A Registry Directory ~ based on BLOCKCHAIN that is ROOTless & NOT Centralized

From IIW

Registry also known as Directory based on blockchain that is rootless and not centralized

Tuesday 2C

Convener: Lionel Wolberger

Notes-taker(s): Lionel Wolberger and Dave Sanford

Tags for the session - technology discussed/ideas considered:

Discussion notes, key understandings, outstanding questions, observations, and, if appropriate to this discussion: action items, next steps:

The XDI Registry working group presented, in general terms, it's implantation of a registry on the bitcoin blockchain.


  • names v numbers: by focusing on numbers (UUIDS type 4) we avoid semantic, intellectual property, and other issues
  • discussed proofs of ownership and proofs of existence
  • full and partial blockchain
  • the role of blockstore

The XDI working group has put together a proposal that will support a decentralized and rootless registry – based on mapping a Universal Unique Identifier (UUID) to an endpoint. UUIDs are 128 bit values and in this case it is a Version 4 (random number). The endpoint might be identity data, personal data, etc.

Discovery of end-points is similar to the DNS system mapping a URI to an IP address. We want to be able to put this mapping function into a blockchain to get the shared, decentralized ledger (e.g. bitcoin blockchain) benefits.

Ethereum is also robust and has a big community. XDI.org has a Registry Working Group charter and proposal at this type.

Once you have the endpoint data (personal identity, business, etc.) and pointer mechanisms – what you get from blockchain combined with the XDI overlay:

1) Proof of ownership

2) Proof of ‘continued’ existence (e.g. vs. brand new account)

3) Escrow services – key recovery and/or replacement

The XDI group looked at Elinor Ostrom's 8 Principles for Managing A Commons as a basis for some mechanisms. XDI requires a look up service, but that requirement is generic. This proposal provides a specific proposal using the bitcoin blockchain to do that lookup.

Revocation exists, but does not erase a record of the past. First item in the ledger may say A = B, a later revocation entry in the ledger indicates that A no longer equals B – however a record of the period in which it was valid stays in the blockchain record.

One of the purposes is to enable decentralized economic incentive model – that will support the continued existence of the model and its goals – but it could use other methods of proof of existence.

Goals include:

  • Endpoint via IRIs
  • Proof of ownership, multiple types of proof
  • Owner digital signing of last change request
  • Proof of existence
  • Who are the trusted registrars (and their proofs)
  • Recovery – this is just secondary proof of ownership
  • Quality of registry response

At this point the XDI committee is trying to keep this discovery proposal not specific to identity discovery and leave identity attributes abstract – potential endpoints (e.g. blob, token, assertions). Mainly the intent is to enable XDI based graph discovery.

The bitcoin blockchain is immutable, but allows mutable transactions to be built on top of it – still allowing everything to be auditable, because of the layer below.