Look at the block height. Not the tweet.
On August 7, 2025, Bitcoin stood 185 blocks from a moment most of the market has not priced: block 961,632, where Bitcoin Knots โ a niche node implementation โ begins rejecting blocks that do not carry version bit 4. Not because the blocks violate proof-of-work. Not because they break any consensus rule Bitcoin has enforced since 2009. Because BIP-110 demands a signal the network never agreed to give.
The code does not lie, only the narrative. The narrative says this is a non-event. The data says otherwise: a consensus divergence between node implementations, triggered by a proposal with 2.62% miner signalling support, enforced by a single software client that Bitcoin Core โ the reference implementation โ explicitly refuses to execute.
This is not a price story. It is a settlement-layer story. And if you are building on Bitcoin, you need to understand what happens when the rules diverge. Trace the wallet, ignore the tweet.
Context: The Proposal That Would Not Die
BIP-110 is not new. Its lineage runs straight back to the 2015โ2017 block-size wars, when SPV-optimization and block-data-compression schemes circulated as technical alternatives to the politically charged "bigger blocks" agenda. The stated goal: compress block data structures and improve Simplified Payment Verification efficiency. Technically unremarkable. Conceptually sound. Politically orphaned.
Standard Bitcoin soft-fork activation follows BIP-9 mechanics. Miners signal intent via a version bit; if 95% of blocks signal over a 2,016-block difficulty epoch, the fork locks in and activates. The threshold exists for a reason: it ensures a rule change reflects network-wide consent. Consensus changes without consent are not upgrades. They are coups.
BIP-110 breaks this model in three distinct ways.
First, the signalling threshold is 55% โ 1,109 of 2,016 blocks โ not 95%. That alone departs from every modern precedent. Second, and this is the anomaly most analysts miss, the mandatory signalling period begins at block 961,632 regardless of whether the threshold is ever reached. Executing nodes will reject unmarked blocks at that height whether 2% or 90% of miners have signalled. Third, even if nothing locks in, the reduced-data rules are scheduled to activate at block 965,664 anyway.
Read that again. Activation is decoupled from miner consent. This is not a soft fork in the traditional sense. It is unilateral rule enforcement scheduled against an indifferent network.
Core: The Divergence Mechanics
Walk through the physical consequences, because the code is explicit where the commentary is vague.
Bitcoin Knots nodes โ and any OCEAN pool endpoints serving BIP-110-compliant blocks โ will, from block 961,632, treat any block without bit 4 as invalid. Bitcoin Core nodes will accept the same block, because it satisfies every consensus rule Core enforces. One block. Two validity verdicts. Two chains of legal history, from the same parent.
This is the crux. Soft forks historically maintain a single chain because old nodes accept new blocks โ the rules get stricter, not different. Here, the "stricter rule" belongs to a minority client, and the majority network has already said it will not comply.
What happens physically?
Scenario A: OCEAN โ which switched its default job-declaration endpoint to a BIP-110-signalling option on July 15, a full month before the deadline โ continues producing compliant blocks with bit 4 set after 961,632. Knots nodes extend that chain. Core nodes extend whatever chain has the most work, which includes OCEAN's blocks if they are otherwise valid. The chains reconcile, because a block with bit 4 still satisfies Core's rules. The system stumbles but does not fall.
Scenario B: OCEAN's compliant hashrate is trivial against the network. Based on publicly available pool estimates, I calculate OCEAN's share of global hashrate at roughly 1โ2%. If the remaining 97% of miners never set bit 4, Knots nodes stall at the last compliant block. They cannot extend the dominant chain. They wait, or they reorg, or they sit in a state that is, functionally, an island. Anyone relying on Knots nodes for block data โ wallets, explorers, indexers, possibly some self-hosted infrastructure โ sees a frozen chain while the rest of the network moves forward.
That is not a fork in the economic sense. It is a node-implementation outage with no scheduled end time.
