S15 E45: Lucas Ontivero & Turbolay on Wasabi Wallet's Future

zkSNACKs shut down the biggest CoinJoin coordinator on Bitcoin, and most of the timeline treated it like a funeral. I invited the two developers who refused to leave the wake: Lucas Ontivero and Turbolay (Clément), the people still shipping the Wasabi client after the company money and the official coordinator both walked out the door.

Lucas has been around this project since before it was even called Wasabi. He was there with nopara73 when the repository got renamed, when rounds barely happened a few times a month, and when Matthew Haywood left for Blockstream. This was his second appearance on my show and his first on video. Turbolay had been a backend developer funded by zkSNACKs for about two years. His first time here. Together they are keeping CoinJoin and the client alive as open source, with a temporary maintenance runway from the old company and a very loud "war mode" to-do list.

The question hanging over the whole hour and a half was simple: after the fattest coordinator dies, is Wasabi still a wallet worth caring about? Lucas's answer, boiled down, was that privacy is not a feature toggle. "Without privacy none of the properties of Bitcoin are real," he told me later, and the rest of the episode was about keeping the one wallet that still takes that sentence seriously from becoming a museum piece.

War Mode: Make Wasabi Unfuckable

Turbolay's first confession was that Wasabi had never been a normal open source project. It was "really well funded," with a crowd of paid contributors and a codebase shaped for that crowd. When zkSNACKs decided to shut down, that luxury ended overnight. "We have to reduce everything, downsize, make it maintainable by a handful of people," he told me, plus sporadic outside patches for critical bugs.

Downsizing means both smaller features and a thinner server. They already ripped out password recovery, the built-in brute-force tool that treated your Wasabi password like a passphrase and tried PIN-style guesses. Their narrative: dedicated open source password tools already do that job better, and they no longer have spare cycles to babysit non-critical code. They also want less load on the old zkSNACKs-shaped backend and more work on the client, so coordinators and backends become cheap enough for random people to run.

Lucas put the priority in one brutal sentence. He does not want Wasabi to "suffer the same that Samourai suffer," where users open the app one day and discover the backend is gone and the wallet is useless. Separate the APIs so coordinators and backend providers can run independently. Keep CoinJoins resilient across the roughly ten community coordinators that already exist (only a handful have real liquidity). Privacy toys for Wasabi 3.0, payments inside CoinJoins (already in the protocol, not yet in the UI), Silent Payments, and everything else WabiSabi already enables can wait. First goal, in his words: make Wasabi "unfuckable."

I told him that sounded like an actual product roadmap.

Why Wasabi Lived and Samourai Froze

The contrast with Samourai was the emotional center of the early conversation. Samourai's developers got arrested, the community largely waited for the same people to walk out of jail, and Whirlpool liquidity stayed glued to clients that needed new trusted binaries. Wasabi did something uglier and healthier. zkSNACKs announced the coordinator shutdown, and within about two weeks Lucas remembers "eight new coordinators running." "It's amazing," he said. "It's absolutely amazing."

That speed was not a zkSNACKs payroll miracle. Kukks, working on BTCPay Server, turned Wasabi coordinator code into a BTCPay plugin. Other people built bots and liquidity dashboards. Ginger appeared as a near one-to-one fork that kept a business-shaped model and, as I understand it, a willingness to negotiate compliance if that is what it takes to stay open. Ex-zkSNACKs contributors kept filing docs and fixes with "zero money." Random GitHub strangers showed up. Someone almost finished a Start9 packaging. Someone else shipped NixOS packages. "They are not letting Wasabi to die," Lucas said, and he called the feeling "really emotional."

Turbolay's diagnosis of the Samourai freeze was practical: if Whirlpool users had been forced onto a brand new client with brand new binaries, gathering trust and liquidity would have been brutal. Wasabi users still had a working private wallet even without CoinJoin. Block filters, Tor, coin selection, all the fancy privacy work zkSNACKs once funded, still ran. New entities only had to point liquidity at a living client instead of asking everyone to reinstall faith.

For the older Wasabi episodes on this site, start with nopara73 on beginnings, Max Hillebrand on the development roadmap, the S5 E6 privacy scrap with nopara73, and Max on privacy advocacy. The ShopinBit "buy anything" era with Max and Lando still lives here.

Not a Mixer: An Integrated Privacy Wallet

I asked the rude question every Joinstr and JoinMarket fan asks: why should anyone still care about Wasabi instead of migrating? I had just talked to F/Dev (1440000bytes) about Joinstr, the Electrum plugin that coordinates over Nostr without a central coordinator. Lucas refused the zero-sum frame. "All the solutions that provide some privacy to Bitcoin are welcome," he said. Fighting other privacy projects like a territorial business is, in his word, "silly."

The difference, as he drew it, is that Wasabi is not a mixer you visit and abandon. The wallet "takes decisions based on the anonymity set," remixes change, and tries to make you reveal exactly what you choose to reveal. People who treat Wasabi as a tumbler that mixes then moves elsewhere are doing it wrong. "CoinJoin is just one part of that," he said. "It's an integrated system to protect your privacy."

