The Fork That Didn’t Change Bitcoin
In 2026, supporters of a proposal called BIP-110 tried to add new rules to Bitcoin. The change did not reach the support it needed, so a group associated with Bitcoin developer Luke Dashjr launched a separate chain, this time with a different mining system called BLAKE2b, that went live on September 1, 2026. It is a new network with its own coin that shares Bitcoin’s history up to the split. Bitcoin itself did not change. Whether the new chain lasts depends on whether miners, wallets, exchanges, and users show up. In the weeks after launch, very little of that happened. Forking the code was permissionless. Changing the network was not.
What They Wanted to Change
The dispute was about data. For years, some Bitcoin users have been frustrated that blocks get filled with arbitrary, nonfinancial data rather than payments. Their proposed fix, BIP-110, was a set of temporary consensus rules restricting several ways arbitrary data could be embedded in Bitcoin transactions, including limits on OP_RETURN outputs and data pushes. Supporters argued it would ease the storage burden on people who run nodes and keep Bitcoin focused on money. Critics argued that fees and each node’s own policies, not a network-wide rule, should decide how block space gets used.
This is a sincere disagreement between people who care about Bitcoin. One side wanted to protect what they saw as Bitcoin’s monetary purpose. The other saw a new restriction they did not want written into the rules.
Why the Change Did Not Happen
BIP-110 was designed to activate only if enough of the network signaled for it. It did not get close.
The deployment threshold was 55 percent of blocks in a single difficulty period, roughly 1,109 of 2,016 blocks. Support never approached it. In the final voluntary signaling period, 51 of 2,016 blocks signaled for BIP-110, or 2.53 percent. When the enforcing version ran anyway, it split onto a separate branch in August 2026 that held only a tiny fraction of Bitcoin’s total hashpower and stalled after two blocks, because nearly all of Bitcoin’s mining kept extending the existing network instead. In plain terms, the change was put forward, the network did not adopt it, and the main chain never moved.
That is the mechanism from the explainer, playing out in real time. Developers can propose a change. They cannot make everyone else run it.
So They Left, and Changed the Mining Instead
Losing the vote on Bitcoin’s own chain left one option that has been taken before. Start a separate chain. That is what the group did, and they made one telling adjustment.
The first attempt failed partly because a minority chain still depended on Bitcoin’s miners, and those miners did not come. So the new chain changed the mining system itself. It replaced Bitcoin’s long-standing algorithm, SHA-256, with a different one called BLAKE2b. Bitcoin’s existing SHA-256 ASICs cannot mine a BLAKE2b proof-of-work chain, and the change was intended to free the new chain from Bitcoin’s existing mining base. The proof-of-work change is permanent; some of the data restrictions are set to expire in 2027.
What Actually Happened at the Fork
The August branch was not the end of the experiment. On September 1, 2026, the group launched a separate BLAKE2b chain as a flag-day fork, changing the proof-of-work algorithm as part of the new network’s rules. From a set point, the new chain’s software began following BLAKE2b and rejecting the old SHA-256 blocks. From then there were two networks. Bitcoin, unchanged, and a separate BLAKE2b chain with its own coin, sharing Bitcoin’s transaction history up to the split.
Launching is not the same as lasting, and the weeks since launch made the gap plain. The mining swap that was meant to free the new chain from Bitcoin’s miners also left it with very few miners. BLAKE2b is also used by Sia’s proof-of-work system, so existing hardware designed for that algorithm can potentially be used on the new chain, while Bitcoin’s SHA-256 ASICs cannot. Very little of even that hardware showed up. What mining the chain did attract, a handful of petahashes across a few dozen small pools, was a rounding error next to Bitcoin’s, whose network runs many thousands of times larger. As of late September, no major exchange or market aggregator had listed the coin, no mainstream wallet had added support, and no Lightning implementation had followed. The chain also launched without universal replay protection, the safeguard that keeps a transaction on one chain from being repeated on the other, offering only an opt-in method that does not cover every wallet automatically.
On September 21, the chain’s developers extended its coinbase maturity to about 45 days, so miners must wait roughly that long before they can spend newly mined rewards. The change came amid concerns that a few pools were growing large enough to threaten a network with so little total mining power. It is a revealing detail. A chain that small can be destabilized by a single participant getting big.
None of that settles whether the chain will ever find its own footing. It might still gather a mining community and an economy of its own, the way earlier forks did at a smaller scale. What these early weeks do show is the thing this article is about. The launch did not change Bitcoin. The main network kept running under the same rules it had the day before, and almost none of the mining or the market moved to the new one. Launching a chain proves the software runs. It does not prove a network will form around it.
