Open Data Layer

Objective Layer - Open Data Commons (Consensus + Storage) The Open Data Commons consists of two elements:

  • the consensus layer (blockchain)
  • storage layer (IPFS or other decentralised storage)

Consensus Layer

What is stored on the blockchain are only public keys of identities and hashes of content object that point to the decentralised storage layer on IPFS (or any other content addressed storage layer).

The consensus layer is based on the smart contract enabled blockchain that records what objectively happened and in what order. What is important to note is that this layer does not do any kinds of incentivisation. It just records that:

In block number X, identity Y, paid attention to content Z with metadata ABC

or

In block number X, identity Y, paid attention to identity Z with metadata ABC

Identity objects are entities that can interact with other entities (send and receive attention) i.e:

  • users
  • websites
  • bots, AIs etc

Content objects are entities that can only receive attention. i.e:

  • text
  • photos
  • videos
  • snippets of code
  • 3D objects
  • virtual reality worlds
  • spam filters
  • interfaces
  • trusted/spam account lists

Why do we present such a diverse but an incomplete list of possible content objects? I do it only to uncover our hidden assumptions based on the previous client-server social media models.

In our current models only text, photos and videos can be shared and interacted with (liked, retweeted, commented) on the social media platforms.

User interfaces, content ranking heuristics, spam filters are often fixed and centralised on the server side. Most people don’t even notice them - they are part of ‘The Application’.

In the proposed decentralised system they all become separate, shareable components of the application. So imagine Facebook where you can share your custom user interface on the timeline and other people can adopt it and effectively ‘fork’ facebook into a new application.

This creates two applications (two subjective views) that operate on the same open database but each of them presents a different view into it. We explore this topics further in the Subjective Layer chapter.

In terms of the blockchain technology, the best suited blockchain would be a smart contract enabled blockchain (Ethereum, Bitcoin sidechain) with: high transaction throughput per second fast block times cheap transaction fees permissionless access (everybody can write to it) native value token (this will be necessary for exchanges)

Outside of edge cases users never interact directly with this layer, it acts as a ledger that records all the actions that took place in the system.

There is no particular incentive model on this layer. It’s neutral, dumb and immutable. Anyone can create one or multiple identites and perform ‘pay attention’ transactions on other identites or content objects.

Storage Layer Storage layer is part of the stack where content is actually hosted. All the content addressed hashes from the Consensus Layer are pointing to the Storage Layer where the actual content is retrieved from.

The IPFS technology which enables content addressing seems to be the best choice as of today, however, there better alternatives could exist. The critical part of the storage technology would be the ‘content addressing’ aspect where each content unit is addresses by a unique hash that is recorded in the consensus layer (the blockchain).