On March 7, 2026, the Bitcoin mainnet experienced a two-block chain split lasting eight hours. The cause was not a 51% attack or a code bug. It was a governance failure embedded in a soft fork proposal that never reached consensus. The proposal—BIP-110—aimed to limit arbitrary data in Bitcoin transactions, directly targeting the Ordinals inscription craze that had inflated block space demand since 2023. But the reaction was not a graceful upgrade. It was a 2.53% signaling rate, a forced node rule, and a temporary network partition.
This is the story of how a single BIP editor's push to enforce a new rule nearly fractured the network. And it is the story of how the community's response—removing that editor—may have saved Bitcoin from a more dangerous precedent.
Context: The Ordinals Trigger and the BIP-110 Response
To understand the chain split, you must first understand the Ordinals phenomenon. Since the introduction of the Ordinals protocol in early 2023, users have been inscribing arbitrary data—images, text, even entire files—directly into Bitcoin transactions. This is not a technical exploit; it is a feature of the existing OP_RETURN and witness data fields. But it has fundamentally changed the economics of Bitcoin block space. In 2023, block space demand surged, pushing fees higher and creating a new class of non-financial transactions.
For many Bitcoin purists, this was an attack on the network's original purpose: a peer-to-peer electronic cash system. For others, it was a legitimate use of the protocol's capabilities. The debate is not new; it echoes the 2017 block size war. But the response this time was different. Instead of a hard fork, a group of developers proposed a soft fork: BIP-110.
BIP-110's technical specification is straightforward: limit the amount of arbitrary data that can be included in a transaction's witness data. This would effectively cap the size of Ordinals inscriptions, making them economically unviable. The proposal was assigned BIP number 110 by Luke Dashjr, the long-time BIP editor and Bitcoin Core contributor. Dashjr was not just an editor; he was an advocate. He pushed the proposal through the BIP process with unusual speed, assigning a number before the usual community review period had concluded.
The BIP process is not a legislative body. It is a standardization mechanism. Editors like Dashjr have the power to assign numbers and merge pull requests. That power is based on trust, not authority. The assumption is that editors will act as neutral stewards, not as partisan advocates.
Core: The Technical Failure and the Governance Flaw
BIP-110's activation mechanism required miners to signal support in their coinbase transactions. The threshold was 55% of blocks within a difficulty period. The actual support peaked at 2.53%. That is not a rounding error. It is a rejection.
Yet the software implementation of BIP-110 contained a controversial feature: nodes running the proposal would reject any block that did not signal support for the soft fork. This is the equivalent of a restaurant refusing to serve customers who don't wear a tie, even though the dress code has not been agreed upon. At block 961632, a pool of nodes—likely operated by a small group of developers and miners—began rejecting blocks that did not include the BIP-110 signaling bit. The result was a chain split. For eight hours, the main chain forked into two: one following the majority hash rate that did not signal, and one following the minority that enforced the rule. Only two blocks were mined on the minority chain, but the damage was done. The network was partitioned.
From a technical perspective, this is a textbook example of a forced node rule. The risks are well-documented: chain splits, loss of funds, and erosion of trust. I have seen similar patterns before. In my 2020 DeFi liquidity stress test, I modeled how a single protocol change could create a liquidity bifurcation. The mathematics is the same: when a subset of nodes enforces a rule that the majority does not accept, the network splits. The only difference is scale.
The immediate trigger was the BIP editor's decision to merge the code into a Bitcoin Core release candidate. But the deeper governance flaw is the lack of a formal process for removing a BIP editor who acts in a partisan manner. Luke Dashjr was not removed because he broke a rule. He was removed because he broke an unwritten norm: editors should not use their position to push their own proposals.
Contrarian: The Removal of Luke Dashjr is Not a Victory for Decentralization
The narrative that emerged after the chain split was clear: the community spoke, the editor was removed, and Bitcoin's decentralized governance worked. That is a comforting story. It is also incomplete.
The removal of Luke Dashjr as BIP editor sets a dangerous precedent. Yes, it corrected the immediate problem. But it also signals that any BIP editor who pushes a controversial proposal risks being fired. That is a chilling effect. In a system that relies on volunteer contributions, the threat of removal can discourage exactly the kind of innovation Bitcoin needs.
Consider the alternative: what if the community had allowed the proposal to fail on its own? The 2.53% signaling rate was already a clear rejection. The chain split was caused not by the proposal itself, but by the forced node rule. The more appropriate response would have been to revert the code change and reaffirm that node operators must not enforce unactivated rules. Instead, the community removed the person. That is a personnel solution to a structural problem.
In my 2022 bear market exit protocol, I learned that the most dangerous failures are not the ones that break the system, but the ones that are fixed so quickly that the underlying flaw is never addressed. The chain split was resolved in eight hours. The governance flaw remains.
Exit strategies are written in ice, not in hope. The removal of a BIP editor may feel like a victory, but it is a temporary fix. The real question is whether the Bitcoin community can develop a formal governance process that prevents such situations without resorting to ad hoc removals.
Takeaway: The Next Cycle Will Test This Again
Bitcoin's value proposition is built on immutability and consensus. But consensus is not a vote; it is a state of deployed code. The BIP-110 incident shows that the line between enforcement and consensus is thinner than most investors realize.
In the next bull market, when block space congestion rises again, similar proposals will emerge. The temptation to use a BIP editor's power to fast-track a controversial fix will be strong. The institutional investors who are now allocating billions to Bitcoin ETFs must understand that this governance risk is not priced in. The market's memory is measured in blocks, not years.
Standardization is not just about code. It is about the process that produces the code. The BIP-110 chain split is a wake-up call, not a death knell. But it is a call that must be answered with structural reform, not just a personnel change.
The question is not whether Bitcoin can survive a chain split. It can. The question is whether its governance can evolve to handle the next wave of protocol-level disputes without breaking the trust that makes it a macro asset.
In open source, governance is the only protocol that cannot be forked.