BIP-110: The Soft Fork That Silently Signals Nothing

The block arrived at height 876,543 with a version bit set to 0x40000000. It was one of the 2.64% of recent blocks signaling support for BIP-110. Across the network, 97.36% of blocks carried no such signal. The forced signaling window is now nine days away. History verifies what speculation cannot: a soft fork with such marginal support does not activate. It becomes a ghost proposal—a technical artifact that reveals more about governance failure than about protocol improvement. BIP-110, titled "Reduced Data Temporary Softfork," is a Bitcoin soft fork designed to limit the size of transaction data fields—specifically the witness data in SegWit transactions and the OP_RETURN output length. Its stated target is the ordinals inscription ecosystem, which has used these fields to embed arbitrary data since early 2023. The proposal imposes a temporary cap: each transaction may commit no more than 80 bytes of OP_RETURN data and 400 bytes of witness data per input. For context, a typical ordinals inscription requires between 300 and 1,000 bytes of witness data. The math is simple: BIP-110 would render most existing and future ordinals minting operations invalid. The activation mechanism is where this proposal diverges from standard Bitcoin governance. BIP-110 uses a mandatory signaling window, a variant of the BIP-8 approach. Under BIP-8, after a defined start height, miners are required to set a specific bit in the block version field to signal readiness. If 95% of blocks within a 2,016-block difficulty period carry the bit, the soft fork locks in. If the threshold is not met by a final deadline, the soft fork automatically activates anyway—the so-called "flag day" activation. BIP-110 appears to follow this pattern, though the specific parameters are still under discussion on the Bitcoin-Dev mailing list. The forced window is the critical design choice. It means that even if miners refuse to signal, the soft fork will eventually become active for upgraded nodes. Those nodes will then reject any block that does not contain the signal bit, effectively creating a chain split. Silence is the strongest proof of truth: the current support rate of 2.64% indicates that the vast majority of hash power is either opposed or indifferent. Foundry, the largest mining pool with approximately 32% of global hash rate, has not issued a public statement. Antpool, at roughly 20%, is similarly silent. Only Ocean, a smaller pool known for its ideology-driven approach, has signaled support consistently. Based on my experience analyzing protocol upgrades since the 2018 winter, I have rarely seen a BIP advance with such minimal community backing. The 2016 BIP-68 soft fork (relative lock-time) had over 90% support before its activation window. The 2017 SegWit activation, though contentious, ultimately achieved the 95% threshold through the BIP-91 compromise. BIP-110 is different: it has neither consensus nor coercion. The 2.64% figure is not a rounding error; it is a statistical death sentence. To understand why, we must examine the game theory. Miners earn transaction fees in two forms: standard payments and fees from ordinals inscriptions. At peak ordinals activity in late 2024, inscription fees accounted for up to 15% of total block reward for some pools. Large pools like Foundry and Antpool have institutional clients—hedge funds, public mining companies—who expect maximum profitability. Supporting BIP-110 would cut off a revenue stream without offering any compensating benefit. The proposal provides no subsidy or fee redistribution. It is pure restriction. Under those conditions, rational miners will not signal unless forced. And since the forced window is designed by developers, not miners, the conflict is structural. Pressure reveals the cracks in logic. The contrarian reading of BIP-110 is not about ordinals at all. It is about the erosion of Bitcoin's governance legitimacy. The mandatory signaling window is an escalation: it bypasses the traditional BIP-9 voting mechanism, which allowed miners to reject a proposal by simply not signaling. Under BIP-8, the fallback activation creates a hostage situation. Upgraded nodes threaten to fork away from the majority chain unless miners comply. This is not governance by consent; it is governance by ultimatum. The 2.64% support is a signal that the broader mining community rejects this approach itself, not merely the technical content of the soft fork. There is also a more technical blind spot. BIP-110 restricts witness data per transaction, but ordinals can be inscribed via alternative methods—such as using the scriptPubKey of Taproot outputs, which are not capped by this proposal. The soft fork would only shift the inscription mechanics, not eliminate them. Complexity hides its own failures: a patch designed to fix one loophole often creates three others. The Bitcoin Core maintainers have previously avoided such scoped restrictions precisely because they lead to an arms race. Looking ahead, the forced window will open, the support rate will remain below 5%, and upgraded nodes will begin rejecting non-signaling blocks. However, those nodes represent an infinitesimal portion of hash power—probably less than 1% of the network. They will form a minority chain that produces empty or orphaned blocks. No exchange will support it; no wallet will transact on it. Within days, the minority chain will halt as miners abandon it. The structure of Bitcoin's incentive alignment will outlast this sentiment-driven rebellion. Patience is a technical requirement. The ordinals debate will continue, but BIP-110 will not be its resolution. The real question is whether the Bitcoin development community can design a governance process that accommodates dissent without threatening the network's unity. The current episode shows that mandatory activation is a dead end. History verifies what speculation cannot: soft forks require consensus, not coercion.