Bitcoin has failed to activate every proposed soft fork since Taproot went live in 2021. Paul Sztorc, the inventor of Drivechains and CEO of LayerTwo Labs, says the latest failure shows how difficult it has become for Bitcoin to approve major changes to its consensus rules.
Sztorc said BIP-110’s failure was part of a larger problem.
“All soft forks since Taproot have failed to activate, and this was no exception,” he said.
BIP-110 was a proposed temporary soft fork designed to limit certain types of non-financial data stored in Bitcoin transactions. During one 2,016-block difficulty period, only 51 blocks signaled support, equal to 2.53% of the total.
The proposal later triggered a split when BIP-110 enforcing nodes began rejecting blocks that did not signal support. The minority chain split from the main Bitcoin network at block 961,632 on Aug. 8 and produced only two more blocks before stopping.
The main Bitcoin chain continued normally.
BIP-110 required 55% of blocks to signal support through its voluntary activation process. By Aug. 2, that target had become mathematically impossible after only 28 of the first 1,108 blocks had signaled for it.
Even after that, the enforcing software entered its mandatory signaling period at block 961,632. Nodes running the software rejected blocks that did not meet the proposal’s requirements, but most miners continued working on the main Bitcoin chain.
By Aug. 9, the minority chain was stuck at block 961,633 while the main chain had moved 111 blocks ahead. OCEAN data showed around 257 petahashes per second assigned to the minority branch at that point, while Strategy Executive Chairman Michael Saylor estimated that about 99.85% of Bitcoin’s hash power stayed with the main chain.
The situation became harder for the minority chain because it inherited Bitcoin’s mining difficulty of 127.48 trillion. With very little hash power, miners on that chain could not produce blocks quickly enough to reach the next difficulty adjustment.
BIP-110, officially known as the Reduced Data Temporary Softfork, proposed seven restrictions that would remain active for 52,416 blocks, or about one year.
The rules included an 83-byte limit on OP_RETURN outputs, a 256-byte limit on certain data pushes and restrictions involving some Taproot functions.
Supporters, including Bitcoin Knots maintainer Luke Dashjr, said the changes would reduce arbitrary data storage connected to inscriptions and keep Bitcoin focused on financial use.
Critics, including Saylor and Blockstream co-founder Adam Back, argued that the proposal could weaken Bitcoin’s neutrality by rejecting transaction types that the network currently accepts.
Sztorc said the problem is not limited to BIP-110. He also pointed to other soft-fork proposals that have remained inactive since Taproot.
Taproot activated at block 709,632 on Nov. 14, 2021, after miners signaled their support through the Speedy Trial process. The upgrade added Schnorr signatures and new Taproot spending rules while allowing older nodes to remain on the same blockchain.
Since then, several proposed changes to Bitcoin’s scripting system have been discussed but have not activated.
Sztorc specifically mentioned OP_CAT, a proposed return of an old Bitcoin opcode that would allow scripts to join two pieces of data.
When asked how BIP-300 could overcome the resistance facing other consensus changes, Sztorc pointed to OP_CAT as an example.
“Nothing can — not even OP_CAT, which is just 13 lines of code and was in the original software and had lots of support,” he said.
He added that Bitcoin cannot activate soft forks “for the foreseeable future.”
OP_CAT was disabled by Satoshi Nakamoto in 2010 because of security concerns. A later proposal seeks to bring it back through a backward-compatible soft fork.
Supporters say OP_CAT could help developers build covenants, vaults, bridges and other programmable Bitcoin applications. But activation would still require enough agreement among developers, miners, node operators and users.
BIP-360 is another proposal that faces a similar challenge. It aims to introduce a new Bitcoin output type that could help users move funds into post-quantum, or quantum-resistant, addresses.
Drivechains are designed to take a different approach to experimentation. Instead of requiring Bitcoin’s base layer to adopt every new feature, Drivechains would allow developers to build separate sidechains with their own rules.
Under BIP-300, users could move BTC between Bitcoin and these sidechains through a two-way peg. Different Drivechains could then be used for things such as privacy, smart contracts or faster transactions.
The idea is that users could choose whether to use a particular sidechain while Bitcoin’s main rules remain unchanged after the initial Drivechain upgrade.
Sztorc said each Drivechain could have its own identity and software, similar to projects such as Liquid and Lightning.
However, Drivechains still face the same problem he identified with other Bitcoin upgrades: BIP-300 itself requires a consensus change on Bitcoin.
Without that change, the proposed Drivechain withdrawal system cannot be activated.
Another major debate around Drivechains involves miners.
Under BIP-300, Bitcoin miners would play an important role in approving withdrawals from Drivechains. Withdrawal requests would remain pending while miners vote on them through Bitcoin blocks. A request receiving enough support over the required period could then release BTC from the sidechain.
Critics have raised concerns about miners potentially approving invalid withdrawals, censoring transactions or working together to attack the system.
Sztorc argues that miner incentives would depend heavily on how much economic activity a Drivechain generates.
“If the chain is popular, it will be generating fees for miners. If this fee revenue is large, relative to the number of circulating coins on the L2, then it will be secure,” he said.
Under that model, a successful sidechain could create fee income for miners, giving them an economic reason to protect the system. Critics, however, argue that the value of BTC held in the sidechain could still create incentives for attacks.
The BIP-110 dispute also showed how mining companies and major Bitcoin holders can become involved in disagreements over network upgrades.
Foundry USA Pool asked its mining customers to vote on BIP-110 signaling before the mandatory period. Strategy, one of the largest corporate Bitcoin holders, publicly opposed the proposal through Saylor.
The failed fork also created a practical risk for users because BIP-110 did not include automatic replay protection.
Bitcoin developer Kevin Loaec warned that transactions made on one chain could potentially be copied onto the other. Users moving or selling coins on the minority chain could therefore face risks if they did not first separate their funds.
BIP-110 supporters also prepared code for a possible proof-of-work change. That would allow the minority chain to move away from Bitcoin’s existing mining algorithm.
The code included a configurable hard-fork time, but developer Chris Guida described it as a backup plan and had not announced an activation date.
The BIP-110 episode has therefore become another example of the difficulty of changing Bitcoin’s consensus rules. For Sztorc, that difficulty is exactly why systems such as Drivechains could become important, although Drivechains themselves still require Bitcoin to approve a consensus change before they can be deployed.