He also answered Floppy's jab that a government could run a Wasabi coordinator. "Yes," Lucas said, almost cheerfully, "and we are designing Wasabi in a way that even if a government runs a coordinator they can learn nothing about you. So welcome governments who want to run Wasabi coordinators."

Autopilot, Rockets & Other Underrated Toys

I spent a stretch praising features the timeline still underrates. Autopilot coin selection is the automatic gearbox of UTXO management: the wallet picks coins for you, and before you confirm you can nudge amounts to avoid change. Pocket selection shows visually who already knows about which coins. PayJoin exists. Payments inside CoinJoins exist in the protocol. The late-arriving transaction accelerator hides CPFP and RBF behind a rocket icon so grannies and Peter McCormack types do not have to learn mempool jargon.

Lucas's favorite quiet cleverness: if he pays me, gets change, then pays me again, the wallet prefers coins I already saw, so I learn nothing new about his stack. Useful when the counterparty is an exchange.

Turbolay admitted the automatic selector is still a black box, and Bitcoiners hate black boxes even when the Average Joe would do worse by hand. "Most of the times the automatic coin selector will do a much better job than the Average Joe," he said, "still it's a black box and so people have to trust it." Their answer was an advanced send workflow, freshly merged at recording time, where you pre-select which coins are available and then still let the engine build the transaction from that subset. "Best of both worlds," in his phrase.

Address Reuse, Graphs & a Bounty Nobody Took

Then I teased them with F/Dev's address-reuse screenshots, which got louder once CoinJoins got smaller and more coordinators appeared. Turbolay said there was a temporary bug years ago, investigated "way too much," and that looking at the chain alone cannot tell you whether reuse is a client bug, a deliberate payment-in-CoinJoin to a reused address, or someone trying to "harm Wasabi's reputation." Lucas added the protocol truth: with about ten coordinators and multiple wallet implementations, the only way to forbid reuse would be a smart coordinator that also forbids payments in CoinJoin. "The client has to make sure," he said. "It's an open protocol."

Their standing challenge still amuses me. Someone claims Wasabi is broken and traceable: take a UTXO, mix it, prove where the money went. There was even money on the table historically. "Nobody have never accepted that," Lucas said. "There are no proof, there are only claims." They still warn that users can wreck their own privacy by treating Wasabi like a mixer and then doing something dumb outside CoinJoin.

Open Protocol, Daemon & Friends Who Do Not Need Permission

BTCPay and Trezor integrations used to look like corporate partnerships. Lucas reframed them as open-protocol accidents. "You don't need permission to integrate anything with Wasabi," he said. The revenue-share glue for Trezor got deleted because there is no zkSNACKs revenue share left; Trezor can still switch coordinators. Hardware wallets cannot see the internet, so Trezor's old job was verifying fee rates and CoinJoin correctness before auto-signing. BTCPay's cooler trick, in Lucas's view, is batching many payments inside CoinJoins automatically.

Turbolay pointed at a second integration layer most people miss: the Wasabi daemon and RPC. Instead of reimplementing WabiSabi, a company can slap a custom GUI (or a mixer-looking front end) on top of the full client. "Almost all the features of the wallet are accessible through the daemon," he said. "We have no idea how many exist." Some mixers may already be doing it quietly.

Hardware wallet support stays on the short-term must-not-break list, even in war mode. Fancy new device integrations can wait until the client is safe against unknown coordinators.

Liquidity Without a Central Coordinator

Centralized zkSNACKs rounds once regularly pushed past 180 inputs, sometimes much higher. Decentralization scared people into imagining five-input samourai-sized rounds forever. Lucas pushed back with fresh numbers: a recent successful round on Kruw's coordinator had about 190 inputs and 196 outputs. Liquidity pages the community built (they kept pointing at wasabi.best) showed real volume across Group, Wasabi-coordinator.io, Ginger, and others in a single day. "We can make the same volume that zkSNACKs [had] in a distributed way," he argued.

Turbolay was more precise. Historical zkSNACKs averages sat around 180 inputs per round; Kruw's busiest coordinator was already around 120 on average. Close, not identical. Smaller coordinators with different value propositions still need a CoinJoin engine that does not assume zkSNACKs-scale liquidity, and that is exactly the work they neglected when there was only one real coordinator. A future hard-fork-ish protocol upgrade is on the mental whiteboard, no date attached, aimed at that lower-liquidity world.

Discovery is the other half. Fresh installs still showed the music box "waiting for CoinJoin" against a dead zkSNACKs endpoint until users manually changed the coordinator. The next release would ship coordinator connection strings: copy a magic string from a landing page, paste into Wasabi, confirm the fee and attributes. Later they want Nostr-based coordinator discovery inside the client. Until then, the free market of fees and AML policies is already visible: Ginger runs chain-analysis screening (same public-blockchain analysis zkSNACKs used, Turbolay insisted, which "does not degrade privacy" beyond what the chain already shows); Kruw.io was offering low minimums and no coordination fee while bootstrapping; other coordinators discount small amounts.

