Mandatory signaling has begun. Miner support: below 3 percent. If you spend enough time in protocol governance, those two numbers trigger a specific form of dread. The first sentence implies enforcement. The second sentence makes enforcement impossible. A mandatory signaling period is a mechanism by which nodes refuse to accept blocks that do not carry a particular version bit. It is the software equivalent of a border guard stamping ‘DENIED’ on a passport. But when fewer than 3 percent of the miners have actually flipped that bit, the guard is not protecting the border; he is inventing a new one. Finding signal in the consensus noise means recognizing when a technical activation window has become a political ultimatum.
BIP-110 is one of those historical proposals that tends to be swallowed by Bitcoin’s larger scaling war. It is not a block size proposal. It is not a new opcode. It is a deployment mechanism — a set of rules that determines how a soft fork becomes active. The standard path, eventually encoded in BIP-9, asks miners to signal support with a version bit over a period of time. If 95 percent of hash power signals within a difficulty period, the soft fork lock-in begins. If not, the proposal simply expires. The failure mode is quiet. BIP-110 offers a different route. Rather than waiting for miner cooperation, it makes signaling mandatory for a subset of network rules. Nodes that run the upgraded client reject blocks that do not include the mandated version bit. In theory, this creates an incentive for miners to comply: if they do not signal, their blocks will be orphaned by the enforcing part of the network. In practice, the question becomes who controls enough of the network to make that warning credible.
The report I was asked to dissect contained four usable facts: mandatory signaling is active; miner support is below 3 percent; the activation is being described as a test; and a hard-fork rollback plan is being discussed. That combination is not accidental. It is a specimen of a governance strategy that has been tested, fragmented, and quietly buried — but whose fingerprints remain on every subsequent forced upgrade. This article is an autopsy, not a price prediction. It is also an attempt to map a failure mode that is still repeating across modern Layer 2 governance, sometimes with the same language, sometimes with prettier dashboards, but always with the same structural contradiction: someone is trying to force a state transition that the people responsible for producing the chain do not want.
Let me start with the number itself. 3 percent is not merely a low participation rate. In Bitcoin’s proof-of-work settlement, miner support is not a popularity contest; it is the physical distribution of block production. A 3 percent signal means that 97 percent of the hash power has chosen, explicitly or by default, not to cooperate with the mandatory signaling window. The enforcing nodes are not facing a minority; they are facing an overwhelming majority of the chain’s secular authority. Under those conditions, a mandatory signaling window does not produce a clean activation. It produces a reorganization risk. If the 3 percent faction mines a chain that rejects non-signaling blocks, while the 97 percent faction continues to mine blocks without the mandatory bit, the two groups are no longer sharing a single state transition. They are sharing a historical ledger until the next recalculation point, after which their views of the chain diverge. This is the moment that most governance models fail to capture: the state transition itself becomes conditional on which client you run. That is exactly the kind of state space where the entropy of Layer 2 state transitions is born — a sudden expansion of possible fork trajectories.
I have spent years parsing the entropy in Layer 2 state transitions, and every time I see a forced consensus change, I ask the same question: who bears the cost if the network disagrees? BIP-110 gives a clean answer. The cost would be borne by every node operator and miner who refused to accept the other side’s block. The 3 percent support figure does not mean the proposal was close to failing. It means the proposal had already failed as a technical default and was now operating purely as a social signal. The core insight is that mandatory signaling converts a disagreement over protocol rules into a disagreement over chain truth, and chain truth is not a majority vote. It is a physical process of ordering blocks. If the ordering process rejects the rule change, the rule change exists only on paper and in the memory of a minority client.
Mapping the invisible costs of abstraction layers is useful here. The abstraction is Bitcoin’s ‘consensus’ as a single, legible entity. The cost is hidden in the moment when that abstraction is stressed: node operators believe they are enforcing the true chain, miners believe they are securing the true chain, and neither group can prove the other wrong without producing a split. The mandatory signaling window converts a conceptual disagreement into a mechanical fork. That is the true cost of abstraction — not the gas fee, not the latency, but the moment when two legitimate users of the same protocol see two different truths.
