On August 10, Luke Dashjr, a long-time Bitcoin Improvement Proposal (BIP) editor, co-founder and CTO of the Ocean mining pool, was removed from the editorial team and lost his editing permissions by his peers. This was due to his actions in pushing BIP-110 by bypassing the discussion process and pre-emptively assigning a proposal number, coupled with his minimal contributions to editorial work in recent years.
The immediate catalyst for this disciplinary action was the first genuine chain split in the Bitcoin network since the Bitcoin Cash hard fork of 2017, which occurred just days earlier. On August 8, the Bitcoin network split at block height 961,632: a subset of nodes refused to follow the main chain, insisting on enforcing BIP-110, a soft-fork proposal prohibiting non-transaction data in transactions, which Luke was deeply involved in drafting. The minority chain produced only one additional block before completely stalling, while the main chain continued operating normally and quickly outpaced it by dozens of blocks.
Who Should Decide the Use of Block Space?
Bitcoin produces a new block on average every ten minutes, packaging a period's transfer records. Once full, miners confirm and link it to the previous block. While blocks mostly contain transaction information, as long as one is willing to pay for the space, the rules do not prohibit including text, images, or even code.
The conflict began with the release of Bitcoin Core 30 in October 2025, the most widely used node software on the Bitcoin network. This version removed the previous 83-byte limit on the OP_RETURN field (used for attaching additional information). This limit was originally just a default recommendation in the software; its removal significantly lowered the cost of including non-transactional data.
However, a segment of the community disagreed with this relaxation. They instead switched to running another software, Bitcoin Knots, which is free and open-source but retains the old limit. Its user share grew from nearly zero in early 2024 to over 22% within two years.
It was against this backdrop that BIP-110 was proposed, aiming to codify Knots' restriction into Bitcoin's core consensus rules, making it mandatory for the entire network. Initially numbered BIP-444, the draft mentioned imposing "legal and moral consequences" on non-compliers, which sparked backlash and was later removed. In December 2025, it was renumbered to 110.
What started as a technical discussion thus escalated into a power struggle.
The Trust Gambled on a Low Threshold
For a Bitcoin soft fork to activate, it typically requires miners (individuals or organizations running specialized equipment to record transactions for the network and earn Bitcoin rewards) to vote by signaling: inserting a marker in their blocks. The new rule only activates once the signaled hash rate reaches a threshold. The historical convention is a high threshold of 95%, requiring near-unanimous miner approval to avoid a chain split.
BIP-110, however, lowered this threshold to 55% and added a forced activation clause: even if the support rate is not met, nodes running the patch will automatically reject blocks that do not comply with the new rule.
While the threshold can be written into code, whether miners will cooperate is not dictated by code.
Since entering the voting period on May 1, support never exceeded 3%; in the final statistical period nearing the forced activation deadline, support only climbed to about 2.53%.
Major mining pools like Foundry USA and AntPool did not signal support. Support votes came almost exclusively from the Ocean pool and some independent miners. F2Pool co-founder Wang Chun even publicly criticized Luke, sarcastically remarking he was "both financially and personally bankrupt" and suggesting he might as well switch to a different proof-of-work algorithm, with the results likely being no better.
Strategy founder Michael Saylor listed "110 reasons" in opposition, arguing that once rules can filter transactions based on content, Bitcoin's neutrality is broken. Security expert Jameson Lopp was more direct, calling it "reckless" and "doomed to fail," noting that data could be encoded differently to bypass the restriction and that it could render certain transaction outputs permanently unspendable.
A Fork That Lasted Only One Block
On August 8, when the chain reached block height 961,632, nodes running the BIP-110 patch rejected this block because it did not carry the signal required by the new rule. They instead created a new block according to their own rules, initiating the split. This new chain initially had the support of about 2.53% of the network's total hash rate—equivalent to roughly 1 out of every 40 mining machines worldwide being willing to mine on it.
However, the minority chain only produced one additional block before it stopped updating entirely. The chain with more hash rate moves faster and farther. The main chain consistently produced a block every ten minutes, leaving the minority chain behind by 243 blocks (data from bip110.mempool.guide).
Japanese exchange bitFlyer's announcement on August 10 merely stated they would continue monitoring the impact of BIP-110 without committing to a specific handling method. Furthermore, no major exchange expressed support for the minority chain, a stark contrast to the guides many issued ahead of the 2017 Bitcoin Cash fork.
Those who truly need to be cautious are self-custody nodes running Bitcoin Knots: BIP-110 lacks "replay protection," meaning the same transaction might be valid on both chains, potentially leading to accidental double-spending if not handled carefully.
The miner "Roughnecks," who mined the block on the BIP-110 chain, initially stopped production but reversed course on August 10, announcing plans to resume mining. However, by then, their signaled hash rate had plummeted from over 15 EH/s to 1.16 EH/s, and no new blocks appeared on the chain. The economic cost of continuing to mine had already exceeded the potential rewards.
But supporters don't seem ready to concede. Proposal author Dathon Ohm and Luke, who just lost his editor status, have indeed begun discussing the path Wang Chun mocked. That is, changing the proof-of-work algorithm to shake off miners using specialized hardware (ASICs) and turning the minority chain into a separate new cryptocurrency. This currently remains in the discussion and code experimentation phase.
Regardless of the eventual outcome, this whole episode confirms the principle etched into Bitcoin from its inception: rules can be drafted by a few, and editing permissions can be granted by oneself, but the real decision-makers are the miners willing to keep burning electricity for a chain, and the exchanges and users who decide which chain counts.
