What Happened
Bitcoin just went through its first genuine chain split in years, and the market barely registered it. BIP-110 (Bitcoin Improvement Proposal 110), formally the Reduced Data Temporary Softfork, reached its mandatory signaling window at block 961,632 on August 8. The proposal would have imposed 1 year of consensus-level caps on arbitrary data in Bitcoin transactions: most new outputs limited to 34 bytes, OP_RETURN (the transaction field built for small data payloads) capped at 83 bytes, and individual data pushes capped at 256 bytes. Supporters pitched it as a temporary spam filter for a chain increasingly carrying inscriptions and token data. Activation required miners to signal agreement, and they did not: of the 2,016 blocks preceding the deadline, 51 carried the signal, roughly 2.53 percent against the 55 percent early-activation threshold the proposal set for itself.
When the first non-signaling block arrived after the deadline, nodes enforcing BIP-110 rules (mostly Bitcoin Knots instances configured for enforcement) rejected it and split onto their own chain. The minority chain produced exactly 2 blocks, mined by an anonymous group called Roughnecks through Ocean's decentralized mining protocol DATUM, then effectively stalled. Roughly 99.85 percent of hashratehashrateThe total computing power securing the Bitcoin network, in hashes per second — a measure of how much work backs the chain.See it on the site → stayed with the main chain, which kept producing blocks on its normal schedule. By Sunday the fork chain had fallen more than 80 blocks behind, and as of this writing it shows no sign of catching up.
The Math That Killed the Fork
The stall was predictable arithmetic, not bad luck. A forked chain inherits the main chain's difficultydifficultyA protocol setting that adjusts every ~2 weeks to keep blocks arriving about every 10 minutes, rising and falling with hashrate.See it on the site → target at the moment of the split but keeps only the hashpower that chooses to mine it. With well under 1 percent of Bitcoin's hashrate, the BIP-110 chain can expect block intervals measured in hours rather than minutes. Worse, Bitcoin's difficulty only adjusts every 2,016 blocks, and at the fork chain's observed pace analysts put that adjustment close to a year away, versus the main chain's routine 14-day cycle. A chain that cannot confirm transactions in reasonable time cannot host an economy, and a chain without an economy cannot attract the hashpower that would fix its confirmation times. The loop closes on itself.
The governance story is more interesting than the technical one. BIP-110 grew out of the long fight over non-payment data on Bitcoin: inscriptions, tokens, and embedded files that proliferated after Bitcoin Core loosened its OP_RETURN relay policy last year. Prominent voices including Strategy chairman Michael Saylor and Blockstream chief executive Adam Back publicly opposed the proposal, and miner signaling never got off the floor. The supporters' decision to enforce anyway, knowing the numbers, was less a bid to win than a demonstration that exit is possible. That deserves to be taken seriously rather than mocked: a network where a dissenting minority can fork off, fail on the merits, and impose no damage on everyone else is a network whose governance is working as designed. The counter-argument, that the episode proves miners rather than node operators hold the real veto, is worth sitting with, but the fork's failure had less to do with miner power than with the absence of any economic demand for the forked chain's coins.
No Replay Protection, On Purpose
The detail holders should actually care about is that BIP-110 shipped with no replay protection. Its authors treated the change as a tightening of Bitcoin's existing rules rather than the launch of a new network, so they never added the transaction-format split that Bitcoin Cash, for example, used in 2017 to keep its transactions from being rebroadcast on Bitcoin. The consequence: any ordinary signed Bitcoin transaction is valid on both chains. If the minority chain were alive, spending coins on one chain could move the same coins on the other, whether you intended it or not.
In practice the risk today is mostly theoretical, and the supporters' logic is internally consistent: a soft fork's ruleset is a subset of the existing one, so transactions are meant to be valid on both chains, and adding replay protection would have turned their tightening into exactly the new-coin launch they insisted they were not doing. But that logic is only comfortable while the minority chain stays dead. The documented pattern from earlier splits is that replay incidents and fork-themed scams arrive when a stalled chain finds a pulse and an exchange lists the ticker. Early signs indicate this chain will not get there. Signs are not guarantees.
What This Means for You
If you hold Bitcoin in self-custodyself-custodyHolding your own bitcoin keys rather than trusting an exchange or custodian. The maxim: not your keys, not your coins., your private keys now technically control coins on 2 chains, one of which is nearly motionless and has no meaningful market. The correct action is no action. There is nothing to claim, no snapshot to chase, and no wallet update required to stay on the main chain. The consistent lesson of prior fork events is that holders were rarely hurt by the fork itself; they were hurt by what they did about it.
Expect the scam wave anyway. Fork events reliably produce phishing campaigns offering to help you claim forked coins, and the mechanics are always the same: a site or tool that asks you to enter a seed phrase or connect a hardware wallet to a claiming service. Any such request is an attack on your key control, full stop. The pattern is a close cousin of the vendor-impersonation phishing OPNorange documented against Coldcard users last week: attackers exploit the moment when holders feel they must act quickly on news they only half understand.
Two postures deserve a check this week. If you run a node, know exactly what you are running: Bitcoin Knots users should confirm whether their version and configuration enforce BIP-110 and decide deliberately, because an enforcing node is currently following a chain more than 80 blocks behind the network and will show a false picture of balances and confirmations. And if your Bitcoin sits with an exchange or an ETF, this event cost you nothing, but note who made that true: your custodian decided which chain your claim tracks. That is the custodial risk trade-off working in your favor this time, and it is worth noticing that the choice was never yours to make.
What to Watch
Watch whether the fork chain produces another block at all, whether any exchange lists a BIP-110 ticker (the precondition for replay incidents and claim scams to become real), and whether Knots maintainers ship a release that drops enforcement by default. The larger fight is not over: the disagreement about what belongs in a Bitcoin block simply lost one activation vehicle, and the same energy is already flowing toward the covenants debate and the quantum-migration timeline. The next contested activation window will look a lot like this one, and the preparation is identical: know your keys, know your node, and do nothing fast.