By: Xiao Bing

On the evening of August 7, Bitcoin split into two chains at block height 961632.

Nodes running BIP-110 software began to reject all blocks that did not express support signals, diverging from the main network. AntPool mined the first standard block at this height, which was accepted by the main network but rejected by BIP-110 nodes. Subsequently, a miner working through the Ocean mining pool mined an alternative block, and the BIP-110 chain followed this block.

Eight hours later: the minority chain had only progressed to block 961633, producing a total of two blocks. During the same period, the Bitcoin main chain had reached block 961681, ahead by 48 blocks.

This fork, which lasted less than a day, has technically declared failure.

What does BIP-110 aim to do?

BIP-110 stands for "Reduced Data Temporary Softfork," published under the pseudonym Dathon Ohm, with Bitcoin core developer Luke Dashjr considered the drafter and primary proponent of the original proposal.

What it aims to do is not complex: to temporarily limit the storage of non-financial data in Bitcoin blocks through consensus rules for about a year. The specific targets are Ordinals inscriptions, BRC-20 tokens, Runes protocol, and other actions that write arbitrary data such as images and text to the blockchain via transaction scripts and witness data.

Supporters' logic is that this data occupies block space, drives up transaction fees, increases the operating costs of full nodes, and deviates from Bitcoin's core mission as a "sound monetary infrastructure." Dashjr refers to this data as "spam," believing long-term accumulation will threaten Bitcoin's decentralization.

Opponents have equally clear logic: Bitcoin's block space belongs to all who pay to use it. Any legitimate transaction that pays enough fees should be accepted; consensus rules should not determine what kind of data is "qualified." Michael Saylor published a lengthy article opposing the proposal in July, stating: "Bitcoin does not need pure guardians; it needs neutral guardians." Adam Back also warned that forcibly pushing BIP-110 without broad consensus could split the network.

BIP editor Murch added a note when assigning the number, calling it "a poorly considered and hasty soft fork proposal," stating that the number only represents procedural compliance and not endorsement.

Why were only two blocks produced in eight hours?

The minority chain became paralyzed after the fork, a mechanism that could be entirely anticipated.

Bitcoin adjusts its mining difficulty every 2016 blocks, aiming to maintain an average of one block every 10 minutes. The minority chain inherited the main chain's complete difficulty setting at the time of the fork but had only a minimal amount of hash power. Over the past two weeks, only 2.53% of blocks expressed support for BIP-110, far below the 55% required for activation, with the vast majority of signaling blocks coming from the Ocean mining pool.

This means the minority chain had to use less than 3% of the main network's hash power to solve the mathematical problem of the main network's 100% difficulty. Given the observed block production speed, it is estimated that the minority chain would need about 350 days to complete the next difficulty adjustment, while the main chain only requires 14 days.

Before reaching a difficulty adjustment, the minority chain will remain in an almost stagnant state. Meanwhile, BIP-110's rules require that within a two-week window before block 963647, every block must express a support signal. At the current speed, the minority chain will not even reach the end of this window.

As of UTC August 9, 11:40, the main chain had progressed to block 961725, while the minority chain remained frozen at 961633. None of the 113 new blocks produced on the main chain expressed support for BIP-110.

Real risks faced by ordinary users

The death of the minority chain does not mean that risks have disappeared. Both chains accept signed transactions of the same format, creating a window for replay attacks.

Assuming a user wants to sell forked coins on the minority chain (if there is a trading market for the forked coins), he signs a transaction to send the forked coins to a buyer. The problem is that this signed transaction is also valid on the main chain. The buyer can broadcast the same signed transaction to the main chain, withdrawing an equivalent amount of real BTC from the seller.

The slow block production speed of the minority chain exacerbates this risk, as the confirmation time for forked coin transactions is extremely long, giving attackers ample time to operate on the main chain. In the absence of built-in replay protection mechanisms in BIP-110, any transaction involving forked coins poses the risk of unexpectedly losing main chain BTC.

For the vast majority of Bitcoin users, the safest approach is to ignore the existence of this minority chain, not to move any forked coins, and not to participate in any related transactions.

Governance issues did not disappear with the fork

BIP-110 failed, but the governance rifts it exposed will not automatically heal.

The last significant soft fork on Bitcoin was SegWit (BIP-141) in 2017, which underwent two years of governance tug-of-war, requiring 95% miner signaling to activate and ultimately barely passed. The 2021 Taproot upgrade achieved broad consensus and activated smoothly. BIP-110 lowered the activation threshold to 55%, yet still only garnered 2.53%.

Dashjr issued a more aggressive signal before the fork: if BIP-110 fails, changing Bitcoin’s proof-of-work (PoW) algorithm "might be the only option left," implying that if miners refuse to cooperate, node operators might bypass miners to modify rules directly. This represents a fundamental challenge to Bitcoin's security model.

The essence of this debate is: who actually owns Bitcoin's block space.

Saylor's position represents one viewpoint: block space is a neutral public resource; whoever pays can use it, and consensus rules should not distinguish the "purpose" of transactions. Dashjr's position represents another: the core function of Bitcoin is currency, and all uses that deviate from this function are an abuse of limited resources, with protecting the network's long-term viability being more important than short-term "freedom."

Today, miners vote with hash power, choosing neutrality; this outcome is favorable for supporters of Ordinals and BRC-20, but a setback for those who hope to keep Bitcoin's block space "pure."

However, Dashjr will not stop here. As long as inscriptions and Runes continue to consume block space, this debate about what Bitcoin "should be used for" will reappear in different forms.

BIP-110 proved one thing with its two blocks: in the world of Bitcoin, idealism without computing power support cannot go far.