Model "address scopes" in signed peer records

[raulk]

There is such thing as “address scopes” for discovery mechanisms, which we have not modelled yet, e.g. a public discovery mechanism should only advertise public addresses, a multi-layer DHT with a private layer or mDNS should advertise private addresses.

When it comes to signed peer records, these two address sets should be contained in records with different asserted scope, and these records are additive to one another, rather than exclusive.

[Winterhuman]

@Stebalien Any update on this for use in ipfs/kubo#6932?

[Stebalien]

Unfortunately, not. There really hasn't been any progress on address records in general as it would require a DHT protocol upgrade to actually work.