Bitcoin is on the verge of its first protocol change in three years. Or is it?

For better or for worse, change has never come easy for Bitcoin. Will that still be true now that a proposed protocol change has widespread support within the community?

The proposed modification, called Taproot, has been in the works for years. It’s widely seen as a worthwhile modification thanks to the privacy and efficiency gains it would bring to Bitcoin. Taproot’s code was recently merged with Bitcoin Core, the dominant Bitcoin client. The only step left is to actually deploy it — and there doesn’t seem to be any substantial resistance.

Still, it’s the first protocol change since Segregated Witness (SegWit). And even though that episode occurred three years ago, the memory of the messy conflict that grew out of the debate over SegWit remains fresh in the minds of many in the Bitcoin community.

The situation is much different this time around. But the community still lacks a detailed activation plan for Taproot. And the 2017 episode serves as a stark reminder that leaderless Bitcoin is unpredictable when it comes to making decisions.

In other words, Taproot is far from a done deal, no matter its popularity.

A clever ‘cryptographic trick’

At a high level, Taproot makes it possible for users to do more complicated things with Bitcoin without revealing that they are doing anything out of the ordinary.

Today, anything more complicated than a “one-key-one-signature” transactions from ordinary wallets — Lightning network payment channels, escrow services, and multisignature wallet schemes to name a few examples — are “visibly distinct” on the blockchain, says Andrew Poelstra, one of Taproot’s creators. With Taproot, they will look like every other transaction.

The premise of Taproot is that nearly every single use of Bitcoin ”can be encoded as a single public key and a single signature,” says Poelstra. Thanks to a “cryptographic trick,” it’s possible to place hidden conditions inside an ordinary-looking public key — for instance smart contract-based instructions about who can spend coins, he says. This has benefits for both privacy and scalability.

Taproot won’t do this all by itself, though. The proposed update package also includes a change to the digital signature algorithm that Bitcoin uses. For its whole life, Bitcoin has relied on the Elliptic Curve Digital Signature Algorithm (ECDSA), but the network’s developers have long desired to replace ECDSA signatures with a different kind, called Schnorr signatures.

Switching to Schnorr signatures goes hand-in-hand with the introduction of Taproot because the new Schnorr signatures will make it easier to encode spending policies in a single public key, says Jonas Nick, one of the Core developers behind the soft fork proposal. They also can also improve the privacy and efficiency of Layer 2 protocols like the Lightning network, he says.

Schnorr signatures are not required for Taproot to work, says Nick. But they make it simpler to achieve things that are theoretically possible with ECDSA yet “have been so complicated that they haven’t seen significant usage,” he says.

“Without Schnorr signatures, Taproot is effectively limited to single party transactions,” says Poelstra. Ordinary wallets would be able to take advantage of the benefits, but not multi-signature wallets, escrow services, or anything else that involves more than one party.

Running unopposed

From what Nick and Poelstra can tell, almost everyone in the community seems to think that a soft fork — a backwards-compatible protocol change that won’t kick old nodes that don’t update off the network — to add Taproot and Schnorr signatures is a good idea.

“Not only the Bitcoin Core folks,” says Poelstra. Wallet developers, exchanges, privacy advocates, and researchers would all like to see it happen, he says. “The impression I get is that there is a tremendous amount of popular support for Taproot and not really any opposition.”

The so-called Bitcoin Improvement Proposals (BIPs) describing the changes have been worked on in public for more than 18 months, says Nick, who helped author all three BIPs. “We would have heard technical concerns by now,” he says. “I’m not aware of resistance at this point — quite the opposite.”

If that’s true, it's a far cry from the conflict that erupted around SegWit three years ago. The root of the fracture was a disagreement between Core developers and some in the mining community over how best to go about improving Bitcoin’s transaction capacity. The developers saw SegWit as the solution, while a contingent that included Bitmain, the dominant company in the mining space, resisted it and pushed instead for an increase to the size of Bitcoin’s blocks.

The conflict dragged on two years, and eventually culminated with a split between the two factions. Those in favor of larger blocks executed a hard fork and created a second currency, Bitcoin Cash.

Lessons learned

Poelstra says an important lesson from the SegWit saga is that developers should think more adversarily when coming up with the specific activation strategy, which will include directions for how miners should signal intent to update, the percentage of the network’s hashpower that must signal intent to activate, and the deadline by which this signalling must occur. The SegWit process showed how a minority with veto power can be disruptive, says Poelstra.

Developers should be very careful about how their work affects miners, says Poelstra. SegWit changed the way that transactions were hashed into blocks in a way that broke something called AsicBoost, a technology that some miners had been using to gain an advantage, he says. For a number of reasons, “nobody knew this until well after things had blown up in a much bigger sense.”

“What’s nice about Taproot is that it does not touch the Bitcoin block headers in any way,” says Poelstra.

The SegWit saga also had important consequences that are difficult to quantify. “There’s an aversion from a lot of people who have technical or intellectual leadership to take that sort of technical or intellectual lead in deploying these things because there was so much heat and it was such a tumultuous time,” says Poelstra.

By design, Bitcoin isn’t supposed to have a leader. But it might help if there was someone or some entity whose job it was to make sure a popular change like Taproot actually gets deployed. Poelstra says he’s observed many people assuming that the soft fork is a sure thing. Crucially, however, there is still no specific technical proposal on the table for the activation mechanism.

Enthusiasm for Taproot on social media and the apparent lack of opposition to it is “not quite the same thing as having technical buy-in of a form that says, ‘This is happening and I am deploying code to make it happen,’ — which is what you need for this to be a sure thing,” he says.

There have been at least a couple attempts at some level of coordination, though. Core developer Anthony Towns, who co-authored two of the Taproot/Schnorr BIPs, recently conducted an informal survey of “smart dev-type” people to gauge their viewpoints on various components of a potential activation strategy. Towns wrote that he found it difficult to infer many obvious patterns in the responses, however.

Last week, mining pool Poolin, which accounts for over 10% of Bitcoin’s mining capacity, launched an initiative to gather feedback from the mining pools. Already, six mining pools representing more than 50% of the network’s hash power have signaled support for Taproot, though they still need to get on the same page about a specific roadmap for deployment.

Poolin’s goal is to “gather as much information as we possibly can” and encourage communication between the mining community and the developer community, says the firm’s vice president Alejandro De La Torre. There was not enough of this kind of communication during SegWit, he says, and that was a major problem.

“One of the major problems in those days was not knowing what the mining pool operators wanted, or not even being able to reach them,” says De La Torre. “We are trying to mitigate that issue now.”