Now we arrive at the uncomfortable economic question. I have seen too many governance post-mortems blame greed or ignorance. But the miners who did not signal were behaving rationally. The BIP did not include a block reward increase, a transaction fee redesign, or any other direct economic benefit. It asked them to change their client behavior, coordinate with pools, and risk disrupting their revenue stream. The expected value of signaling was negative. In that sense, the 3 percent was not a rejection of the technical idea; it was the output of an incentive calculation. The BIP’s authors forgot to change the payoffs. They designed a governance mechanism that treated miners as voters and then forgot that miners are also employees of a capital-intensive, commodity-sensitive industry.
In my audits of optimistic rollup challenge periods, I have seen the same failure take a different form: a protocol assumes that validators will behave like altruistic participants, but the game theory collapses when the cost of participation exceeds the penalty of abstention. BIP-110 is the same skeleton. The 3 percent signal is the statistical signature of a boycott. It is not an accident. It is the aggregate result of thousands of mining decisions, each one made by a firm that looked at the upgrade and concluded that the upgrade was not worth the risk. A protocol can call that lazy. A better analyst calls it rational.
Deconstructing the incident from a governance perspective, the core conflict is not ‘centralized versus decentralized.’ It is the gap between the issuance of rules and the production of blocks. Bitcoin’s implicit governance model has always been tripartite: developers propose, miners dispose, users and exchanges ultimately ratify. BIP-110 tried to change the order. It allowed the developer-node axis to bypass the miner-producer axis. With 3 percent miner support, the proposal lacked the physical majority needed to become a fact. The existence of a hard-fork rollback plan reveals that the developers knew this possibility. They were not betting on success; they were betting on leverage. The rollback plan was not a contingency. It was an implicit admission that the mandatory signaling window was a test of coercion, not a test of adoption.
Unraveling the spaghetti code of legacy DeFi has taught me to spot this pattern: when a governance mechanism needs a definition of failure that is itself a hard fork, it has already abandoned the protocol’s social contract. The hard fork rollback plan in BIP-110 is not a technical detail. It is a confession. It says: if this experiment does not work, we will have to break the chain to undo it. That is the opposite of a safe deployment process. A safe process has an expiration date. A safe process has a default. A safe process does not require the same network to contemplate two competing histories just to unbundle one failed signal.
Traders often misread governance conflict as a binary event: either the fork happens, or it doesn’t. In reality, the market prices in the risk of a split long before the first competing block. In a sideways or low-volatility market, this type of protocol uncertainty can be amplified because there is no bullish catalyst to overwhelm it. The news of a mandatory signal with 3 percent support does not produce immediate selling. It produces a slow repricing of tail risk. Custodians start drafting contingency procedures. Exchange listing committees quietly game out how they would allocate a new asset if the split actually occurred. The visible price effect is often delayed until the most uncertain date on the calendar, typically the activation deadline. By the time the deadline arrives and the signal fails, the market has already moved to the next event. This is why most governance post-mortems look anticlimactic. The drama is not in the result; it is in the period when everyone knows that the result is unknowable.
Regulators will not intervene in a failed soft fork. But the existence of a mandatory signaling window with a rollback plan creates a paper trail that institutional risk teams cannot ignore. When a protocol’s code says that a hard fork is possible, the institution that holds the asset must calculate the cost of being on the wrong side of the split. That cost is not visible on-chain. It lives in legal opinions, accounting memos, and custody operations. The hard fork fallback plan is not an insurance policy; it is a disclosure document. Every mandatory signaling period is therefore also a compliance event. It asks every institutional holder to admit that the asset they hold is not a single, stable thing but a set of possible chains, each with a different balance sheet. This is the invisible cost that never appears in transaction fees.
The ecosystem impact of BIP-110 was small because the proposal was small. But the pattern is not small. When developer-client and miner-producer fall out of line, the entire downstream stack has to prepare for the worst. Wallets, exchanges, and custody providers must all be ready to support two conflicting state histories. This is an operational tax on every participant. In 2015, that tax was manageable. Today, with hundreds of L2s and token standards, the same disconnect would be catastrophic. The reason I keep returning to this episode is that it highlights a principle the modular stack often forgets: the closer the base layer is to a split, the more every abstraction built on top of it becomes a liability. If Bitcoin’s consensus view forks, every L2 secured by those state roots forks too. The entropy in Layer 2 state transitions is not generated by L2s themselves. It is inherited from unresolved L1 governance conflicts. Parsing that entropy is a discipline, not a hobby.
