support this. You just need to have a single handler for more general prefix and perform additional logic inside this handler, which is trivial for status dataset without authorization.

You just need to have a single handler for more general prefix and perform additional logic inside this handler

Yes this will work temporarily (until a second dataset/command is added).
It's similar to face query operation.

I forgot this option.

Currently ForwarderStatus puts multiple elements directly under the Content of a Data packet.

In #2107, ForwarderStatus publisher is implemented as a StatusDataset, which causes a problem described in #2107 note-40.

The root cause is the lack of an outer element that wraps these elements, which conflicts with this definition in StatusDataset:

Status information of each entity is represented by a TLV block. Those TLV blocks are concatenated into a BLOB.

The temporary fix is commit:ecbf584299f4ae18d412c669874450b5405aaed4 : it parses ForwarderStatus and appends each element individually to the StatusDataset.
This solution can work, but it's ugly.

We should consider changing protocol to add an outer element that wraps those elements, to make implementation easier.

20151008 conference call rejects note-7. Alex thinks it's not a valid argument to change protocols for easier implementation.

Prefix changing will still proceed as planned.

In the future, we should consider changing ndn::mgmt::StatusDatasetContext to better support publishing this.