The BIP-110 Post-Mortem: Two Blocks, Eight Hours, and the Arithmetic of Bitcoin's Hashrate Veto

Hook

Block 961,632. Block 961,633. Block 961,681.

The first two numbers represent the entire block production of a Bitcoin fork over eight hours. The third is where the main chain stood when that window closed. At one block per ten minutes, the theoretical output for 480 minutes is 48 blocks. The BIP-110 chain found two.

That is a hashrate share of roughly 4.16 percent.

This is not a consensus. It is not a movement. It is not even a protest. It is a diagnostic readout — a handful of miners, possibly a single pool, pointing hardware at a chain that no economically significant actor recognized. The ledger does not lie, only the narrative does, and the narrative here collapsed in a single work shift.

I have traced consensus failures before. In 2018, I spent 200 hours manually walking the ERC-20 logic of a failed ICO, submitting a vulnerability patch anonymously because the bounty would have compromised the finding. In 2022, I reconstructed the Terra death spiral from 50,000 transactions and concluded it was deterministic failure, not panic. The BIP-110 fork is a simpler specimen. Its autopsy takes eight numbers, not 50,000.

Context

BIP-110 was never a scaling proposal. It did not promise throughput, lower fees, or finality. It was an attempt to restrict what Bitcoin's block space could carry — specifically, a rule change that would reject blocks containing non-financial data writes. The target was Ordinals. The target was inscriptions. The target was BRC-20 tokens and every data-intensive use case that had turned Bitcoin block space into something other than a pure settlement ledger.

The mechanism was the interesting part. This was not a standard BIP-9 miner-activated soft fork, where miners signal support over a difficulty period and activate a rule once 95 percent (or 55 percent, in this proposal's unusual threshold) agree. BIP-110 node operators executed a user-activated soft fork variant — a flag-day enforcement. At height 961,632, nodes running the BIP-110 client began rejecting any block that did not carry the signaling bit, regardless of how much hashrate actually backed the rule.

The support numbers made the outcome predictable before the first rejected block. In the previous difficulty period, only 51 of 2,016 blocks — 2.53 percent — carried the BIP-110 signal. The proposal's own activation threshold was 55 percent. The gap between 2.53 and 55 is not a debate. It is a data point about the absence of miner consensus.

And yet the fork was triggered anyway. That is the part worth dissecting. Someone chose to activate a rule that the mining ecosystem had rejected by a factor of twenty. That is not governance. That is a deployment error.

Core

The Technical Autopsy: A Fork Is a Claim on Hashrate

A blockchain fork is not a code change. It is a claim on a production resource — the hashpower that secures, orders, and finalizes transactions. The code is the least important component. The BIP-110 chain proves this: its code executed as written, and it still failed, because code does not mine blocks. Miners do.

Let me establish the arithmetic. Bitcoin's average block interval is ten minutes. Over eight hours, the network should produce approximately 48 blocks. The fork chain produced two. Under a Poisson process, the probability of finding two or fewer blocks in 48 expected blocks, given the full network hashrate, is effectively zero. The only consistent explanation: the fork attracted roughly 4 percent of the network's hashrate — just over 4 EH/s against a mainnet order of magnitude near 500 EH/s.

That is not a marginal showing. That is below the threshold required for even a stable, insecure chain. A chain with 4 percent hashrate faces an adversarial reorg risk that is not theoretical but structural. Any miner with 10 percent of the network — and there are several — could rewrite the fork's entire history on a whim. A single determined pool could double-spend every transaction on it. The fork's two blocks were not the beginning of a new ledger. They were two confirmations of a dead ledger.

The security implication is so obvious that it is easy to miss: a fork that lacks hashrate is not dangerous to the main chain. It is dangerous only to its own users. Anyone who received "fork BTC" in those eight hours holds a token that can be reorged, erased, or spent back by the very few miners who produced it. (Story: omitted — but the risk is symmetric.)

The flag-day trigger deserves more scrutiny. Standard Bitcoin improvements move through BIP-9 signaling: miners express support in their block headers over a defined period, and only at a high threshold does the rule activate. The design exists precisely to prevent a minority from imposing a rule on a majority. BIP-110's activation path — nodes unilaterally rejecting blocks at a fixed height without hashrate support — is the UASF playbook. UASF worked conceptually for SegWit because it had overwhelming economic and mining support behind it in the end. Here, the support never materialized. The activation was not a last resort. It was an ambush with 48 soldiers against a division.

