Miners Ignored a Fork Built to Ban Ordinals

For two years, one faction of Bitcoin has wanted to evict the Ordinals and BRC-20 tokens that pack blocks with images and data instead of payments. This week they tried to force the issue in code — and Bitcoin’s miners answered by doing almost nothing.

BIP-110, a soft fork designed to restrict non-financial data for a year, reached its mandatory-signaling block, 961,632, with support from just 2.53% of miners — 51 of the preceding 2,016 blocks, against a 55% threshold. Nodes enforcing the rule began rejecting blocks that failed to signal, splitting off a minority chain. Starved of hashpower, that chain stalled and fell behind the main network’s proof-of-work within hours.

How a fork actually passes

Bitcoin has no CEO and no ballot. A rule change activates only when a supermajority of miners signal for it and the economic nodes — exchanges, wallets, custodians — run the software that enforces it. Volume on social media counts for nothing; hashrate and node adoption count for everything.

BIP-110 had neither. Miners earn fees from the very Ordinals transactions the fork sought to curb, and had little reason to torch that revenue. Without them, the change was a suggestion the network declined.

The Ordinals fight isn’t over

The failed activation settles this attempt, not the argument. Ordinals defenders call the data a legitimate use of blockspace that pays real fees; critics call it spam that raises costs for ordinary users and bloats the chain. Both sides will keep pushing — through relay policy, through mining-pool defaults, and through the next proposal.

What the episode shows is the system working as designed. A contentious change that lacked broad consent could not be imposed by a determined minority. The same mechanism that frustrates would-be reformers is the one that stops anyone — a government, a company, a mob — from rewriting Bitcoin by force.

Why it matters: Bitcoin’s rules don’t change because a faction is loud. They change when miners and nodes actually run the code.

Block height and signaling percentages are as reported in the cited review; a single-source account of an on-chain event that can be verified against block explorers.