Bitcoin's BIP-110 soft fork — a one-year proposal to restrict non-financial data in transactions — triggered a minority chain split on August 9, 2026, at block 961,632. The fork mined two blocks in eight hours, then stalled. The main chain advanced 48 blocks in the same period. Miner signaling pe...
"If BIP-110 doesn't pan out, the only option is a PoW change." — Luke Dashjr, Bitcoin Core contributor and Bitcoin Knots maintainer
Bitcoin's BIP-110 soft fork — a one-year proposal to restrict non-financial data in transactions — triggered a minority chain split on August 9, 2026, at block 961,632. The fork mined two blocks in eight hours, then stalled. The main chain advanced 48 blocks in the same period. Miner signaling peaked at 2.64% against a required 55% threshold.
The episode amounts to the most consequential failed activation attempt since the 2017 SegWit2x collapse. Within 24 hours, BIP Editor Murch filed a motion to remove Luke Dashjr — the proposal's principal architect — from the BIP editorial board, citing conflicts of interest. Developer Chris Guida has rebased Dashjr's 2017 proof-of-work hard fork code onto Bitcoin Knots as a contingency. The fallout now extends beyond technical debate into a governance crisis over who controls Bitcoin's upgrade process.
BIP-110, formally titled the Reduced Data Temporary Softfork (RDTS), proposed seven consensus-level restrictions on data embedded in Bitcoin transactions. The rules would have capped most new outputs at 34 bytes, limited OP_RETURN to 83 bytes, and restricted data pushes to 256 bytes. The restrictions targeted inscription techniques used by Ordinals, BRC-20 tokens, and Runes — protocols that use Bitcoin's block space to store images, tokens, and arbitrary data.
The proposal was designed as a temporary measure: enforcement would last approximately one year (52,416 blocks) before expiring automatically. Activation required 55% miner signaling — a threshold notably lower than the 95% convention used in prior Bitcoin soft forks. The proposal contained no FAILED state, meaning it would proceed to mandatory activation at block 965,664 (projected early September 2026) regardless of signaling levels.
The proposal was authored under the pseudonym Dathon Ohm, with Luke Dashjr credited for the original draft and technical advice.
Miner signaling for BIP-110 began on December 1, 2025. Support remained in the sub-1% range through most of 2026:
No major mining pool besides Ocean publicly endorsed the proposal. Foundry, AntPool, ViaBTC, F2Pool, and other pools controlling the vast majority of hashrate did not signal.
At block 961,632 on August 9, BIP-110 nodes began rejecting blocks from non-signaling miners. This created a chain split. Nodes running BIP-110 rules forked onto a minority chain, while the main Bitcoin network continued uninterrupted.
The minority chain mined its first block at height 961,633. Eight hours later, it had produced a second block — and stopped. In the same period, the main chain reached block 961,681, a gap of 48 blocks.
The technical cause was straightforward: the minority chain inherited Bitcoin's full difficulty setting of 127.48 trillion but commanded only approximately 2.5% of the network's hashpower. At that hash rate, each block would take roughly 6.5 hours to mine. The next difficulty adjustment — which could have lowered difficulty to a sustainable level — was estimated to be approximately 350 days away. The minority chain was effectively frozen.
According to CoinDesk, the main chain had advanced by more than 48 blocks while the minority chain produced two, representing a day's worth of activity versus near-zero output.
The core dispute is over Bitcoin's fee market. Ordinals and Runes have materially contributed to miner revenue since their introduction:
For miners, fee revenue from inscription-related transactions is not trivial. Block rewards halve every four years; as the subsidy declines from 3.125 BTC to 1.5625 BTC in 2028, fees become more critical to the long-term security budget. BIP-110 proposed eliminating a category of transactions that, during active periods, accounted for 5% to 10% of total block value.
Mining pools have a direct economic incentive to include any transaction paying competitive fees, regardless of data content. This structural misalignment between BIP-110's goals and miner economics explains the sub-3% signaling rate.
BIP-110 drew public opposition from two of Bitcoin's most prominent stakeholders.