The absence of a public testnet phase, the absence of a peer-reviewed spec release in the report, and the absence of a realistic signaling campaign all point to the same conclusion: this was not a technical proposal that failed its engineering test. It was a political proposal that failed its political test. The code worked. The consensus was absent.

The Arithmetic of a Dead Chain

Let me run the numbers again, from a different angle. Eight hours. Two blocks. At the fork chain's observed production rate, the expected time to reach 100 confirmations is roughly 400 hours — 16.7 days. Bitcoin's main chain would produce about 2,400 blocks in the same period. No exchange listing, no wallet integration, no merchant adoption survives that kind of divergence. Liquidity does not build on a chain that cannot produce blocks faster than one every four hours.

The "two blocks" finding is also informative about who mined them. With only 51 signaling blocks in the prior period — roughly 2.53 percent — the actual pool of supporters was small. The fork's block producers were likely drawn from a niche subset: ideologically committed miners, a single commercially marginal operation, or opportunists who saw a window to mint the first fork coins before the chain died. The last hypothesis is the most damning, and it cannot be excluded. In low-hashrate forks, the first blocks are effectively free mints. There is no cost to mining a chain that no one values, and there is strategic value in being the first entity that can claim to have "supported" a piece of Bitcoin history.

Structure outlives sentiment; code outlives hype. The structure of Bitcoin's security model — hashrate-weighted finality — did exactly what it was designed to do. It absorbed a minority rule change and excreted it as a two-block footnote.

The Tokenomics of Block Space

BIP-110's failure was not only technical. It was economic, and the economics made the technical outcome inevitable. Consider the incentive structure of the actors involved.

Miners earn two things: the block subsidy and transaction fees. Since 2023, Ordinals inscriptions have become a non-trivial fee stream. Data-heavy transactions — inscriptions, BRC-20 mints, indexer traffic — pay above-average fees because they consume disproportionate block space. This is not a rounding error. At peak activity, inscription-related fees represented a meaningful percentage of total transaction fee revenue. For a miner deciding whether to adopt a rule that bans those transactions, the calculation is simple: the rule reduces fee income by some material fraction, and in exchange offers nothing to the miner. No subsidy increase. No security improvement. No competitive advantage over other pools. The BIP-110 proposal was, from a miner's perspective, a pure cost.

This is the deepest insight of the fork's failure. The anti-Ordinals faction framed the debate as a defense of Bitcoin's "purity" as a monetary network. But the miners who matter treated it as a question of revenue. They did not defect to the fork because the fork would have made them poorer. The absence of 55 percent signaling was not an ideological statement. It was a line-item budget analysis performed by every major pool simultaneously.

There is a structural parallel to my Terra investigation. UST's death spiral was often described as a market panic, but the ledger showed it was a deterministic extraction mechanism: a mint/burn arbitrage that rewarded rational actors for de-pegging the system. BIP-110 is the mirror image. The ledger shows miners rationally declining to endorse a rule that taxes them. No one had to coordinate. No one had to campaign. The incentive structure produced the outcome automatically. Panic is just poor data processing in real-time; so is triumphalism. The data here was a 2.53 percent signal rate in the prior period. The market processed it correctly.

On the other side of the ledger, the fork's token is worthless in economic terms. Fork BTC shares the codebase of mainnet Bitcoin but lacks the security, the liquidity, the exchange listings, and the user base. It is a ledger with no settlement guarantee. Any exchange that lists it is listing a liability. Any user who accepts it in payment is accepting a claim on a two-block chain. "Collateral was a mirage; solvency was a myth" applies to the fork's entire asset base: there is no collateral, no solvency, no issuance schedule that survives the missing hashrate.

The Governance Verdict: Multi-Party Settlement

Bitcoin governance is not a company board. It is a settlement process between four classes of actors: developers who write proposals, node operators who set validation rules, miners who produce blocks, and users who transact. A change is real only when all four converge. BIP-110 attempted to collapse this process into a single actor — the node — and the other three refused to follow.

The result was a clean demonstration of the "hashrate veto." Miners do not need to vote "no" in a formal sense. They need only decline to produce blocks under the new rule. That is what happened. The fork chain stopped at 2 blocks while the main chain continued to 961,681 and beyond. The veto was silent, automatic, and final.

