December 20, 2023

By CjS77

What's up? Docs!

The Tari core developers have been fighting the good fight on two fronts: wrapping up changes to the Minotari codebase based on the audit feedback (finished!), and driving towards a Tari network testnet release candidate (soon?).

As usual, the major casualty of an extreme code offensive has been documentation. Today we’re happy to announce that we’re setting this straight– well, straighter, by releasing several new RFC documents that describe the Tari Digital Assets Network.

Firstly, RFC-303 has been completely rewritten and updated. It describes the DAN architecture and how it fits into the Tari ecosystem.

Then, RFC-350 describes the Tari Virtual Machine (TVM) and how it executes Tari smart contracts.

Then, we dive deeper into the Consensus layer of the DAN with a series of 5 documents, 4 of which are brand new:

  • RFC-305 presents the design philosophy and defines the roles of the consensus engine.
  • Many of you will never have heard of Cerberus, yet it sits at the very core of the Tari network. RFC-330 attempts to explain the fairly complex algorithm behind this very novel consensus algorithm in plain English.
  • A core part of what makes Tari unique is the dual-layer design to leverage the security and liveness guarantees of a proof-of-work blockchain, with the scalability and throughput of the Cerberus BFT consensus algorithm. Getting these two layers to play nice so that we can get the best of both worlds is a tricky challenge. RFC-325 covers a proposal for epoch management, our favoured idea for aligning the two layers.
  • State management is a huge challenge in massively scalable smart contract platforms. RFC-331 describes Tari’s indexing strategy for retrieving contract state from shards reliably and quickly.

We also have some RFC proposals that have entered the implementation phase:

  • RFC-141 describes Sparse Merkle Trees as a replacement for the somewhat hacky Mutable Merkle Mountain range data structure used to track the UTXO set on the base layer. SMTs have already been implemented and are currently being tested in nextnet.
  • This is a biggy. RFC-385 outlines a design for a suite of privacy-enabled stablecoins on Tari. Self-sovereign money is one thing, but self-sovereign private money takes us to the next level. This is still a very early design, but the core devs are already working on an early implementation of the design to be ready for the DAN testnet release next year.


Don’t feel like wading through over 20,000 words of new documentation? No worries, mate.

We’ve printed it all out on environmentally friendly, 100% recycled virtual paper and dumped it in Tari Bot’s lap, told it to memorise it all, cancelled its Christmas leave and made it work 24-hour shifts answering all the questions the community has about the DAN. You’re very welcome, and boas festas!