On March 8, Bitcoin’s main chain fractured for eight hours. Two blocks. A few hundred transactions orphaned. The network healed itself — no funds lost, no permanent split. But the fault line was not a bug. It was a governance choice. A client running a modified consensus rule decided to reject blocks that didn’t signal support for BIP-110. The chain split. And then the BIP editor who pushed that proposal was removed from his role. Most coverage focuses on the politics: Luke Dashjr, the long-time Bitcoin Core contributor, removed by a vote of his peers. They frame it as censorship, as a power struggle between the “purists” and the “innovators.” They miss the technical reality. The chain doesn’t care about your feelings. It cares about consensus. And this event exposed a fundamental flaw in how Bitcoin’s governance operates — one that could lead to a real split next time.
Context: Why BIP-110, and Why Now The Ordinals inscription craze of 2023 pushed arbitrary data into Bitcoin blocks. Transactions that normally carried a few dozen bytes of op_return data now carried entire images. Block sizes swelled. Fees for regular users spiked. Some developers saw this as an existential threat to Bitcoin’s peer-to-peer cash vision. BIP-110 was their answer: a soft fork that would limit the amount of arbitrary data a transaction could write. It was a conservative move — a restriction, not an expansion. But the method of its introduction was anything but conservative. Luke Dashjr, as the BIP editor, assigned the number BIP-110 and merged the pull request into the Bitcoin Core repository before the broader community had reached anything close to consensus. He then released a version of Bitcoin Core that included the logic to enforce the rule — but only if miners signaled support. The problem? The logic was active from the start. The client would reject blocks that did not signal, even if the signal threshold was not met. This was not a waiting implementation. It was a loaded gun.
Core: The Technical Autopsy of the Eight-Hour Split On-chain data tells the story. Block 961632. The chain diverged. The minority chain — running the BIP-110 client — produced two blocks that were rejected by the majority chain. The majority chain produced multiple blocks that were rejected by the minority. For eight hours, two versions of Bitcoin existed simultaneously. The immediate cause is clear: the BIP-110 client implemented a rule that said “if a block does not signal support for this BIP, reject it.” This is a classic test of economic majority. The majority chain won. The BIP-110 chain died. But the fact that it happened at all is a red flag.
Let’s examine the numbers. BIP-110’s signal support topped out at 2.53% of blocks. The activation threshold was 55%. Even if every miner who signaled had actually upgraded their node, the proposal was nowhere near consensus. The client that enforced the rejection was running on a tiny fraction of the network. Yet it was willing to fork off. This is not a healthy property for a decentralized system. “The chart doesn’t care about your BIP number,” I’ve written before. The network’s economic majority will always choose the chain that minimizes disruption. In this case, the disruption was minimal. But the precedent is dangerous.
Now, the removal of Luke Dashjr as BIP editor. The narrative from his supporters: he was fired for trying to protect Bitcoin from Ordinals. The reality is more technical. The removal was not about the content of BIP-110. It was about the process. Dashjr assigned the BIP number, merged the PR, and released the client code without waiting for the normal consensus-building process. The Bitcoin Improvement Proposal process is not a rubber stamp. It is a social contract. The BIP editor is a gatekeeper, not a dictator. By fast-tracking a contentious proposal and releasing code that could cause a chain split, Dashjr violated that contract. The other maintainers removed him to preserve the integrity of the process. Speed is safety when the exploit is already live. But when the change is a soft fork, speed is a liability.
Contrarian: The Unreported Blind Spot The mainstream crypto media loves a good villain story. Luke Dashjr is cast as the martyr. The narrative: “Bitcoin developers silenced a voice against Ordinals.” This is wrong. The real story is about a governance failure that could have been catastrophic. The BIP-110 client did not wait for consensus. It enforced a rule that the network had not agreed to. If the economic majority had been smaller — say, 40% instead of 2.53% — the chain split would have been permanent. Bitcoin would have forked into two assets. That is not a “soft fork.” That is a chain split triggered by a single client’s decision to go live before the community was ready.
“We don’t use the word ‘decentralized’ until we’ve seen the code and the governance,” I often say. The code here was clear: the client had a hard-coded rule that forcibly rejected non-signaling blocks. The governance was not clear: the BIP editor bypassed the normal process. The combination is toxic. The removal of Dashjr was not censorship. It was a corrective action by the maintainers who realized that the network’s security was at risk. The contrarian truth is that the real enemy is not Ordinals or the “censorship” of a BIP editor. It is the creeping assumption that a single developer, or a small group, can impose a consensus change through code alone. Bitcoin’s strength is that it requires social consensus before code consensus. The BIP-110 incident was a violation of that principle.
Takeaway: The Next Fracture The eight-hour split was a warning shot. The next time, it might not heal. The takeaway is not about the fate of Luke Dashjr. It is about the need for a more robust governance layer in Bitcoin. The BIP process is informal. It relies on the judgment of a few editors. When one editor oversteps, the network is vulnerable. The solution is not to remove the editor — that treats the symptom. The solution is to formalize the process so that no single person can push a controversial change to the brink of a chain split. The question every Bitcoin holder should ask: who watches the watchmen? If the answer is “no one,” then the network is only as safe as its weakest governance link. The next BIP-110 will come. Will the chain survive?