Originally published at https://medium.com/mix-blockchain/introducing-link-blockchain-b447576ff438
Update: LINK is now called MIX.
LINK is an uncontrolled linked data ecosystem living on the blockchain. It is fully public. It cannot be censored and no one can be prevented from participating. It is fully programmable at every level. It is a development platform for interconnected distributed content apps that empower the individual.
LINK is currently running as a testnet. It has a block explorer and network stats.
The linked data is encoded with Google Protocol Buffers. It has an upgradable schema system which means that LINK’s content types are inheritable.
For example, a standard story type could be extended by a publication with information specific to its proprietary app. Regular apps would still be able to display the story.
Other potential content types are user profiles, Reddit posts, tweets, badges, media metadata and business metadata.
The encoded data is compressed with Google Brotli and stored in transaction logs using the BlobStore smart contract which I have developed.
Because LINK is an Ethereum blockchain it has the full power of Ethereum’s smart contract system. This allows for many useful smart contracts to be developed that can reference the content. For example, a key smart contract would be a feed contract that allows users to maintain a feed of content like RSS / Twitter / YouTube. Other important smart contracts would provide functionality such as content voting, tagging, badge awarding, etc.
A key feature of LINK is that it is uncontrolled. At a base level no one can be prevented from publishing. Certain filter bubbles will become popular and some users may not like the way they operate. However, this creates the incentive for more filter bubbles to be created. Because LINK is programmable at every level a new filter bubble can be created that is based on the same content as the other filter bubbles. This is in stark contrast to existing services such as Twitter, Medium, Reddit or YouTube where the content can only be accessed via the filter bubble provided by each siloed service.
An advantage to storing small pieces of content as “transaction attachments” using BlobStore instead of IPFS or Swarm is that there is incentivization for old content to be stored even after the creator is no longer prepared to pay for it. It would be really annoying if old content on Twitter or Reddit simply disappeared. Node operators are incentivized to store and serve old content because they will get paid by light clients. Potentially, old or rare content could have a higher price tag. Because publishers have to pay a fee that is in proportion to the size of the data, node operators cannot be spammed with content to store.
LINK apps using Ethereum’s light client functionality will be able to very efficiently query full nodes for multiple content items in a single round-trip.
Once the size of the content database increases, nodes may wish to reduce the amount of content they are storing. They can trim their transaction logs based on age, language, popularity, or content type, for example.
Larger items such as images or video are not well suited to storing in transaction logs. LINK will have media metadata content types that contain identifiers to the actual media in various fidelities and delivery systems.
Issuance
LINK has a built-in 2000 day bootstrap mechanism to fund the project. 5 years worth of mining is pre-allocated into a smart contract that will gradually release it to myself over the 2000 days at an ever slowing rate. This revenue will be sold as necessary to fund the project and (ultimately) to provide profit for myself. Thus I am financially incentivized to invest the LINK that is released in a way that will increase the value of LINK.
This “continuous crowdfund” is quite different to a lot of Ethereum startups that are seeking a large amount of capital upfront.