The 2.6% Fork That Could Move Your Bitcoin: BIP-110 and the Replay Attack Nobody Is Pricing In

Every hard fork is a referendum on memory. Not the memory stored in UTXOs—the memory of who we think we are. For two decades in cryptography, I have watched this industry swear "code is law," while the actual legal tender has always been narrative. And every few years, someone tries to cash that narrative in. The latest attempt is BIP-110, a Bitcoin protocol proposal to strip non-payment data—images, text, inscriptions—out of transactions. On paper, it is a technical argument over block space. In practice, it is a trap for anyone who gets curious about "free" tokens. Here is the number that should stop you cold: 2.6 percent. That is the current miner signaling support for BIP-110. It is nowhere near activation, and the proposal lacks remotely enough community consensus to force a network upgrade. And yet this marginal proposal could, under the right conditions, put your actual Bitcoin at risk through one of the oldest attack vectors in the book: replay. Not because the fork will win. Because a fork that is treated as a joke is exactly when people stop taking precautions. That, based on my audit experience, is when they get hurt. Context: The Block Space Wars, Revisited This is a war we have fought before. In 2017, the Block Size War split Bitcoin Cash away with roughly 15 to 30 percent miner support. Today's BIP-110 skirmish commands less than one-tenth of that. The technical intent is straightforward: require miners to reject blocks carrying non-payment data, starving the Ordinals and inscription ecosystem that has congested Bitcoin's block space since 2023. If enough miners signal support, nodes enforcing the new rule would begin rejecting non-compliant blocks at height 961,632. Kevin Loaec's warning about the proposal is not just a technical heads-up; it is a pre-enactment of a public safety event—an attempt to get the community to prepare before the window opens. But here is the Rubicon moment: BIP-110 is a hard fork, not a soft fork. The two resulting chains—the main chain and the purist chain—share every single historical transaction. Your balance exists identically on both. That sounds like a gift. It is actually the mechanism of injury. Early on, neither chain has replay protection. A transaction signed for one ledger is perfectly valid on the other, because the inputs are the same UTXOs, the keys are the same keys, and the math does not know which chain it occupies. I have seen this movie before. In 2017, at the peak of ICO mania, I audited more than fifty whitepapers and watched project after project promise "instant settlement" with zero concern for what happens when two ledgers disagree. I published a guide called "The Ethics of Empty Vests," arguing that technical substance is an ethical obligation, not a feature list. The replay attack teaches the same lesson: the gap between chains is not a bug report. It is where human money goes to disappear. Core: The Free-Money Trap Let me walk through the exploit, because the danger is counterintuitive. You hold Bitcoin. A fork happens. An exchange announces—as exchanges do, to retain customer loyalty—that it will credit the forked token to every BTC holder. Suddenly, there is value in your account you did not ask for. So you do the natural thing: you sell a little. You sign a transaction spending those forked coins. Here is the attack. That signature gets broadcast to the fork network, and a miner confirms it. But the same signature spends the same UTXOs on the main Bitcoin chain. An attacker—or an opportunistic miner—simply copies the transaction to the main network. Your forked-coin sale just became a real-Bitcoin sale. The buyer receives their tokens on the fork chain, and you lose real BTC on the main chain. This is not a hypothetical; it is a function of how ECDSA signatures and the UTXO model work. Code is law, but people are the soul—and the law here does not care that you "didn't mean" to send real Bitcoin. The standard industry advice is that this risk applies only to people who move coins during the fork window; if you hold and do nothing, you are safe. I have made that point myself, and it is mechanically true. But I have also spent enough time in DAO governance forums to know that "do nothing" is not what humans do when they see a free asset. The advice is technically sound and sociologically naive. The exposure curve follows curiosity, not greed. The moment an exchange credits the forked token, millions of people suddenly have a new balance to poke at, test, or "just try to sell." That is precisely the population that becomes replay victims. The token economics here form a negative-sum game, not a value-creating one. The fork coin's supply mirrors Bitcoin's—twenty-one million, no new issuance—but its demand is anchored to a chain with a whisper of hashrate. Low hashrate means unstable block times. Unstable blocks mean a miserable user experience. A miserable experience means thin liquidity. Thin liquidity means miners leave. That downward spiral is the destiny of any chain born at 2.6 percent support. The theoretical "1:1 airdrop" value is so close to zero that it cannot compensate for the operational risk. And yet the existence of that token—however worthless—is what motivates the dangerous transaction in the first place. Don't govern the exit, govern the entrance. BIP-110 obsesses over what data enters blocks; but the real governance failure sits at the entrance of user attention, where an unrequested token becomes a psychological nudge toward self-harm. Contrarian: The Fork Is the Distraction Here is the contrarian view, and it will annoy both sides: the fork's viability is irrelevant. Even if BIP-110 activated, the minority chain would face catastrophic difficulty adjustments, hours-long block production, and a hashrate so small that a 51 percent attack becomes trivially affordable. It would be a ghost chain, not a competitor. The maximalists behind it would claim moral victory; the Ordinals crowd would call it censorship; the market would move on with a price wobble of two or three percent. My read of developer chatter suggests the event is roughly halfway priced in—which is another way of saying the market, too, has decided this is noise. The real danger sits in the infrastructure layer, not the protocol layer. When exchanges decide whether to credit the forked token; when wallets decide whether to ship replay protection; when block explorers decide which chain is canonical—those off-chain decisions determine who loses money. History supports this. The 2016 Ethereum-Classic split produced initial chaos and then long-term amnesia. The 2017 Bitcoin Cash split burned more retail traders through rushed arbitrage than it enriched. In every case, the damage was mediated not by consensus rules but by the speed with which infrastructure deployed isolation tools. The good news: reputable exchanges and wallets usually ship replay protection and split scripts within hours to days. The risk window is not infinite. But hours is all an automated attacker needs. There is also a paradox worth naming. If a minority faction forces this chain into existence, BIP-110 may achieve the opposite of its stated goal. Ordinals will gain a martyr narrative, a sharper identity, and a chain history that now includes an attempted purge. The inscription ecosystem, which was losing the attention war, suddenly becomes the protagonist in a story about censorship resistance. A failed fork is often the best marketing a counterculture can receive. And as for the proposal's numbering—BIP-110 is itself contested, with BIP-420 more commonly associated with inscription debates in community lore. That confusion is fitting: a fork born from unclear governance carries unclear labels. Takeaway: What I'm Watching So what should a reasonable holder do? Three things. First, if you hold BTC, do nothing during the window around block 961,632. No transfers, no "let me test whether my forked coins are tradable," no clever arbitrage. The potential reward is a fraction of a token on a dying chain; the risk is your real Bitcoin. The asymmetry is absurdly in favor of inaction. Second, watch the exchanges. The moment a major exchange announces it will credit the forked token, the replay risk surface expands by an order of magnitude. That announcement is your cue to go quiet. Third, understand that the block space question is not going away. BIP-110 is a dress rehearsal for harder debates—about AI training data ownership, about decentralized identity, about who gets to decide what is "worthy" of a shared ledger. The chain may split, but the community doesn't have to. I have spent a career watching forks fail because their architects forgot that the chain is the infrastructure, not the institution. Code is law, but people are the soul. The question at block 961,632 is not which chain survives. It is whether we finally learn to govern the entrance before we are forced to exit.