Block 961,632. That is where the fiction ended.
BIP-110 nodes rejected every subsequent block that lacked their activation signal. At 961,633, a single miner produced the first block under the new, restricted rules. Then, silence. Eight hours later, the Bitcoin main chain had advanced forty-nine blocks to height 961,681. The BIP-110 fork chain had produced exactly two blocks in that same window. Not forty-eight. Not twenty-four. Two.
The math is unambiguous. At Bitcoin's ten-minute average block target, eight hours of expected production is roughly forty-eight blocks. The fork chain produced two. That implies a hashrate share of approximately 4.2 percent of the main chain's mining power. In plain terms: the BIP-110 fork attracted less computational weight than a dying ghost chain. It was not a competing network. It was a corpse with a block reward.
And yet — this is the part that warrants forensic attention — the fork was triggered deliberately. This was not an accident of a stealth miner wandering into an incompatible rule set. Node operators running BIP-110 software force-activated the rule change at a predetermined block height. They knew the signaling numbers. They had watched 51 out of 2,016 blocks in the previous difficulty window carry their signal — a paltry 2.53 percent. The activation threshold was 55 percent. The distance between those figures is not a negotiation gap. It is a chasm. And they pulled the trigger anyway.
That decision, and the mathematics of its aftermath, is the subject of this report.
Context: What BIP-110 Actually Proposed
BIP-110, on its face, is a proposal about block space hygiene. It restricts non-financial data writes to Bitcoin blocks — a direct, technical assault on Ordinals inscriptions and BRC-20 token operations. Inscription transactions embed arbitrary data in the witness or script segments, effectively turning Bitcoin's block space into a medium for content storage. Since the first inscription wave in early 2023, this practice has grown from a curiosity into a genuine economic phenomenon: digital artifacts, token protocols, and metadata schemes have all found their way onto Bitcoin's base layer.
To its proponents, this is pollution. The argument runs as follows: Bitcoin's block space is not a general-purpose database; it is the settlement layer for a global monetary network. Every inscription transaction displaces a financial transaction. Every byte of arbitrary data raises the cost of running a node. Every BRC-20 token mints a claim on block space that has nothing to do with payment finality. From this perspective, Ordinals represent rent extraction — a degradation of the system's monetary role.
BIP-110's solution was not a soft tap. It was a hammer. The proposal never followed the standard activation protocol. The conventional path in Bitcoin is BIP-9: miners signal support within a difficulty adjustment window, and the rule change activates when the threshold — typically 95 percent — is reached. BIP-110 was not deployed that way. Its node software included a flag-day trigger: at block height 961,632, any block lacking the BIP-110 signal would be rejected outright. This is a user-activated soft fork (UASF), a coercion mechanism that pressures miners into compliance through node-level economic weight.
It is a clever design in theory. It is a catastrophic design without hashrate. The numbers going in were dire. In the prior difficulty adjustment cycle, only 51 of 2,016 blocks carried BIP-110's signal. The proposal required 55 percent to activate through the conventional path. It received 2.53 percent. The fork proceeded anyway.
This analysis examines three things: the hashrate mathematics that dictated the failure, the miner economics that made it predictable, and what the governance autopsy reveals about Bitcoin's actual decision procedure.
Core Evidence Chain I: The Hashrate Mathematics
The forensic starting point is the block interval data. Let me lay it out as a timeline.
At block 961,632, BIP-110 node software began rejecting blocks without the signal. Immediately after, a block at height 961,633 appeared — the fork's first block. The fork chain then produced one more block in the ensuing hours, for a total of two. The main chain, meanwhile, advanced from 961,632 to 961,681 in the same eight-hour window. That is forty-nine blocks at an average interval of 9.8 minutes, consistent with the network's expected ten-minute schedule.
Using the ratio of fork blocks to main-chain blocks as a hashrate proxy:
- Main chain: 49 blocks in 8 hours
- Fork chain: 2 blocks in 8 hours
- Implied hashrate ratio: 2/49 = 4.08 percent
This is, if anything, an upper bound. The fork chain may not have inherited the same difficulty level, and early blocks can benefit from a stale difficulty adjustment window. In practice, the fork's hashrate share likely sat below 5 percent. To frame this in absolute terms: the Bitcoin network operates at approximately 500 exahashes per second (EH/s). A 4 percent share is roughly 20 EH/s. That would require on the order of 200,000 modern ASIC miners — hardware worth billions of dollars. No rational economic actor committed that level of resources to a fork with no exchange listing, no wallet support, and no institutional endorsement. The more likely scenario is that the two blocks came from a small set of mining rigs running experimental software, or from a miner who accidentally found a block while testing configuration changes.
