Sjors Provoost’s technical book on problems and trade-offs in Bitcoin development. This overview separates a historical account from address specifications and the current state of implementations.
Sjors Provoost wrote Bitcoin: A Work in Progress. The copyright page identifies the first paperback edition as April 2022, published by Purple Dunes. A later source-file revision is not the first publication date. [1]
The official contents connect addresses, node synchronization, cryptography, network attacks and activation of changes. Topics include SegWit, Utreexo, Guix and Taproot. The book presents development problems, not a binding roadmap. [2]
In the public chapter, the author explains an address as a representation from which a wallet derives transaction-output conditions. The blockchain does not store addresses as bank accounts; it works with scripts and outputs. [3]
Address encoding helps detect transcription errors. A valid checksum does not establish the recipient’s identity. A correctly written address belonging to someone else still belongs to someone else: encoding and recipient verification are different tasks. [3]
We check the technical detail directly in BIP 350: witness version 0 uses Bech32; versions 1 through 16 use Bech32m. The new variant addresses an insertion weakness; it does not rename all older addresses. [4]
The Eclipse Attack chapter distinguishes rule validation from network access. An isolated node may receive only an attacker’s data while still rejecting invalid blocks. Validation alone does not guarantee a complete network view. [5]
The author shows why isolation alone cannot create valid confirmations for free: an alternative block must satisfy Proof of Work. After reconnection, a valid branch with more accumulated work may prevail, removing a payment visible in the earlier view. [5]
This overview uses the copyright page, contents and public chapters on addresses and Eclipse Attack, supplemented by the encoding specification. We do not claim to have read the whole book or listened to the podcast. Historical network parameters require further checking before being presented as current. [2]
For the clearest picture, read this entry together with Bitcoin Core, Taproot, Consensus Rules, Wallet.
01Does a valid address guarantee the right recipient?+
No. Encoding and checksums help detect transcription errors, not establish who owns an address.
02Can an isolated node accept any block?+
No. It continues checking the rules. An attacker can restrict its network view, but an alternative block must still meet validity requirements, including Proof of Work.