Forks & Chain Divergence

A fork is a protocol change: the existing network branches onto new rules.

Soft fork: the new validity rules are a subset of the existing ones. Nodes that do not implement them stay compatible with nodes that do, so the change is backward-compatible.

Hard fork: the new validity rules are incompatible with the existing ones, and nodes on each side recognize different chains. There is no backward compatibility, so as long as nodes and hash power supporting the old rules remain, the result is two chains (if the old side disappears entirely no split persists, but in practice a split is assumed).

Soft forks are asymmetric. Because the new rules must be a subset of the old ones, changes can only tighten, making previously valid things invalid. Loosening changes, such as raising the 21 million supply cap or the block limit, are necessarily hard forks. This asymmetry is the technical basis for saying Bitcoin's rules are hard to change.

Bitcoin conservatively prefers soft forks. SegWit (2017) and Taproot (2021) were both soft forks. Hard forks split chains easily, so they are avoided unless the whole community agrees.

A UASF (User Activated Soft Fork), as practiced during SegWit in 2017, activates on the will of node operators, meaning users, rather than on miner signaling.

The fiercest internal debate in Bitcoin's history. How to deal with the 1MB block size limit split the community in two.

"Big Blockers": simply raise the block size, in a hard fork. The main advocates were Roger Ver and Jihan Wu (Bitmain), and several expansion proposals appeared: BIP 100/101/102/109 (BIP 148, discussed below, was not an expansion proposal but a tool of the small-block side).

"Small Blockers": scale through SegWit, the Lightning Network, and similar work. This was the mainstream view among Core developers, and BIP 141 SegWit was the camp's central proposal.

The New York Agreement (May 2017) was a compromise signed by major companies and miners: "SegWit + 2MB hard fork." The 2MB portion was later abandoned, and only SegWit activated.

August 1, 2017: BIP 148 (UASF) came into force. BIP 91, which had locked in ahead of it on July 21, brought miner signaling into line, and SegWit locked in on August 9 (block 479,808) and activated on August 24 (block 481,824). On that same August 1, the big blockers split off Bitcoin Cash (BCH) in a hard fork.

Bitcoin's first major hard fork. It raised the block size to 8MB, later 32MB, and stressed the practicality of on-chain payments.

Roger Ver ("Bitcoin Jesus") marketed it aggressively, and for a while BCH drew real support in the US and Asia as "the true Bitcoin successor."

November 2018: BCH further split internally. Craig Wright (self-proclaimed Satoshi Nakamoto) and Calvin Ayre's Bitcoin SV (BSV) branched off from Bitcoin Cash ABC.

November 15, 2020: another chain branched off BCH, initially named Bitcoin Cash ABC (BCHA). It was renamed eCash (XEC) in July 2021, with a re-denomination splitting one token into 1,000,000 XEC.

As of August 2026, published market-cap trackers put BCH below 1% of Bitcoin's (figures vary by tracker). Bitcoin decisively won the "brand battle."

Craig Wright forked from BCH under the "Satoshi Vision" banner, claiming Satoshi's original vision, and removed the block size limit entirely.

Wright's persistent claim to be Satoshi Nakamoto fueled fierce controversy. In 2024 the UK High Court ruled that he is not Satoshi.

Major exchanges dropped it: in 2019 Binance, Kraken, ShapeShift, and others delisted BSV in response to Wright's legal threats against individuals.

2021: BSV suffered 51% attacks, with multiple block reorganizations that exposed its security weaknesses.

As of 2026 its market cap has shrunk further, and it is remembered as the archetypal "failed fork."

BIP 16 P2SH (2012): "Pay to Script Hash." It lets a script, a multisig for example, hide behind an ordinary address.

BIP 66 Strict DER (2015): a strict signature format, which tightened the cryptography.

BIP 68/112/113 CSV (2016): relative timelocks, activated July 4, 2016 (block 419,328). They are the basis for Lightning Network's time-locking.

BIP 141 SegWit (2017): Segregated Witness. It fixed transaction malleability and made the Lightning Network possible.

BIP 340/341/342 Taproot (2021): Schnorr signatures, MAST, and Tapscript. Privacy improvements, and the basis on which Ordinals was later built.

The Block Size War crystallized the fundamental question: "who governs Bitcoin?"

The candidates: developers (Core contributors), miners, exchanges, large holders, ordinary users running nodes, and corporations. Which of them actually decides?

What the war showed is that a majority of miners alone cannot change the rules. No matter how many miners agree, if full nodes reject their blocks, the result is merely a split-off chain.

But it does not follow that node operators alone decide. Nodes decide which blocks to accept; which chain carries economic value as "Bitcoin" is decided by exchange tickers, market price formation, and user choice, that is, by the economic majority. A node veto is a necessary condition, not a sufficient one.

This aligns with Bitcoin's core principles of "permissionless" and "censorship-resistant" operation.

The UASF (User Activated Soft Fork) idea is the practical form of "users are the network's guardians."

Soft fork candidates (all figures below are as of August 2026): OP_CAT has reached "Complete" specification status as BIP-347, but no mainnet activation parameters are set. OP_CHECKTEMPLATEVERIFY (CTV, BIP-119, Draft) entered a signaling window on March 30, 2026 via an independent activation client, yet miner signaling has stayed at 0% (the configured activation threshold is 90%). OP_CHECKSIGFROMSTACK (CSFS, BIP-348, Draft) is discussed as a companion to CTV. Inclusion in the BIPs repository is not a decision to activate.

Quantum migration: draft BIP-360 (Pay-to-Merkle-Root) proposes an output type that mitigates long public-key exposure, but it is not itself a post-quantum signature and does not solve short exposure after a spend reaches the mempool. Draft BIP-361 discusses migration and legacy signatures. Neither has an activation schedule or an adopted replacement signature.

Hard forks: since Bitcoin SV in 2018, no fork claiming to be "the true Bitcoin" has won market support. The market grew tired of forks, and after Bitcoin's victory the community pays new ones little attention.

"Bitcoin is Bitcoin": a simple point, engraved into the community after a decade of debate, and a demonstration of how strong network effects can be.