Eight hours. That is how long Bitcoin’s most recent governance experiment actually lasted. On the morning of August 8, 2026, a group of nodes running the BIP-110 soft fork rejected a block that refused to signal support and started building their own chain. They managed exactly two blocks. Then the chain went quiet. The main network kept producing blocks at its usual rhythm, almost as if nothing had happened. By the time most people checked their phones, the minority branch was already a historical footnote. Yet the episode still matters. It tells us more about how Bitcoin actually decides things in 2026 than any carefully worded proposal ever could.
What The Failed Soft Fork Actually Revealed
The Reduced Data Temporary Softfork, better known as BIP-110, set out to limit the amount of arbitrary non-financial data that could be written into Bitcoin transactions. Its supporters wanted to push back against Ordinals inscriptions, BRC-20 tokens, Runes, and the larger Taproot data payloads that have become common. The idea was straightforward on paper: cap most new outputs at 34 bytes, keep OP_RETURN under 83 bytes, and limit data pushes to 256 bytes. The restrictions would last for roughly one year, after which they would expire. Supporters framed the measure as a temporary clean-up rather than a permanent cultural shift.
Activation required 55 percent of blocks in a difficulty period to signal support. That number itself was already a concession. Traditional soft forks have usually aimed for 95 percent. The lower bar showed that the authors understood they were not going to win overwhelming consensus. What they did not anticipate was just how far short they would fall. When the mandatory signaling window opened, support sat at 2.53 percent. Fifty-one blocks out of 2,016. Not fifty-one percent. Fifty-one blocks.
That gap is not a negotiation that almost succeeded. It is a clear statement. The mining pools that control the vast majority of hashpower looked at the proposal and decided that restricting fee-paying transactions ran against both their economic interest and their view of how the protocol should work.
Why Miners Walked Away
The economic case is the easiest part to understand. Inscriptions and related activity pay fees. During peak periods those fees have pushed average transaction costs well above twenty dollars and delivered meaningful extra revenue to miners. BIP-110 would have removed that revenue stream for a full year. No major mining pool publicly endorsed the proposal in the weeks leading up to the activation window. That absence of support was not an oversight. It was a deliberate choice.
Yet the money argument only goes so far. The deeper objection was philosophical. A number of prominent voices argued that Bitcoin’s value rests on its neutrality. Once the network starts deciding which valid, fee-paying transactions are acceptable, the door opens to broader forms of filtering. Michael Saylor published a lengthy essay listing one hundred and ten reasons the proposal was a bad idea. The central claim was simple: if inscriptions can be treated as spam today, other categories of transactions can be treated the same way tomorrow. The boundary between “financial” and “non-financial” data is not as clear as the proposal’s authors suggested. Multisignature setups that carry metadata, timestamp proofs, or sidechain anchors could all be reclassified under similar logic.
Bitcoin’s value comes from its neutrality. The moment you start deciding which types of valid, fee-paying transactions are acceptable, you open a door that cannot be closed.
Adam Back, whose earlier work on hashcash helped shape Bitcoin’s proof-of-work design, joined the opposition. He argued that trying to enforce disputed rules without broad support posed a greater risk to the network’s integrity than the data itself. When figures with that level of institutional and technical weight speak in the same direction, miners and node operators tend to listen.
The Replay Risk That Was Never Solved
One of the quieter failures of BIP-110 was the complete absence of replay protection. When a chain splits, a transaction that is valid on one branch is often valid on the other. Without protection, a user who spends coins on the main chain can see the same transaction replayed on the minority chain, or the other way around. The result is an unintended movement of funds. Historical hard forks such as Bitcoin Cash included replay protection from the start. BIP-110 did not. Its designers argued that because the change was a soft fork, the rules were simply a subset of existing consensus and therefore did not require special handling. In practice the chain split created exactly the conditions where replay becomes possible.
Node software providers quickly advised users to take no action, avoid splitting coins, and wait for the minority chain to fade. That advice turned out to be correct within hours. Still, shipping a proposal that its own authors knew might trigger a split without addressing replay risk looks like an internal governance failure of the BIP-110 effort itself.
Echoes Of The Blocksize War
Bitcoin has lived through larger governance fights before. The blocksize debates of 2015 to 2017 stretched across years, produced a permanent fork, and nearly delivered a hard-fork increase in the block size limit that was called off at the last minute when economic nodes made clear they would not follow. The parallel with BIP-110 is useful but incomplete. The earlier conflict was about scaling philosophy. This one was about what kinds of data belong on the chain at all. Both featured a minority trying to push a consensus change against the preference of the majority. Both failed.
The difference in speed is striking. The blocksize war took two years. BIP-110 was effectively settled in eight hours. That compression suggests the network’s governance culture has matured. In 2017 the question of whether miners or economic nodes ultimately controlled consensus was still open. In 2026 it is largely settled. Hashpower alone cannot force a rule change. Economic nodes, exchanges, wallets, and the businesses users actually rely on must agree. When roughly 97.5 percent of miners and virtually all major infrastructure providers reject a proposal, the outcome is not really in doubt.
SegWit itself activated through a user-activated soft fork that bypassed resistant miners. The lesson many drew was that users running nodes, not miners, hold the final say. BIP-110’s failure applies the same principle in reverse: a small group of nodes and miners cannot impose rules the rest of the network refuses.
