PeerID Encoding/Decoding is inconsistent with specs

[aschmahmann]

The PeerID text spec https://github.com/libp2p/specs/blob/master/RFC/0001-text-peerid-cid.md, makes no reference to allowing us to decode identity multihashes as PeerIDs and yet we do here.

Similarly, the Encode(peerID) function in go-libp2p-core is also not spec compliant since it just encodes every peerID as base58 instead of returning it as a CID (e.g. identity multihashes are returned as base58 encoded)

Is the bug in the spec or code? Are we able to keep the spec as is at this point, or do we have to change it to accommodate people already using the non-spec compliant code?

^ @Stebalien @raulk @lidel

[Stebalien]

That was the original RFC. The spec has since been fixed: https://github.com/libp2p/specs/blob/master/peer-ids/peer-ids.md

[aschmahmann]

Ah ok. So that means we just have a bug in Encode then right?

[Stebalien]

No? Encode encodes in the legacy format for now.

[aschmahmann]

You're right it is consistent with the specs as it encodes in the legacy (base58) version. There's a separate issue to be raised here about whether we should upgrade to the newer version.

[aschmahmann]

Separate issue on upgrading is #976