The security implications are not theoretical. A chain with 4 percent of the main chain's hashrate is vulnerable to a 51 percent attack by any miner or pool controlling more than 2 percent of global hashrate — which is to say, dozens of existing entities. The main chain miners themselves could trivially reorganize the fork chain at will. Even if the fork had continued producing blocks at its observed rate, transaction confirmations on it would carry zero meaningful finality. Double-spend risk would be structural, not hypothetical.
And here is the critical point: the failure cannot be attributed to BIP-110's code quality in any simple sense. There is no publicly documented consensus bug, no cryptographic flaw, no catastrophic edge case. The proposal failed because it lacked the physical resources to exist. Bitcoin's consensus, at its core, is not a product of elegant code. It is a product of energy and work. BIP-110's developers wrote the rules, but they did not bring the watts.
This is the first insight worth holding onto: BIP-110's fork failure is not a software bug. It is a power failure.
Core Evidence Chain II: The Miner Economics
The 2.53 percent signaling number was not the cause of the failure. It was a symptom. The underlying cause is an economic variable the BIP-110 authors appear not to have modeled: miner fee revenue from Ordinals-related transactions.
Look at the data honestly. Since Ordinals inscriptions entered Bitcoin's block space, they have become a meaningful contributor to fee revenue. Inscription-heavy blocks have frequently produced fees well above those of standard transaction blocks. During congestion periods, inscription fees have spiked to levels rarely seen in Bitcoin's pre-inscription history. Some individual blocks have captured fees worth several Bitcoin — paid by users racing to inscribe content and to mine BRC-20 tokens. Over the life of the Ordinals ecosystem, this has accumulated to tens of millions of dollars in miner revenue.
This is not rounding error. It is a revenue stream.
Miners are not ideology machines. They are commercial operators with electricity bills, hardware depreciation schedules, and debt obligations. When a proposal arrives that would restrict or eliminate a meaningful fee stream, the rational response is not a principled debate about the nature of money. The rational response is to ignore the proposal. Do not signal for it. Do not mine for it.
The causal chain is short:
- BIP-110 restricts non-financial data writes.
- Ordinals inscriptions are, in nearly all cases, non-financial data writes.
- The restriction would end inscription fee revenue.
- Miner income would decline by the proportion of fees attributable to inscriptions.
- Miners signal and mine according to their economic interest.
- The proposal received 2.53 percent signaling.
- The fork produced two blocks.
You do not need a governance theory to explain this sequence. You need a spreadsheet. The entire story is visible in miner fee data. I built custom SQL queries on Dune Analytics during the 2021 DeFi mania to track liquidity flows across hundreds of tokens, and the pattern I keep seeing is that market participants respond to incentives with astonishing consistency. Bitcoin miners are no exception.
There is an uncomfortable implication here for those who want to read BIP-110's failure as a principled victory for permissionless innovation. The miners did not reject BIP-110 because they believed in the freedom to inscribe. They rejected it because inscription fees are profitable. The alliance between miners and the Ordinals ecosystem is a coincidence of incentives, not a marriage of ideology.
Consider the counterfactual. If inscription fees were negligible — if Ordinals had never become economically significant — BIP-110's signaling would still have been low, but the fork might have received a few more blocks from curiosity-seekers. The economic pressure to oppose the proposal would not have existed. And if inscription fees had become large enough to threaten the fee market balance in the other direction, miners might have actively defended the practice with public statements.
This is the second insight: what looks like Bitcoin "voting with hashrate" is really Bitcoin "pricing with hashrate." The medium of exchange in Bitcoin governance is not arguments. It is work.
Core Evidence Chain III: The Governance Autopsy
The final piece of the investigation concerns governance mechanics. BIP-110's fork attempt was a UASF — a deliberately adversarial activation mechanism. How did it produce such unequivocal failure?
Bitcoin's consensus change process is built around a specific assumption: miners and node operators share a rough alignment of interest. BIP-9 works because miners signal only when the broader ecosystem — economic nodes, exchanges, wallets, institutional users — has reached sufficient consensus. The 95 percent threshold is intentionally conservative. It exists to ensure that no change activates without near-universal acceptance.
BIP-110's authors rejected this model. They substituted a flag-day UASF, which requires only that enough node operators run the software to create economic coercion. The theory is that if a significant percentage of economic nodes reject blocks lacking the signal, miners will fall in line out of fear of mining orphaned blocks. The flaw in this theory is that "enough node operators" was never measured — and, by every available proxy, it was never close. There is no evidence of major exchange support, wallet adoption, or institutional endorsement of the BIP-110 rule set. Without those, the UASF was not a pressure campaign. It was a statement. A declaration of intent backed by approximately zero institutional weight.