The Proof-Of-Work Pivot And What It Signals
The most unexpected turn came after the fork had already failed. A mining operation that supported BIP-110 announced it would keep mining the stalled chain and look for a “sensible proof-of-work change” that would let it move past the pools that refused to follow. Changing the hashing algorithm means abandoning SHA-256. Every existing Bitcoin ASIC becomes useless for that chain. The minority branch would need new hardware, a fresh difficulty schedule, and enough exchange and wallet support to function as something that looks, in practical terms, like a new cryptocurrency.
History is not kind to projects that take this path. Bitcoin Gold switched to a different algorithm in 2017 and now trades at a fraction of its early value. The market’s verdict on proof-of-work changes that split from the main chain has been consistent and harsh. Whether the BIP-110 supporters actually follow through remains open. The fact that they are discussing it at all shows the underlying disagreement has not disappeared. The faction that believes non-financial data should be kept off Bitcoin still exists. They have simply concluded that the current governance process will not deliver the result they want.
Implications For Every Future Soft Fork
BIP-110’s collapse reaches beyond the inscription debate. Bitcoin still has pending proposals that would require soft forks, including covenant designs and various vault constructions. Each of them will have to clear the same bar that BIP-110 could not. Some observers have begun describing Bitcoin’s consensus rules as something closer to a constitution. Changes, in this view, should be rare and should never serve any single faction’s short-term convenience. If that framing becomes dominant, the practical threshold for any soft fork moves from technical majority toward near-unanimity.
The result is a protocol that is becoming harder to change both by design and by culture. Whether that is a strength or a weakness depends on what you think Bitcoin needs to remain useful. If the current rule set is already sufficient for its role as a monetary network, gradual ossification is a feature. If the network needs new capabilities to stay competitive with newer chains, the same ossification becomes a risk. BIP-110 did not settle that larger question. It did demonstrate that in 2026 the bar for changing Bitcoin’s rules sits higher than it has ever been.
What Would Shift The Picture
Two developments would force a rethink. First, if the minority chain successfully switches its proof of work and attracts meaningful hashpower, the episode stops being a failed experiment and becomes a permanent split. Exchange listings would be the clearest signal. As of mid-August 2026 no major venue has indicated plans to list the BIP-110 chain. Second, if inscription volume falls sharply for market reasons rather than protocol restrictions, the anti-spam argument loses force and the whole dispute becomes largely academic. Fee data and inscription counts over the next few months will tell us whether the underlying tension is still live or simply fading on its own.
In the meantime the practical checklist is short:
- Watch for any sustained block production on the BIP-110 branch beyond the original two blocks.
- Track formal proposals to change the hashing algorithm on the minority chain.
- Monitor exchange and wallet announcements. Continued silence after thirty days would confirm the market has moved on.
- Follow inscription fee revenue. If Ordinals and related activity keep generating meaningful miner income, the economic case against data limits grows stronger.
- Observe how future soft-fork proposals adjust their activation strategies. Any plan that sets a threshold below the traditional 95 percent now carries the memory of BIP-110’s rejection.
A Clearer Picture Of Who Decides
Perhaps the most useful way to read the episode is as a stress test of Bitcoin’s informal constitution. The network did not need a formal vote or a lengthy conference to reach a decision. Hashpower, economic nodes, and public argument combined to produce a result that was both rapid and decisive. The minority that wanted tighter data limits lost. They retain the option of forking further, but that option comes with the usual costs of starting something new.
I’ve found that the most interesting moments in Bitcoin’s history are rarely the ones that succeed on the first try. They are the ones that fail quickly and cleanly, because those failures reveal the real distribution of power. BIP-110 failed quickly and cleanly. The main chain kept moving. The minority chain stalled. The argument over what belongs on Bitcoin is not over, but the method for settling that argument has become clearer than it was a decade ago.
In the end the network treated the proposal the way it has treated most contested changes in recent years: it waited to see whether enough people would actually run the code, discovered they would not, and continued under the existing rules. That is not always elegant. It is, however, effective. And for a system whose primary job is to remain reliable under pressure, effectiveness still counts more than elegance.
Looking Ahead Without False Certainty
No one can say with confidence whether a similar anti-spam proposal will ever gather the support BIP-110 lacked. The neutrality arguments that proved decisive this time will remain available to opponents of any future attempt. Replay protection will almost certainly be expected next time. The activation threshold will be scrutinized more carefully. Cultural resistance to contested soft forks appears to be rising rather than falling.
At the same time, the economic incentives that made inscriptions attractive to miners have not vanished. As long as people are willing to pay for block space to store data, that activity will continue to compete with ordinary payments. The tension is structural. Governance can manage it, delay it, or channel it, but it cannot eliminate the underlying trade-off between openness and efficiency.
For now the practical outcome is simple. The Reduced Data Temporary Softfork is dead. The main Bitcoin chain continues under the rules it had before August 8. The minority chain sits at two blocks and shows no sign of recovery. The conversation about what Bitcoin is for will keep going. The mechanism for deciding which proposals actually change the protocol has, if anything, become more conservative. That conservatism may frustrate those who want faster evolution. It also reassures those who value the network’s resistance to sudden shifts. Both perspectives are now part of the living record of how Bitcoin governs itself in 2026.