Maximum Fee Zero & Free Coordinators Under Fire

Trusting a random coordinator with your coins is the nightmare. Turbolay's answer was the design claim Wasabi has made for years: "you don't have to trust coordinators whether with your money or your private information," and the coordinator should learn no more than a chain observer. Reality still has non-trivial exploit surfaces, so the post-shutdown release (they pointed people at version 2.0.8.1) added a maximum coordination fee setting that defaults to zero. Free coordinators cannot charge you if your max is zero. Other checks cover minimum input counts and mining fee-rate lies where a coordinator could quote 1,000 sat/vB, pay miners 1, and pocket the rest.

Lucas admitted the psychological shift: for six years there was effectively one coordinator, so malicious-coordinator scenarios stayed abstract. Now they are shipping protections in a rush. Free coordinators, he added, had already been DoS-attacked "in one month more than zkSNACKs coordinator was attacked in six years." If nobody cared, nobody would bother.

Why Keep Working When the Paycheck Ends?

I asked the personal question an hour in: why not take a boring corporate job? Turbolay corrected me slightly. zkSNACKs is still funding maintenance for a while as they wind down, which he called "really nice." Beyond that, Wasabi is "one of the best way to have financial sovereignty" and "opportunity of a lifetime" for his generation. Impact on people who need private Bitcoin is bigger than another well-paid ticket.

Lucas was blunter about money: the company money "is over or will be over one day," he will hunt open source grants, and he still will not quit. He had wanted this work since he followed nopara73 before the rename. "Without privacy none of the properties of Bitcoin are real," he said. Borderless and permissionless collapse if anyone who dislikes your payment can censor it. "Privacy is the starting point." The anonymous coordinator operators putting themselves at risk make him feel less like "the only idiot." "It's not all tokens, it's not all Michael Saylor," he said. There are people who still want Bitcoin to be money you can actually use without a dossier.

I shared my own fear that a fully KYC'd Bitcoin becomes the CBDC people pretend to hate. Lucas hoped WabiSabi becomes obsolete sooner rather than later because better layer-one or layer-two privacy arrives. Covenants interest him vaguely; Turbolay wants anything that someday makes mining-fee-heavy CoinJoins less punishing for people with under a thousand dollars. I ranted about OP_CAT, OP_CTV, and even MimbleWimble extension blocks. They stayed humble about soft-fork preferences and focused on not dying before those debates finish.

nopara73 showed up in the live chat with the perfect roast: without a marketing team, Wasabi devs must crawl out of their dungeons. Turbolay owned it. He likes to code more than talk.

Showing up anyway mattered because the project is "nor dead nor dying," just making a "huge" turn from company product to community open source on easy mode: production-ready code, brand, and community already exist. Updates now check GitHub instead of zkSNACKs servers.

The @WasabiWallet Twitter account transferred to the open source project. Slack and GitHub are open for code, docs, and support. Technical users who hate the GUI can use the daemon. "If there is a lacking feature in the UI," Turbolay said, "it is not lacking on the daemon."

Follow Lucas as lontivero (more active on Nostr) and Turbolay on GitHub, Slack, or Telegram under that handle. Watch the WalletWasabi repo. Pick a coordinator that matches your paranoia. zkSNACKs closed the biggest booth on the fairground. The fairground is still open, and for Bitcoin privacy that is the only part that ever mattered.

Listen to Lucas Ontivero & Turbolay on Spotify, Apple Podcasts, YouTube, and More

Time stamps:

00:00:42 – Introduction
00:02:25 – Maintaining Wasabi
00:07:10 – Decentralization and New Features
00:10:55 – Community Takeover and Comparison with Samourai Wallet
00:15:25 – Adapting to New Paradigm
00:20:57 – Why Choose Wasabi?
00:21:56 – The anonymity and privacy features of Wasabi
00:23:27 – Government running a Wasabi coordinator?
00:24:38 – Underrated features of Wasabi
00:27:37 – Automatic coin selector and advanced coin selection workflow
00:29:38 – Concerns and criticisms about Wasabi
00:32:30 – Address reuse and privacy risks in Wasabi
00:39:15 – Integrating Wasabi with external services
00:42:43 – The daemon integration
00:44:24 – Decentralization of CoinJoins
00:47:49 – Liquidity and coordinator updates
00:50:42 – Coordinator connection and discovery
00:56:10 – Coordinator fees and user protections
01:02:46 – Motivation for working on Wasabi
01:04:29 – Lucas Ontivero's Challenges
01:06:49 – Project Commitment
01:09:52 – Community Support
01:13:30 – Privacy in Bitcoin
01:15:59 – Future Solutions
01:17:32 – Regulatory Concerns
01:21:32 – Maintaining Wasabi
01:25:35 – Hardware Wallet Compatibility
01:26:40 – Community Support and Future Plans
01:27:23 – Technical Control and Graphical Interface
01:28:39 – Acknowledgments and Sponsorship
01:30:10 – Following the Contributors
01:31:17 – Importance of Wasabi for Bitcoin Privacy