My background is on-chain forensics. I have audited consensus-adjacent failures since the 2017 ICO cycle, when I cut through fifteen whitepapers and flagged three fraudulent tokenomics before launch. That experience taught me a simple rule: the most dangerous failure modes are the ones that look like nothing โ a node that silently stops extending the chain, an indexer that serves stale data, a warning that ships without a fix.
The warning shipped here. Bitcoin Knots published a security notice on August 7, 2025, stating that non-executing software โ explicitly including Bitcoin Core โ may not fully validate BIP-110 rules and may leave an unsafe chain state. Let me translate: the client imposing the rule is also the client flagging that everyone else is operating insecurely. That is not a bug report. That is a positioning statement.
Audits reveal the skeleton, not the soul. The skeleton of this proposal is a compressed-block scheme. The soul is a power struggle over who defines Bitcoin's validity predicate.
The State-Machine Risk Nobody Is Discussing
The third-party reproduction work is where the technical story gets genuinely interesting. BlockSlop reproduced a narrow upgrade-delay issue in regtest: when switching from BIP-110-executing Knots back to non-executing Knots or older software, the data directory retained blocks accepted under the old rules. On normal startup, the node did not immediately reconnect inherited history, leaving the node temporarily in a rule-inconsistent state.
No physical database corruption. No mainnet event. But consider the implication: a node implementation can, under specific upgrade sequences, hold two conflicting rule sets in memory while it reconciles its chain state. If that window coincides with a reorg or a disputed block, the node's verdict on validity is undefined for a period.
Knots has since merged protections โ scanning inherited block headers for forced-signalling violations, invalidating offending blocks, and executing a reorg. Good. But the fix only covers violations visible in block headers. Transaction-level or script-level violations that are not visible in headers still require reconnection validation, and potentially a full reindex.
That is the real technical exposure. It is not the BIP-110 rules themselves. It is what happens when execution state and non-execution state mix in the same data directory, on the same machine, on the same network.
I have seen this pattern before. In May 2022, I ran monitoring scripts across ten stablecoin protocols in the days before the Terra collapse. The early warning that saved client positions was not price action โ it was a liquidity anomaly in Curve pools that preceded the narrative by 48 hours. The lesson generalizes: the infrastructure reveals the failure before the market does. The infrastructure here is talking loudly.
The Tokenomic Non-Event and the Economic Asymmetry
BIP-110 changes nothing about Bitcoin's supply schedule. No new issuance. No block-reward alteration. No halving acceleration. The tokenomics are a zero.
The economic risk lives entirely in the fork scenario, and the fork scenario is economically lopsided. In 2017, the BCH split created two chains where every holder received an equal balance on each side, and the market repriced both. If BIP-110 ever produced a persistent alternative chain, its initial value would be near zero. No exchange support. No wallet support. No community consensus. What would remain is a cluster of privacy-maximalist nodes extending an orphaned ledger โ a museum piece, not a market.
The DAO fork of 2016 was the mirror image: overwhelming support for a hard fork that produced ETH and ETC. This is a quasi-hard fork with negligible support. The market's indifference โ 2.62% signalling, zero futures pricing, no narrative traction โ is rational in the short term. But markets price liquidity, not protocol pathology. The pathology compounds in the infrastructure layer, not the order book.
Volatility is the tax on ignorance. The ignorance here is not about price. It is about which nodes you trust to tell you which chain is canonical.
Ecosystem Fault Lines: Who Feels This
Map the dependency chain.
Miners, in aggregate, are unaffected. They continue mining under existing rules, and their economic incentives are unchanged. OCEAN's miners face the only real decision: whether to point hashrate at the signalling endpoint and accept the risk of orphaned work if Knots nodes are the only ones extending that chain.
Core node operators are unaffected. They accept every block that satisfies existing consensus rules, including OCEAN's bit-4 blocks. No action required.
Knots node operators are the exposed class. From 961,632, they lose the ability to extend the dominant chain unless compliant blocks arrive. Their view of Bitcoin diverges from the network's view. The divergence manifests as stale height, failed reorgs, and in the worst case, permanent isolation.
