BIP-110 Is Dead. What Its Two-Block Fork Proved About Who Really Controls Bitcoin

No CEO stepped in. No regulator ruled. A proposed change to Bitcoin's rules asked the network for permission, the network declined, and by Sunday the whole thing was over.

A retrospective on the year-long spam war, the weekend a rival chain died of arithmetic, and what Luke Dashjr and Matthew Kratter are saying now. The spam debate was never the interesting part. How it got settled was.

Block 961,632 arrived on Saturday, August 8. For a small, fierce group of Bitcoiners, this was the moment everything had been building toward. Months of arguing, a countdown, mandatory signaling. At that height, nodes running the BIP-110 rules would begin rejecting any block that failed to follow their rule changes. The line in the sand was finally drawn.

Then the chain mined one more block, 961,633, and stopped.

Not paused for a few minutes. Stopped. While the BIP-110 chain sat frozen, the Bitcoin the rest of the world runs kept humming along, block after block, until it was dozens ahead and pulling away for good. By the end of the weekend the breakaway chain had produced a grand total of two blocks. The mining pool carrying it, Roughnecks, posted a note that it was pausing operations. A movement that had spent a year insisting it could reclaim Bitcoin had, in practice, mined about twenty minutes of it.

Here is the thing worth sitting with before we assign winners and losers. That outcome was not a bug, a hack, or a censorship event. Nobody unplugged anything. It was Bitcoin's governance doing exactly what it was built to do, in public, at full volume. So let's walk back through how we got here, because the mechanism is more interesting than the drama, and the drama is genuinely good.

Rewind twelve months

You see, this fight is old. Older than the acronym by years.

For most of a decade, Bitcoiners have argued about "spam": stuffing arbitrary data (images, text, tokens, the occasional JPEG of a cartoon) into transactions rather than using the chain purely to move money. One camp sees this as parasitic. Every byte of someone's picture has to be stored and verified forever by every node runner on Earth, quietly raising the cost of participating and dragging Bitcoin away from its purpose as money. The other camp shrugs. You paid the fee, you bought the block space, the network is permissionless, and "permissionless" does not come with an asterisk that reads "unless we dislike your bytes." Not to mention that all of these fads fade a away and most people don't see them as a long term threat.

Reasonable, principled people land on both sides of that.

BIP-110, formally titled "ReducedData" and published in February 2026, was the anti-spam camp's most serious attempt to win the argument in code instead of on Twitter. It proposed a temporary set of restrictions, one year long, that would sharply limit the arbitrary data a valid transaction could carry. Championed by Luke Dashjr, the long-time developer behind the Knots node software, and authored under the name Dathon Ohm, it was structured as a user-activated soft fork with a 55% miner signaling threshold. Hit 55%, and it activates smoothly. Miss it, and the design still forced a "mandatory signaling" showdown at block 961,632, where BIP-110 nodes would start rejecting non-compliant blocks whether the miners liked it or not.

It was, in plain terms, a game of chicken. The bet was that miners, staring down the threat of having their blocks orphaned, would blink and signal rather than risk a split. A rerun, the organizers hoped, of 2017's BIP-148, the scrappy user-activated soft fork that famously bent the mining industry to the users' will and is still celebrated in some circles as Bitcoin's "Independence Day."

That was the theory. Now the arithmetic.

The weekend the fork died of math

In the two weeks before activation, BIP-110 signaling limped in at roughly 2.5% of blocks. Not 55%. Not 30%. About one block in forty. Prediction markets had priced it as a near-certain failure for weeks, and the miners, overwhelmingly, simply declined to play.

When block 961,632 hit, the mandatory-signaling logic did precisely what it promised. Nodes running Knots with BIP-110 enforced began rejecting the ordinary chain's blocks, and a separate chain split off, carried by the sliver of hashrate that backed the rules. The Ocean and Roughnecks pools mined 961,632 and 961,633 on that new chain.

A new chain doesn't get to start on easy mode. It inherited Bitcoin's current mining difficulty, a punishing figure around 127 trillion, calibrated for the entire network's horsepower. But only about 2.5% of that horsepower had come along. Bitcoin re-tunes its difficulty every 2,016 blocks to keep new blocks arriving roughly every ten minutes. On the main chain, that next adjustment is about two weeks out. On the BIP-110 chain, at the pace of two blocks in a weekend, reaching the same adjustment would take something on the order of 350 days. Nearly a year of near-frozen blocks before the network could even catch its breath.

No rational miner burns real electricity for that. Even hashrate with historical sympathy for the cause voted with its feet: Simple Mining, a pool affiliated with Ocean, went and mined on the ordinary chain instead. The breakaway wasn't censored or attacked. It was abandoned, because staying was a guaranteed way to set money on fire.

That is the whole event. A rule change asked the network to adopt it. The network, meaning the people who actually run the software and point the machines, said no. And there was no committee to appeal the decision to.

Hayek wrote the postmortem in 1945

Let me put a name to what happened, because an economist described this machine long before Satoshi wired it together.

