How a failed soft fork exposed the real mechanics of Bitcoin’s governance — and why miners said “no.”
BIP-110
Bitcoin doesn’t have a clear “CEO, a board, or a shareholder vote”. So when people say Bitcoin’s rules changed, what they usually mean is: a small group of developers writes code, miners either run it or didn’t, and the market watched to see whose incentives “wins”. BIP-110 just gave us a clean, real-time demonstration of exactly how that works — and as someone who spends his time engineering
tokens, I think it’s one of the best case studies we’ve had in years.
Here’s the story, and why it matters way beyond Bitcoin maximalist Twitter……
What BIP-110 actually proposed
BIP-110’s Proposition
BIP-110, officially the “Reduced Data Temporary Softfork,” was written by a pseudonymous developer going by0 Dathon Ohm, with Bitcoin Core contributor, Dashjr credited for the original draft and technical input. It was assigned in December 2025 and reached “Complete” status on the official Bitcoin BIPs repository by June 2026.
The proposal was narrow on paper: for about one year, it would have capped new transaction outputs at 34 bytes, limited OP_RETURN payloads to 83 bytes, and restricted certain Taproot witness data. Anything created before activation would stay exempt forever — this really wasn’t retroactive, neither does it ever named its target directly, but technically everyone knew it was aimed at Ordinals inscriptions, BRC-20 tokens, and Runes ( the protocols that let people attach images, text, and token balances to individual satoshis using space that Taproot made cheap to fill). To one camp, that’s spam clogging up a payments network. To another, it’s simply people paying fees for valid transactions, and Bitcoin itself doesn’t get to have opinions about what a transaction is for.
So far, this is the interesting part; it’s failure. But interesting?
BIP-110
Yes, Bitcoin’s biggest protocol changes have historically needed something close to unanimous miner buy-in. Taproot activated with well over 90% signaling support. BIP-110 asked for a comparatively modest 55% threshold and hence, yet it couldn’t get there. By the time its signaling window approached in early August, support sat under 1%.
Michael Saylor came out hard and strong against it, publishing a lengthy essay laying out dozens of objections. His core argument wasn’t really technical rather constitutional. He argued that once Bitcoin’s consensus rules start policing which valid transactions are acceptable based on their purpose, you’ve created a precedent that can be pointed at privacy tools, stablecoin settlement, or anything else a future majority decides it doesn’t like. Adam Back and JAN3’s Samson Mow pushed back too, and Bitcoin Core developers themselves declined to fold the change into the reference client.
That near-total rejection is exactly what token engineers should be paying attention to. This wasn’t a failure of the code — the specification was technically sound. It was a failure of the mechanism design around how the change was supposed to get adopted.
Ask me from the token-engineering lens: this was a coordination game, and the game was rigged against consensus.
Try stripping away the Bitcoin-specific vocabulary and BIP-110 is a textbook multi-stakeholder coordination problem. You’ve got miners (who earn fees from the exact transactions being restricted), node operators (who bear the cost of storing and relaying that data), long-term holders (who care about money), and builders on Ordinals/BRC-20/Runes (who have real economic activity riding on the status quo). Each group has different, partially conflicting incentives, and there’s no formal voting mechanism to reconcile them — only a norm that “rough consensus” across all these groups is required before miners will even bother signaling.
This is where the economics gets genuinely deep. University of Chicago economist Eric Budish’s well-known 2018 paper on Bitcoin’s security model, which updates the model for proof-of-stake systems (worth noting that Budish later revised and extended this work into a peer-reviewed piece published in the Quarterly Journal of Economics 2024, co-authored with Andrew Lewis-Pye and Tim Roughgarden) too shows that Bitcoin’s whole security guarantee rests on a delicate balance: miners must earn enough in ongoing rewards that attacking the network is never worth more than playing along honestly. Ordinals and BRC-20 traffic isn’t just noise — it’s real fee revenue that helps sustain that balance, especially as the block subsidy keeps halving. Restricting that traffic without a plan to replace the fee income isn’t just a philosophical question, it’s a security-budget question. That’s likely a big part of why miners had zero appetite to lock it in.
Moreso, there’s also a governance angle worth sitting calmly with. Legal scholar, Angela Walch has argued for years that Bitcoin’s “decentralized” branding functions as a kind of liability shield — in practice, a small set of core developers and large miners hold outsized influence over what counts as a valid rule change, even though nobody formally elected them to that role. BIP-110 is a live example: a pseudonymous author and a handful of prominent developers and businesses effectively decided whether to try reshaping Bitcoin’s block space economics for everyone, and the market’s answer was a near-unanimous no. That’s governance happening in public, in real time, without a vote — which is either beautifully emergent or deeply opaque depending on which side of the debate you’re on, ask yourself.
What this means if you build or write about tokens
A few takeaways I keep coming back to:
Fee markets are tokenomics, whether or not a chain calls it that. Bitcoin doesn’t have a “protocol revenue” line item, but Ordinals and BRC-20 activity function exactly like a demand-side subsidy for miner security. Any proposal to restrict that demand is, functionally, a proposal to cut protocol revenue — and should be modeled that way.
Soft forks are a governance mechanism, not just a technical one. The 55% threshold BIP-110 needed wasn’t arbitrary , it’s a coordination bar meant to approximate rough social consensus. When a proposal can’t clear even a lowered bar, that’s the market pricing doing the work itself in disagreement about legitimacy, not just about bytes.
“Credible neutrality” is a real design constraint, not a slogan. Saylor’s objection — that policing transaction purpose breaks Bitcoin’s promise of treating all valid transactions the same — is the same principle we (serious protocol designers) reach for when we build permissionless systems: rules should apply to what a transaction does, not why someone sent it.
BIP-110 is probably going to fade from the headlines fast, especially with no activation and a market that’s already moved on. But as a case study in what happens when tokenomics, security-budget economics, and informal governance collide in public, it’s about as good as it gets. I’d bet we’re still citing it in five years.
If you’re building or researching token systems and want to dig into the sources: Eric Budish’s “The Economic Limits of Bitcoin and the Blockchain” (NBER Working Paper No. 24717, 2018) and Angela Walch’s “In Code(rs) We Trust: Software Developers as Fiduciaries in Public Blockchains” (in Regulating Blockchain: Techno-Social and Legal Challenges, Oxford University Press, 2019) are both worth the full read.
BIP-110 and The political Economy of Bitcoin’s “Non-Monetary” Data was originally published in Coinmonks on Medium, where people are continuing the conversation by highlighting and responding to this story.