Michael Saylor, co-founder of Strategy (the largest publicly traded corporate Bitcoin holder), published an essay listing 110 objections to the proposal. Key arguments included:
Adam Back, CEO of Blockstream, argued the proposal conflicted with Bitcoin's permissionless design. His position: Bitcoin's strength comes from its resistance to top-down rule changes, and policing what kinds of transactions are "allowed" could be weaponized later for different purposes.
The BIP-110 collapse triggered immediate institutional consequences within Bitcoin's developer governance.
On August 10, 2026, BIP Editor Murch filed a formal motion to remove Luke Dashjr from the BIP editorial board. The allegations:
Dashjr responded that Murch's opinions "alone do not establish new processes" and claimed the Bitcoin development mailing list was "completely captured and corrupt," alleging his replies would "likely be censored."
The dispute reveals a structural tension in Bitcoin governance: the BIP editorial process relies on a small group of volunteer editors with no formal accountability mechanism beyond peer pressure and community norms.
BIP-110's failure has not ended the campaign to restrict inscription data. Developer Chris Guida has rebased Dashjr's 2017 proof-of-work hard fork code onto Bitcoin Knots. The code would change Bitcoin's mining algorithm, effectively creating a new network where existing ASIC miners could not participate.
Dashjr has stated publicly that a PoW change is "the only option" if BIP-110 does not succeed. This mirrors tactics from the 2017 block size wars, when the user-activated soft fork (UASF) threat ultimately pressured miners into accepting SegWit.
The comparison has limits. In 2017, UASF had support from major exchanges, wallets, and a substantial portion of the user base. The BIP-110 faction commands a fraction of that economic weight. A PoW hard fork would require re-bootstrapping Bitcoin's entire mining infrastructure — a multi-billion-dollar proposition with uncertain exchange and custody support.
A separate hard fork called eCash is scheduled to split at block 964,000 (around August 21, 2026), airdropping a new coin one-for-one to every Bitcoin holder. According to AMINA Bank's research division, this constitutes the second fork event in a two-week span — a concentration of protocol disputes not seen since 2017.
BIP-110's trajectory exposes several structural features of Bitcoin governance:
1. Miner veto power is real. Despite the proposal's user-activated design, 97.5% of hashpower declining to signal rendered the fork economically unviable within hours. The difficulty adjustment mechanism — designed to stabilize block times — became an execution barrier for minority chains.
2. Economic consensus outweighs technical consensus. Saylor's argument that 99% of Bitcoin's economic weight rejected BIP-110 proved predictive. The proposal's failure was not primarily a technical judgment but a capital-weighted vote.
3. The 55% threshold backfired. By lowering the activation threshold below Bitcoin's 95% convention, BIP-110's authors implicitly acknowledged they could not achieve broad consensus. This concession became an argument against the proposal, cited by both Saylor and Back as evidence of insufficient support.
4. Inscription economics are entrenched. With Runes generating 25% of transaction fees during active periods and thousands of tokens trading on major exchanges, the economic constituency for inscription-based activity has grown beyond the capacity of a minority soft fork to displace.
BIP-110's failure demonstrates that Bitcoin's consensus mechanism functions as designed: proposals lacking broad economic support cannot override the network's majority. The fork's two-block lifespan was not a bug but a feature — the difficulty adjustment mechanism acting as an immune response against minority chain splits.
The governance fallout, however, is ongoing. The motion to remove Dashjr from the BIP editorial board, the PoW hard fork code rebasing, and the upcoming eCash split suggest the underlying dispute over Bitcoin's purpose — monetary network versus general-purpose data layer — remains unresolved. The inscription economy generates real revenue for miners, and that economic reality is difficult to override through consensus rule changes.
CoinDesk framed the outcome as "free-market capitalism in its purest form." No regulator or central committee rejected BIP-110. Distributed economic consensus did. Whether that same mechanism can address the governance tensions the episode has surfaced — particularly around BIP editorial process and the role of individual developers — is the open question.