Learn the fundamentals of blockchain by building it from scratch. In Rust.
Some sections are less important than others and may be skipped if you are in a hurry.
Less important sections are marked with a *.
We formalize the notion of a state machine and implement several examples. We do not yet discuss the blockchain data structure or consensus. Examples range from simple toys for learning purposes, to realistic multi-user state machines common in real-world blockchain systems.
- Part 1 - Switch-based state machines - Two dead simple state machines to learn the basics.
- Part 2 - Laundry Machine - A toy state machine modeling the lifecycle of clean and dirty laundry.
- Part 3* - Automated Teller Machine - A semi-realistic, but significantly simplified state machine modelling a common ATM.
- Part 4* - Accounted Currency - A realistic state machine used as the foundation for many cryptocurrencies such as Ethereum and Polkadot.
- Part 5 - Digital Cash - A realistic state machine used as the foundation for many cryptocurrencies such as Monero, Dogecoin, and Litecoin.
We introduce the blockchain data structure and scaffold it from a simple hash-linked list to a proper blockchain with the Body distinct from the Header, and a consensus digest included. This is the most important chapter of the book.
- Part 1 - Header Chain - A minimal hash-linked list with no real state or execution logic.
- Part 2 - Extrinsics and State - We extend our chain to track state and introduce a simple notion of extrinsics.
- Part 3 - Consensus - We introduce a basic notion of consensus using proof of work as our first example.
- Part 4 - Batched Extrinsics - We separate the block body out of our header, and show that there are multiple extrinsics in a single block
- Part 5 - Fork Choice - We introduce the notion of a fork choice rule and the idea that consumers of the blockchain data structure must decide which of multiple chains is real for them.
- Part 6 - Rich State - We show that in real-world blockchains the state is not stored directly in the blocks and must be tracked separately. We also introduce the concept of genesis state.
We formalize the notion of consensus see how our previous look at Proof of Work fits into this framework and explore several other consensus schemes including Proof of Authority, a brief look at Proof of Stake, and some higher-order consensus concepts.
- Part 1 - Proof of Work - We re-implement Proof of Work in our new consensus framework
- Part 2* - Dictator - A toy identity-based consensus system where a single authority, the dictator, says what blocks are valid
- Part 3 - Proof of Authority - We implement several identity-based consensus systems, some of them realistic, others just toys. We briefly discuss Proof of Stake
- Part 4* - Even Only - We explore the notion of "arbitrary" consensus rules more formally.
- Part 5* - Interleave - This section is still under development. We will explore how to interleave different consensus rules on a block-by-block basis.
- Part 6 - Forking - We explore how to coordinate consensus handoffs so that consensus rules can change as the result of a fork part way through a blockchain's history.
This chapter is still under development. We begin by extending our blockchain data structure from chapter 2 to be fully generic over both the state machine (using the framework from Chapter 1) and the consensus engine (using the framework from chapter 3). We then continue on to develop a proper blockchain client which is able to import and export blocks, create blocks, manage a transaction pool, and decide on which fork is best. We may even introduce a notion of finality eventually.
Licensed under the terms of the GPL-3 or later.