Downstream infrastructure is the silent victim. Exchanges, wallets, and block explorers running Knots as a backend will serve different data than those running Core. Consider a self-hosted custody service that indexes using Knots: after 961,632, it may report zero confirmations for blocks the rest of the network has already settled. Users see delayed balances. Support tickets multiply. Trust erodes quietly.
mempool.space's version-bit monitoring data is a useful barometer here. Its precision implies deep node-API access, and its signalling dashboard is currently the most reliable window into miner behaviour. Watch the dip in that dashboard at the deadline. Data source matters as much as the data itself.
Core's Refusal Is the Real Story
Let me be precise about where the authority lies. A BIP-110 implementation PR in Bitcoin Core was closed on March 26, 2025, without merge. Core contributor Antoine Poinsot, speaking personally on June 4, 2025, confirmed Core would not implement the proposal.
These two facts matter more than any signalling metric.
Historical consensus evolution on Bitcoin has been conservative by design. It works because implementations converge on the same rules. When they diverge, you do not get a fork โ you get two networks claiming the same name and the same genesis block, with different validity predicates. Bitcoin survived this once, in the contentious Blocksize War, when the chain split into BTC and BCH in 2017. That was a hard fork: explicit, declared, and settled by hashrate.
BIP-110 is worse in one specific respect. It is a soft fork in name but a hard fork in behaviour โ a minority client rejecting blocks the majority deems valid, on a schedule the majority never ratified.
Contrarian: The Fork Is Not the Risk. The Precedent Is.
Take the contrarian position seriously. The most common reading is: "BIP-110 is a zombie proposal from a fringe client; it will fail; move on." That reading is correct about the fork and wrong about everything else.
The genuine risk is the precedent of enforcement without consent.
If Bitcoin Knots proceeds with forced signalling at 961,632 โ with 2.62% support, without Core's participation, without a merged implementation โ and suffers no consequence, the rule of minimum consensus is weakened. Any future implementation with a specific agenda can copy the playbook: ship the rule, set the schedule, warn the dissenters. The community's discovery that "the majority client won't enforce it" becomes a weaker defense the second time it is tested.
Here is the anti-contrarian counterpoint, and it is equally important: Bitcoin's multi-implementation architecture is working exactly as designed. Core does not capitulate. It executes its own judgment. The safety valve exists. But the safety valve is only as good as the clarity of the resulting divergence. The market does not have a protocol for pricing "two nodes, two truths, one ticker." That ambiguity is itself the cost.
My read: the chain will not split. OCEAN's hashrate cannot sustain a competitive fork, and economic gravity will pull the network back to the heaviest chain. But the chain not splitting is the easy outcome. The harder outcome is the slow erosion of the assumption that all Bitcoin nodes agree. The 2017 ICO due-diligence cycle taught me that the most expensive narrative is the one that is technically true but institutionally naive. BIP-110 is technically survivable. Institutionally, it is a stress test of whether Bitcoin can absorb a rule divergence without fracturing its own legitimacy.
Takeaway: What to Watch at 961,632
Three signals matter, in order.
First, hashrate. Watch whether OCEAN's signalling endpoint maintains a continuous, non-orphaned chain of bit-4 blocks for more than a few hours after 961,632. Continuous production means the divergence is real and self-sustaining. An occasional orphan means the network is absorbing the friction.
Second, Knots uptime. If Knots nodes stall, the infrastructure around them stalls. Watch mempool.space and any indexer with known Knots dependencies for stale block-height reporting. A stalled height marker is the visible symptom of an invisible rule divergence.
Third, Core's response. If Core publishes an updated compatibility notice or an explicit statement on chain-state safety, the conflict has escalated beyond the fringe. Silence means Core expects the disruption to be contained.
Pegs break, principles remain, portfolios vanish. Bitcoin's peg to a single consensus rule is the principle that protects every portfolio built on it. The code does not lie, only the narrative. At block 961,632, the narrative says nothing will happen. The code says something will happen โ the open question is how long the network takes to absorb it.
The next 185 blocks decide the precedent. Watch the block, not the tweet.