Maybe I should give BIP-110 more credit: mandatory signaling can work if three conditions are met. First, the enforcing node population must be large enough to create a credible minority chain. Second, the non-signaling miners must face a heavier penalty from being orphaned than from upgrading. Third, there must be a clear date at which the uncertainty collapses. BIP-110 satisfied none of these. The 3 percent support rate tells us that the node-heavy city was not prepared to impose a meaningful penalty. The hard fork rollback plan tells us that the authors themselves were not confident enough to commit. A mechanism with an exit plan is not a mechanism; it is a threat. A threat with no execution probability is a negotiation. Bitcoin opted for BIP-9’s negotiated approach. The cost was slower activation. The benefit was that the chain never had to learn whether it could survive a forced split.
Here is the contrarian reading: BIP-110’s mandatory signaling period was a success, even in failure. It forced Bitcoin to confront the uncomfortable truth that miners are not a veto-proof oligarchy. By opening a mandatory signaling window with almost zero support, the developers established something more valuable than an activation: a precedent. They demonstrated that a small, organized group of node operators can introduce a credible fork threat. That threat became a bargaining chip. It changed future negotiations. In any governance system, the ability to impose a transaction cost on an adversary is power. The cost imposed by BIP-110 was small, but the message was not: miners do not, and should not, have the final word on protocol logic. The mandatory signal was a form of verbal reprimand, an announcement that the code can always be forked. The low support rate was irrelevant. The signal was the signal.
Every actor treated 3 percent as too small. But what if the 3 percent had been enough? Imagine a small group of nodes, segments of mining hardware, and a niche exchange prepared to list the minority chain. It would not have to win the hashrate war. It would only need to create enough ambiguity to force the 97 percent to spend time and money defending their own chain. In a contest over accounting entries, the attacker does not need a majority; he needs enough confusion to make consensus unreliable. BIP-110’s low support rate made it a joke. But the logic of crypto governance is asymmetric: a determined minority with a clear rulebook and a credible exchange listing can always produce a fork. That is the lesson that later UASF proponents internalized. The failure of BIP-110 was not the failure of the force. It was the failure of the audience to believe.
The historical afterlife of BIP-110 matters more than its original activation. BIP-9 eventually became the standard mechanism for soft forks because it made miner coordination the gatekeeper. But the question of what happens when miners refuse to coordinate did not disappear. It resurfaced in the 2017 SegWit debate. That time, the nodes did not wait for 95 percent miner support. A user-activated soft fork threat, BIP-148, was created to reject blocks that failed to signal SegWit. The same logic as BIP-110 — nodes force miners to comply — was resurrected under a more politically palatable name. The fact that SegWit ultimately activated is often cited as proof of UASF’s success. But the deeper lesson is even more uncomfortable: mandatory signaling worked only when enough node operators and economic actors were prepared to credibly walk away from the current chain. BIP-110’s failure was not because the mechanism was flawed; it was because the credibility wasn’t there yet. By 2017, the political environment had changed enough to supply it. The mechanism remained a mutation, ready to be activated when the social conditions allowed.
What should a modern observer take from this story? First, when a protocol reports low validator or miner participation in a mandatory signal, do not interpret it as technical apathy. Interpret it as a negotiated refusal. Second, when a proposal includes a hard fork rollback plan, understand that the proposal is not designed for success. It is designed for leverage. Third, when the market fails to react to governance conflict, remember that the market is often waiting for a clearer date of existential uncertainty. By then, the positioning has already happened behind closed doors. The public signal is delayed. The tail risk is repriced in custody agreements before it ever appears on a candle chart.
The ghost of BIP-110 is still in the machine. Every protocol that tries to force a state transition over the objection of its own production layer — whether it is an L1 telling validators to accept a new rule, or an L2 telling sequencers to stop censoring — is replaying this experiment. The lesson is not that mandatory signaling always fails. It is that the force of code is meaningless without a social settlement that can make the majority blink. BIP-110 had 3 percent support. The next proposal might have 30 percent. The question is not about thresholds. It is about whether the people who run the protocol believe the minority has enough conviction to split. As I watch 2026’s L2 governance debates, I keep waiting for the moment when another BIP-110 appears: a clean mechanism, a tiny support base, and a rollback plan that pretends the hard fork was never the point.