Historical context is useful. The Bitcoin Cash fork of 2017 had substantial miner support, exchange backend coordination, and a coherent business faction driving it. The BSV fork of 2018 had a wealthy patron willing to absorb sustained losses. Both forks survived for years — not because they were technically superior, but because they had real backing. BIP-110 had none of this. No major pool publicly endorsed it. No major exchange announced listing plans. No significant corporate entity backed it. It was, in effect, a solo gesture.
Let me draw on my own audit experience here. In 2019, I spent three months performing a line-by-line review of the Zcash protocol's shielded transaction logic, hunting for edge cases in the proof verification loop. That process taught me something that applies directly to this event: code, no matter how well-written, has no authority outside the network that chooses to run it. You can write the most elegant rule enforcement mechanism in the world, but if the surrounding ecosystem does not honor it, it is a text file. It is not a law.
Bitcoin's BIP-110 client could reject blocks until the heat death of the universe. No miner mining under those rules means the fork remains a database with two records and a very short story. The governance system, in this instance, worked exactly as designed: a proposal with 2.53 percent support, a chain with roughly 4 percent hashrate, and a result that was a non-event. The main chain advanced to 961,681 and kept going. Context reasserted itself. The system's resilience is not a feature of its constitution or its rulebooks. It is a feature of the physical distribution of hashrate.
This is the third insight: in Bitcoin, the flag-day belongs to whoever controls the energy. Everything else is commentary.
Contrarian: What This Actually Reveals
The easy narrative is that BIP-110's failure proves Bitcoin's robustness against hostile forks and confirms Ordinals' legitimate place in the ecosystem. A more careful reading points in a different, less comfortable direction.
First, the failure proves no principle other than "miners maximize revenue." Return to the counterfactual: if Ordinals inscriptions had never generated meaningful fees, BIP-110 would have approached activation with only slightly more drama than a routine GitHub pull request. Ordinals would have been banned — not by vote, not by community referendum, but by default. The entire difference between the Ordinals ecosystem surviving and being protocol-banned comes down to a few thousand dollars of daily miner revenue. That is not a principled defense of open block space. That is a market outcome with no moral weight.
Second, the "victory" for Ordinals is structurally fragile. Consensus-level attacks have been repelled, but the conflict has not ended. The BIP-110 approach was direct: change the rules. Future approaches will be indirect: raise the cost of inscription transactions, filter them at the mempool level, or create block templates that systematically deprioritize non-financial data. These require no fork and no signaling threshold. They require mining pools to adjust transaction selection policies. And mining pools are already doing this to varying degrees based on fee profitability.
Third — and this is the part that deserves more attention than it is getting — why was the fork triggered at all? A rational actor with 2.53 percent support would have deferred activation and spent months or years building consensus. Instead, the activation trigger was pulled immediately. That suggests the goal may not have been activation at all. It may have been visibility. A theatrical fork, in a sense. The resulting attention, debate, and polarization may have been the intended output all along. In that light, BIP-110's "failure" is closer to a successful publicity operation with a two-block garnish.
There is one final risk marker worth flagging. The fork chain's two blocks exist. If anyone continues mining, the chain will produce additional blocks. The token on that chain is not Bitcoin. It is a shell asset — a claimed fork with zero hashrate, zero liquidity, and zero exchange recognition. Rug pulls are just math with bad intent, and this one has the math written all over it. Any exchange that lists this fork token without a full chain-health audit is creating a liability for its users.
The uncomfortable truth is that BIP-110 was never a serious attempt to change Bitcoin. It was a probe. And now we know, with quantitative certainty, where the floor is: two blocks and a dormant chain.
Takeaway: The Next Front
Watch the fee data, not the forks. BIP-110's consensus-level challenge to Ordinals has failed. The next battlefront will be operational: mempool policy, block template construction, and transaction relay rules. These require no fork and no signaling threshold — only configuration changes at the pool level. That is harder to track and harder to fight.
The alliance between miners and the Ordinals ecosystem is contingent on one variable: inscription fees as a share of total miner revenue. As long as that number holds up, the case for inscription-friendly block space is economically self-enforcing. The moment it declines, the balance shifts. The signal is visible today in Dune dashboards. The question is who will read it in time.
Check the calldata, not the headline. Check the block intervals, not the press releases. Hashrate votes. Everything else is commentary.
The two blocks of BIP-110 remain in the chain's archaeological layer — a curious stone with an inscription of its own: barely any hashrate, no support, and a lesson embedded in its orphaned headers.