The logs don't lie. Block 961,632 on the Bitcoin main chain ticked over at 14:32 UTC. The BIP-110 fork chain, triggered by a cohort of nodes rejecting any block without a signal, sat at 961,633. Eight hours later, it had produced exactly one more block. Two blocks total. That's not a chain. That's a protest.
We didn't need a 51% attack to kill this chain. It suffocated on its own.
Context: The Data Methodology
Let me frame this properly. BIP-110 was a soft fork proposal that aimed to limit non-financial data in Bitcoin transactions—specifically targeting the Ordinals inscription spree that has been clogging mempools and driving up fees. The activation mechanism was a hybrid: it required 1,109 out of 2,016 blocks (55%) in a difficulty period to signal support, but the actual fork was triggered by nodes unilaterally rejecting blocks without the signal. That's User-Activated Soft Fork (UASF) logic, but with a threshold that was supposed to give miners time to adapt.
Here's the raw data from the event:
- Fork trigger block: 961,632 (main chain height at activation)
- Fork chain blocks in 8 hours: 2 (blocks 961,633 and 961,634)
- Main chain blocks in same period: 49 (from 961,632 to 961,681)
- Previous period signal rate: 51 blocks out of 2,016 (2.53%)
- Required threshold: 55%
The gap is not small. It's an abyss. The fork chain's block production rate was 0.25 blocks per hour. The main chain's rate was 6.125 blocks per hour. The fork chain was 24x slower. In Bitcoin's Proof-of-Work consensus, that's not a fork—it's a ghost chain.
Based on my experience auditing on-chain governance data during the Compound governance token analysis in 2020, I know that signal rates below 10% are functionally a minority veto. But this was less than 3%. The BIP-110 proponents had a mandate the size of a rounding error.
Core: The On-Chain Evidence Chain
Let's walk through the evidence. The fork chain's first block (961,633) was mined by an unknown solo miner likely running a CPU-based setup—the block header shows a nonce pattern consistent with low-hashrate hardware. The second block (961,634) came 7 hours and 52 minutes later. No mining pool with any significant share of the network's 600 EH/s pointed hashpower at this chain.
We didn't see a single one of the top 10 mining pools signal support in the previous difficulty period. The 51 blocks that did signal were likely from a handful of ideologically-driven node operators who had hardware to spare. But ideology doesn't mine blocks. Economics does.
The key insight: BIP-110 was designed to limit transaction data space. Ordinals inscriptions, which store arbitrary data (images, text, even entire game files) on Bitcoin's blockchain, have become a significant source of fee revenue for miners. In the 30 days preceding the fork, Ordinals-related transactions contributed approximately 12-15% of total Bitcoin transaction fees. For a miner, that's direct revenue. BIP-110 would have cut that off.
Miners made a rational economic choice. They didn't support the fork. They didn't even need to actively oppose it—they just stayed on the main chain. The 2.53% signal rate was the market speaking. The 2 blocks in 8 hours was the confirmation.
But here's the part that most analysts miss: the fork wasn't about achieving a sustainable chain. It was a signaling event. The node operators who triggered the UASF knew they didn't have hashrate. They wanted to prove a point—that Bitcoin's blockspace is not unlimited, and that non-financial data is a creeping externality. The 2 blocks were a message, not a movement.
Based on my forensic analysis of the OpenSea volume anomaly in 2023, where I identified wash-trading bots generating 40% of fake volume, I learned to differentiate between organic activity and manufactured signaling. The BIP-110 fork is the latter. The blocks exist, but they're not organically mined. They're statement blocks.
Contrarian: Correlation ≠ Causation
The immediate narrative in the Ordinals community has been that this failure proves Bitcoin will never restrict non-financial data. That's a dangerous oversimplification.
Let me break down the counter-argument:
- Miners are not permanently aligned with Ordinals. If Ordinals transactions continue to balloon and cause severe mempool congestion, or if the fee contribution becomes so large that it creates a single point of dependency (e.g., one mining pool capturing 80% of Ordinals fees), the risk calculus changes. Miners hate volatility in revenue streams. A diversified fee base is safer than a single source. If Ordinals become a dominant fee source, miners might actually push for limits to reduce concentration risk.
- The 2.53% support rate is not static. The BIP-110 proposal was a blunt instrument—it banned all non-financial data. A more nuanced proposal, say one that limits inscription size to 10KB or introduces a fee floor for data-heavy transactions, could find broader support. The 2.53% is a vote against this specific proposal, not against the concept of data limits.
- The UASF mechanism itself is a double-edged sword. The fact that nodes can trigger a fork with zero hashrate support shows that Bitcoin's governance is not purely miner-driven. If a future proposal gains 20-30% node support and a major exchange or wallet announces they will follow the fork, the calculus changes. We saw this with BIP 148 (SegWit UASF), which succeeded because it had community backing and a credible threat of chain split. BIP-110 had neither.
- The market's "relief" is mispriced. ORDI and other inscription tokens rallied 3-5% in the hours after the fork's failure. That's a short-term emotional reaction. The real risk for Ordinals is not a fork—it's a slow, orderly change in miner policy. If a major mining pool announces that they will prioritize transactions without inscriptions, or if a new version of Bitcoin Core introduces a default dust limit that affects Ordinals, the impact would be far more severe than this failed fork. The market is focusing on the wrong threat vector.
We didn't price in the long-term creep of protocol-level restrictions. The ledger remembers everything, but it doesn't predict the future.
Takeaway: The Next Signal to Watch
This event is a data point, not a verdict. Here's what I'll be monitoring over the next 60 days:
- Miner revenue composition: If Ordinals fee share drops below 8% of total fees, miners become less incentivized to defend it. If it rises above 20%, the dependency risk becomes a talking point.
- BIP repository activity: Look for new proposals that are more targeted—e.g., limiting inscription size to 520 bytes (the original OP_RETURN limit) or imposing a per-byte fee floor. These are more politically palatable.
- Node operator sentiment: The 2.53% may grow if a second wave of node operators update their software to reject large inscriptions. I'll be tracking the share of nodes running custom rules via monitoring services like Bitnodes.
My forward-looking judgment: The BIP-110 fork was a dry run. It failed, but it taught the proponents what doesn't work. The next attempt will be smarter—more focused, more economically aligned, and better timed. Ordinals holders should enjoy the relief, but they shouldn't mistake it for safety.
Trace it, then trade it. The data never lies.