History provides the contrast. The Bitcoin Cash fork of 2017 succeeded in the narrow sense that it produced a persistent chain — because it had major mining pools, exchanges, and a large community of ideological supporters. Bitcoin SV in 2018 had deep-pocketed backers. BIP-110 had neither. It had a flag-day and a 2.53 percent signaling rate. Its failure rate was therefore deterministic, not probabilistic.

The governance lesson is worth stating plainly: in Bitcoin, the cost of initiating a rule change is low, but the cost of making it stick is the entire economy of the network. The system's resilience is not in its code. It is in the redundant layers of consent that a proposal must traverse. BIP-110 was a single-layer attempt. It failed at the second layer, and the failure was visible within eight hours.

There is a hidden risk in this victory, and it is uncomfortable for Ordinals proponents. The mining veto is not an endorsement of inscription culture. It is a defensive position on fee revenue. Miners will abandon Ordinals the moment the fee stream dries up or a better-paying use of block space appears. The alliance between miners and the Ordinals ecosystem is a transaction, not a marriage. Anyone who reads BIP-110's failure as permanent protection for BRC-20 tokens is reading the wrong ledger.

Market, Risk, and the Zombie State

For the market, the event was a risk-release, not a price catalyst. Bitcoin's main chain never stopped producing blocks; the disruption was confined to a chain with 4 percent hashrate. Mainnet pricing remains anchored to ETF flows and macro liquidity, not to a failed minority fork that produced two blocks. The correct market read is structural: the "protocol-level ban on Ordinals" tail risk — the scenario where Bitcoin itself refuses non-financial transactions — has been pushed off the table for the medium term. That is a positive for the Ordinals ecosystem and a negative for anyone who held the fork's token.

The operational risks cluster around confusion. Users may misconfigure nodes, point RPC endpoints at the wrong chain, or accept fork tokens in payment. Exchanges face a listing-review problem if anyone attempts to float a "BIP-110 Coin." The mitigations are mundane: verify chain health metrics, check hashrate, confirm the block height matches the main network. The ledger supplies the test. A chain producing 48 blocks per 8 hours is the main chain. A chain producing 2 is a corpse.

A zombie state remains possible — a low-hashrate chain kept alive by a few ideologues or bots, producing a handful of blocks per week. Such a chain would have negligible economic impact but persistent nuisance value. It would function as a symbol for the anti-Ordinals wing and a recurring source of confusion. The probability is low but not zero; the report's own confidence estimate on this is low-to-medium. I would add that the psychological value of keeping a corpse alive is sometimes greater than the economic cost of doing so. Emotion is a variable I exclude from the equation, but others do not.

Contrarian

Now the part that makes the anti-Ordinals side uncomfortable: the BIP-110 proponents were not entirely wrong. Their framing was flawed. Their tactics were destructive. Their support base was imaginary. But the underlying concern — that block space is finite, that data-heavy inscriptions impose a real cost on the network, and that the monetization of the block via "cultural artifacts" is a form of rent extraction on a shared resource — is a serious economic argument, not a moral panic.

The fee market is the mechanism that resolves this. Ordinals transactions pay their way; if they do not, they are naturally priced out. The BIP-110 error was trying to short-circuit the market mechanism with a rule. That is the same mistake that killed UST — attempting to impose a price or a rule on a system that had its own pricing function. The ledger resolves disputes through cost, not through coercion.

And here is the genuine blind spot for the winners. The failure of the consensus-rule path does not end the conflict. It shifts it. The next anti-Ordinals campaign will not come as a BIP. It will come as mempool policy — miners and node operators filtering or deprioritizing inscription transactions at the relay layer, where there is no vote, no activation threshold, and no block-height flag. That path is quieter, less transparent, and more difficult to contest than a public fork. BIP-110's failure may have taught its supporters the exact lesson they needed: do not fight the consensus rule. Fight the transactions.

Takeaway

Watch the mempool. The battle over Bitcoin block space was never settled by this fork; it was merely relocated from the consensus layer to the policy layer. The next move will be invisible to the block height and visible only in confirmation times, fee data, and the propagation behavior of nodes. That is where the data lives, and that is where the narrative will be written. The ledger does not lie. It just records the attack in a language most people are not reading.