Friedrich Hayek spent his career on a single subversive idea: that the most important kinds of order in human life are not designed by anyone. Language, common law, market prices, none of it was drafted by a central planner and handed down. It emerged, from the bottom up, out of millions of people acting on their own local knowledge and coordinating through a few shared rules, with no one of them in charge of the whole. He called it spontaneous order, and he drew a hard line between an order that is grown and an order that is merely made.

His warning was aimed squarely at the planners who believed a clever enough group could sit at the top and design a better order by decree. He named that belief the fatal conceit. The curious task of economics, he wrote, is to demonstrate to men how little they really know about what they imagine they can design.

BIP-110 was, and I say this with real respect for the conviction behind it, a small brush with the fatal conceit. A group certain it knew what Bitcoin should be tried to impose that vision from the top, through an ingenious bit of forcing logic, onto a system whose entire nature is that nobody sits at the top. The network did not route around the rule because the underlying worry was stupid. It routed around it because that is what a spontaneous order does to a decree it did not choose. The rule had to be adopted voluntarily, node by node and miner by miner, or it was never a rule at all. It was a very well-argued opinion running on 2.5% of the hashrate.

And this is where the easy narrative breaks, so I'll say it plainly. The anti-spam camp is not foolish and its concern is not fake. As one Bitcoiner put it in the wreckage, being against spam was the right instinct, and BIP-110 was still a bad idea. You can hold both at once. You can believe the chain is cluttered with junk and still believe that no developer, however brilliant, gets to unilaterally reach in and rewrite the rules for everyone else. Holding both is arguably the most Bitcoin position on the board.

What Luke and Kratter are saying now

Here is where a normal organization would issue a press release, run a postmortem, and go back to work. Bitcoin has none of those things. So instead it has a very loud, very public argument, and the argument is the most revealing part of the whole affair.

Luke Dashjr is not treating this as finished. Before activation he had already said the quiet part out loud: if BIP-110 did not pan out, in his words, the only option was a proof-of-work change, a hard fork that would swap out Bitcoin's mining algorithm entirely and, not incidentally, brick every existing miner on the network. A developer named Chris Guida had already rebased Dashjr's old 2017 proof-of-work fork code onto Knots days before activation, describing it as something to keep in the back pocket in case the miners "betray bitcoin." Since the failure, Dashjr has floated targeting September 1, the date BIP-110 would have activated, for exactly that fork, and told supporters to hold tight while the group works out how to move forward.

Matthew Kratter, the Bitcoin University educator who went so far as to add "#BIP-110" to his display name, spent Monday morning narrating the aftermath to his audience with unusual candor. He called the stalled chain "the Bitcoin chain in a coma," one of the saddest things he had ever watched, and then reframed it as "seeds of hope." His read of Saturday is worth quoting fairly: what shocked him was "the failure of the mining pools to obey the wishes of the node runners," the fact that not a single major pool besides Roughnecks flipped its bit. To him that smelled less like a market and more like a backdoor deal among a "mining cartel." Roughnecks itself struck the same defiant note, announcing that it did not see the pause as a defeat "but an escalation to the next step," and then, a day later, that it would resume mining the stalled chain and keep going until a proof-of-work change could "replace the centralized mining pools that left." They called it wildcatting.

Kratter's real argument is not "we lost the vote." It is that the game itself has quietly changed since 2017. Back then, he notes, a minority of users with "no hash rate, no exchanges, no lobbyists" forced the industry to back down, armed with nothing but conviction and nodes. What grew up in the years since is the problem, as he sees it: spot ETFs, treasury companies, and custodians holding millions of coins for people who will never run a node, alongside hashrate concentrated into a handful of pools writing block templates for the whole world. In that telling, the 2.5% figure is not proof that users chose freely. It is proof that the machinery of choice has been captured by intermediaries, and that the coins with the most weight are precisely the ones sitting in accounts whose owners cannot vote with them. Dathon Ohm went further still, accusing the large pools of colluding to turn Bitcoin "from money into a toxic data dumping ground."

That worry about capture is not nothing. Pool concentration is real. Anyone who cares about Bitcoin as a sovereignty tool should take the concern seriously rather than dunk on it.

But watch where the logic lands, because it is telling. Faced with a system they believe is captured, the BIP-110 faithful are not asking a regulator to intervene or a court to rule. They are invoking the technologist Balaji Srinivasan's old distinction between "voice" and "exit," reform from within versus leaving to build something new, and Kratter is leaning hard toward exit: fork off, change the proof-of-work, and let a smaller, uncaptured chain carry the flame. "Legacy Bitcoin," he says, is no longer his freedom project. In his own blunt phrase, "in my mind, it is now stock." Others are less sure. Podcaster Guy Swann pushed back that swapping the proof-of-work does not "fire the miners" so much as launch a brand-new coin. I agree that that is exactly what they are doing.

Here is the quiet irony underneath all of it. The recourse they are reaching for, seceding to a chain they prefer, is only possible because Bitcoin permits precisely that kind of dissent. The same openness that let a determined minority split off on Saturday is the openness they are counting on to build their life raft now. The system they are calling captured just handed them the exact tool a captured system would never allow: the freedom to leave.