Final voluntary signaling period, Aug 2026
Chain split, Aug 8, 2026
Weeks after launch, Sep 2026
The BLAKE2b fork swapped out Bitcoin’s SHA-256 mining specifically so it would not need Bitcoin’s existing miners, who had already declined to support the earlier attempt. But BLAKE2b is the algorithm the Sia network already uses, and only that narrow class of hardware can mine it. Very little of it turned up. Existing Bitcoin machines cannot mine BLAKE2b at all. The move that was meant to make the chain independent is the same move that left it nearly empty.
What It Means if You Hold Bitcoin
For most people holding bitcoin, the main chain and the bitcoin you hold on it were unaffected by the split. A separate chain launching does not alter the Bitcoin network or its supply.
The one real caution is replay risk. Because the two chains share history and this fork launched without universal replay protection, anyone planning to interact with the new chain should understand that a transaction could, in some cases, be valid on both. Whether you can access any coins on the new chain depends on how and where you hold your bitcoin and on what support appears. This is a description of the situation, not advice on what to do about it.
The Pattern, One More Time
Strip away the specifics and this is the third time the same story has run. Big-block supporters could not change Bitcoin in 2017, so they made Bitcoin Cash. A faction of that group split again in 2018 and made Bitcoin SV. In 2026, a group that wanted stricter data rules could not change Bitcoin, so they launched a chain of their own. In each case, the result was a separate network and a separate coin, rather than a change to Bitcoin’s existing chain.
Different motives, opposite directions, same ending. They could change the rules on a network they controlled. They could not make Bitcoin’s existing participants adopt those rules. That is not a flaw in the story. It is the point of it.
Anyone can fork Bitcoin. The harder question is whether anyone follows.
All Roads Lead to Bitcoin
New to the concept?
This case study is the live example. The explainer covers soft versus hard forks, why copying the code is not the same as forking the chain, and the full history behind the pattern.
Common Questions
Did the BLAKE2b fork change Bitcoin?
No. It created a separate network with its own coin. Bitcoin’s main network continued under the same rules. A hard fork can produce a new chain, but it does not alter the original network unless that network’s own participants adopt and enforce the change, which here they did not.
What was BIP-110?
A proposed set of temporary consensus rules restricting several ways arbitrary, nonfinancial data could be embedded in Bitcoin transactions, including limits on OP_RETURN outputs and data pushes. It needed 55 percent of blocks in a difficulty period to activate. Support peaked at 2.53 percent, and its enforcing branch stalled after two blocks.
Did anyone end up using the new chain?
In the weeks after launch, very few participants had appeared. The chain drew only a sliver of mining power, a rounding error next to Bitcoin’s, with no major exchange listing and no mainstream wallet or Lightning support. Launching a chain proves the software runs; it does not prove a network will form around it.
Do I need to do anything to protect my bitcoin?
Your bitcoin on the main chain was not changed by the fork. The main thing to be aware of is replay risk if you choose to interact with the new chain, because it launched without universal replay protection. Whether you can access any new-chain coins depends on how and where you hold your bitcoin and on what support appears. This is information, not financial advice.
Why do people keep forking Bitcoin if it never changes Bitcoin?
Because forking the open-source code is easy and permissionless, while changing the actual network requires enough participants to adopt and enforce the change. Groups that cannot win that adoption can still launch a chain of their own. Whether it becomes a lasting network is a separate question, decided by adoption.
A fork can look like a moment when Bitcoin itself is about to change. This one showed the distinction clearly. A separate chain launched, Bitcoin’s main network continued under its existing rules, and the new chain attracted only a tiny fraction of the mining and market support surrounding Bitcoin. The code could be copied. The network could not be carried with it.
Sources and Further Reading
Current as of late September 2026.
- Primary BIP-110 specification, Bitcoin BIPs repository. The proposed temporary consensus rules, the 55% activation threshold, and the deployment mechanics.
- Primary Bitcoin Knots pull request #419, “Long coinbase maturity”, merged September 21, 2026. The change extending coinbase maturity to about 45 days.
- Signaling record What Is BIP-110? Bitcoin’s Data-Limit Proposal and Why It Failed, Simple Mining. The 51-of-2,016-block signaling result (2.53%), the minority branch’s share of hashpower, and the two-block stall. See also the BLAKE2b chain’s hashrate explorer.
- Reporting Bitcoin BLAKE2b fork faces Sept. 1 launch test (crypto.news) and Bitcoin Knots is trying to fork Bitcoin again (CryptoSlate), on the launch, the background, and the replay and block-format issues.
- Reporting A Bitcoin Hard Fork Went Live September 1. Miners, Exchanges, and Traders Ignored It (Yahoo Finance), on the post-launch outcome.
- Series What Is a Hard Fork? (the explainer this case study sits alongside), The Bitcoin myths, explained, and how we test every claim on this site.
Everything on this site is for educational purposes only. It is not financial, investment, tax, or legal advice. Bitcoin carries real risk. Prices move, sometimes sharply. Do your own research, think for yourself, and speak with a qualified professional before acting on anything you read here.