Then, predictably, the fight turned inward. On Monday, August 10, Mark "Murch" Erhardt, one of the editors who maintains the BIP repository, formally recommended removing Luke Dashjr as a BIP editor, citing a broken editorial process and what he called a complete departure from the Bitcoin development ecosystem. Dashjr fired back that the accusations were false, that he had followed the process for years, and that the mailing list itself was "completely captured and corrupt." He countered by proposing Erhardt's removal instead. Notice what no one in that exchange can do: win it by fiat. There is no board that votes, no CEO who decides, no regulator who rules. There is only the same permissionless process that produced the weekend's outcome. Talk all you want, then go convince people to run your software. Everything else is noise.

Meanwhile Michael Saylor, who had spent the summer insisting Bitcoin "still has no spam problem," offered a four-word verdict on the whole episode: "Bitcoin worked exactly as designed." And Adam Back picked apart the movement's central game-theoretic bet, that miners would be forced to switch, calling it flatly wrong, and noting that any user genuinely worried about the fork could neutralize it with a single command on their own node.

Who was actually in control the whole time

So who runs Bitcoin? The weekend answered it about as cleanly as it will ever be answered. Not the miners, who mostly ignored the whole thing. Not the developers, one of the most respected of whom just watched his preferred rules draw 2.5% support and now faces removal from his own editor seat. Not the loudest accounts on Twitter, whichever tag they added to their names.

The answer is the boring one, and it is the entire point: the rules are enforced by the thousands of independent nodes that each choose which software to run, and those nodes only change when their operators decide to change them. A soft fork can ask them to change. Most of them, this time, said no. There is no lever above that layer. This is the same property that makes Bitcoin the one major asset with no insider round and no gatekeeper who can quietly rewrite the terms after you have already bought in, a point worth sitting with if you have ever felt like you showed up late.

That is not a flaw that BIP-110 exposed. It is the feature BIP-110 accidentally demonstrated for everyone. And it is why self-custody is not a side quest but the whole game. If you were running your own node this weekend, you did not have to trust a headline, a pool, or a developer to tell you which chain was real. Your node checked the rules and told you itself. If you held your own keys, no fork, successful or stillborn, could move your coins without your signature.

The most striking proof of that came, of all places, from the losing side. As Kratter weighed whether to follow the fork, the practical advice he gave his own audience was not about tickers or hashrate. It was this: if a hard fork does come, you need to still have possession and control of your own UTXOs, you still need to hold your own private keys. So, in the middle of the most divisive governance fight in years, the one thing both sides agree on, the thing that protects you no matter which chain turns out to be "real," is holding your own keys. Even the man who thinks the network has been captured knows that the person running a node and holding their own coins is the only one who gets to decide anything at all.

Consensus in Bitcoin is opt-in. That is the radical part, and it cuts both ways on purpose. The very property that let a passionate minority secede and start their own chain is the property that guarantees nobody can drag you onto a chain you did not choose. You cannot have one without the other. A network where a brilliant, certain group could force through the change they knew was correct would also be a network where a brilliant, certain group at a central bank could force through the change they know is correct. We already have that one. It is called the dollar.

Not your keys, not your coins. This weekend just added a second line to the creed: not your node, not your rules. The whole point of Bitcoin is that the rules answer to you, and to no one who merely wishes they were in charge.

That is the quiet freedom Bitcoin Well was built to make simple: buy, sell, and hold your Bitcoin in genuine self-custody, on rules that no committee, no developer, and no weekend fork can change without your say.

A note on sourcing: the BIP-110 mechanics (the "ReducedData" proposal, its February 2026 publication, the 55% signaling threshold, and the mandatory-signaling activation at block 961,632) draw on Jameson Lopp's layman's guide, the BIP repository, and reporting from CoinDesk, Decrypt, and Bitcoin.com News. The weekend timeline (roughly 2.5% signaling, the two blocks mined by Ocean and Roughnecks, the inherited ~127 trillion difficulty, and the roughly 350-day retarget math) comes from CoinDesk and Bitcoin.com News. The fullest window into the BIP-110 side's own thinking, the "coma / seeds of hope" framing, the node-runners-versus-mining-cartel read of Saturday, the Roughnecks "escalation" and "wildcatting" statements, the Balaji reform-versus-exit frame, the "legacy Bitcoin is now stock" line, and the plea to hold your own keys and UTXOs, is Matthew Kratter's own Bitcoin University video recorded the morning of August 10, 2026. Quotes from Luke Dashjr, Dathon Ohm, Michael Saylor, Adam Back, Guy Swann, and Mark "Murch" Erhardt are drawn from their public posts on X and the Bitcoin development mailing list, and from TFTC and CryptoTimes reporting, current as of August 10, 2026. Moving figures (signaling percentages, block gaps, node counts, and the status of the proposed September 1 proof-of-work fork) are live and worth re-verifying at publish time. The Austrian framing draws on Friedrich Hayek's "The Use of Knowledge in Society" (1945), Law, Legislation and Liberty, and The Fatal Conceit.