Research
Timechain
Search Donate
Sponsors
Cake Wallet, private Bitcoin and Monero walletSideShift.ai, no-signup crypto exchangeOrangeRock, non-custodial bitcoin spot and perps tradingBraiins Hashpower, buy and sell Bitcoin hashrateLayer Two Labs, drivechains for Bitcoin
S8 · E7 · 1:56:17

Lixin Liu on the Cobo Vault Hardware Wallet (S8 E7 Full Article)

30 Apr 2021 · by Vlad Costea · 7 min read
Published at block 681,165
Cover art for S8 E7, Lixin Liu on the Cobo Vault Hardware Wallet
MP3 (160 MB)

The audio keeps playing while you browse the site.

Read the transcript15,585 words, every line links to the minute it was said

Ask this episode

Embed this episode

Paste this on your site or blog. It is a small player that streams from here: no trackers, no cookies, no ads.

Preview the player

Lixin Liu first came on the show in February 2020 to announce the second generation of the Cobo Vault. A year later, during the hardware wallet season, I wanted to know how it had gone. He started by correcting me. "I'm not the CEO of Cobo," he said. He's the head of hardware, leading the team that builds the device. Cobo later renamed the hardware wallet Keystone, which is why both names appear on this site.

We talked about everything: design choices, QR codes, the chip shortage, open source, and how Lixin thinks about the rest of the industry. The original show notes have the short version.

Designed Against Human Error

Lixin's design philosophy starts with one idea. "Human error is the biggest reason for Bitcoin loss." Most hardware wallets "have tiny screens or tiny buttons, which is really bad for user experience." So Cobo gave the Vault "a big touchscreen, so that everything will be displayed very clearly," because people are already used to their phones.

He also thinks about passwords differently. Phones use four or six digit PINs, but a hardware wallet should encourage strong passwords with "numbers plus characters plus symbols," and that only works if typing is easy.

Why QR Codes

The Cobo Vault talks to other wallets through QR codes, and Lixin gave three reasons. "It's totally air-gapped," because you never connect it to a laptop that "may have some malicious code." Second, "everything getting in and out of your hardware wallet is auditable," since anyone can scan the codes with a third-party scanner. Third, compatibility: "you just need to scan with your camera."

His security mindset is honest. "There is no absolute security, there is only relative security." With USB or Bluetooth, "the information getting in and out of the device, you cannot see that." The downside is size: large multisig PSBTs can need "tens of QR codes or even 100 QR codes," and that's where a microSD card helps.

Thanks to PSBT, the Vault works with many software wallets. Lixin said most users pair it with BlueWallet for single-sig and Specter for multisig. It also does multisig with Coldcard, and both devices store the other co-signers' xpubs so they can verify an unsigned transaction.

From Chinese Miners to Western HODLers

The first Cobo Vault was rugged, built "mainly for miners in China." But during the bear market, Cobo found that "most people in China, they just do trading and they put most of their bitcoins in exchanges." At Bitcoin 2019 in San Francisco, Lixin met many American Bitcoiners and decided "we should pivot our targeted market from China to Western countries," where people hold Bitcoin "as a very serious investment."

The Chip Shortage

We recorded during the 2021 chip shortage, and Cobo felt it. Manufacturing lead times went from 45 to 60 days to "over 90 days." Costs exploded too. "The manufacturing cost is three times more than the cost in 2018. You cannot imagine. Three times. Not 30%, but three times." He blamed part of it on the mining boom, which made even hard drives and video cards scarce in China.

How Open Is It?

I asked about open source, and Lixin didn't oversell. "To be honest, we're not the most open source one. But what I can say is that we have open sourced everything we can open source." He's proud of one first: "we are the first hardware wallet that open sourced the firmware of the secure element," and the first to publish its audit report.

The secure element and microcontroller carry the Cobo name, but Lixin clarified, "we're not making those chips." They hide the real manufacturer because naming it could invite "social engineering on the developers of the chips."

They also work with white hat hackers, including Lazy Ninja. When the "fee attack" appeared, where a software wallet could trick a user into signing a transaction twice and paying a huge fee, Cobo proposed a fix, and "both Coldcard and us were implementing some very similar solutions." Lixin wants to "make it possible for our users to build their own Cobo Vault by themselves." The device runs on Android plus a secure element. Lixin argued that a different base avoids a monoculture: if everyone builds on the same base, "if there's some risk in the base, then every tower will collapse." Still, Cobo was "removing Android layer" for future devices.

Bitcoin First, as a Business Choice

He gets "tons of requests" to support coins that Ledger and Trezor don't list, and he refuses. "As a business owner, I still cannot agree with this kind of strategy." His reasoning stuck with me. "The biggest competency of the hardware wallet is time." "It's not your secure element. It's not your big screen or something. It's time." A hardware wallet needs a plan of at least three to five years, not "shitcoins that may only last for several months."

Pro, Essential and Self-Destruct

The second generation comes in two versions with the same firmware. The Pro adds "the rechargeable battery," "the fingerprint sensor" and "the self-destruct mechanism." Since a Pro is more likely to be used outside the house, it's more likely to be stolen. If someone disassembles it for a side-channel attack, "your private keys and other sensitive information will be all erased from the device." The Essential is for people who rarely touch their wallet.

Friends, Not Competitors

Lixin was generous about the competition. "I really respect Trezor guys," and he called the Model T "the second best hardware wallet currently on the market for user experience." He doesn't see Coldcard, Trezor or Ledger as rivals. "We should stay together to build a defense line to defend against all the hackers." He and NVK talk privately about security: "I think we're friends and we talk frequently."

For Lixin, good design is also about calm. "You should not make people anxious about using the product. Then the people will make mistakes."

Privacy and Shamir Honeypots

To check that a device is genuine, the Vault's secure element talks to a verification website backed by "Amazon's cloud HSM," with matching keys preloaded at the factory. The companion app never stores your xpub on Cobo's servers, because "if we have your xpub in our server, we will actually know all your Bitcoin addresses." They only store the wallet fingerprint. Shipping data is deleted after six months, the usual chargeback window.

He doesn't see Shamir backup and multisig as rivals either: "you can use either of them or you can use both of them simultaneously." Before Shamir, most people simply copied their seed to several places, which protects against floods but makes theft more likely.

My favorite trick was the Shamir honeypot. With SLIP-39, you can extend a 20-word share into a valid 24-word phrase, put a little bitcoin on it and store it at a friend's place. If those coins move, you know someone looked. "You can turn one piece of the Shamir backup into a honeypot."

But his main advice is simple. "Don't make it overcomplicated. Don't shoot yourself." Choose "some solution that you can manage."

What Came Next

Lixin shared the roadmap: integrations with Caravan and Casa, then "a more open source, dedicated Bitcoin-only device, which removes the Android layer" while keeping QR codes, microSD and the big touchscreen. "We are keeping an eye on Taproot," he said. Cobo later released that device under the Keystone name, and Taproot activated in November 2021.

He also imagined a safer way to buy hardware wallets: a world tour of Bitcoin conferences where makers "sell our product directly to the end user with cash."

What Makes This Episode Worth Listening To

You get the reasoning behind one of the most different-looking hardware wallets, from the person who builds it. Lixin is candid about what's open and what isn't. He also keeps coming back to the one thing that loses most bitcoin: human error.

You can read the full transcript, the original show notes, or visit Lixin Liu's guest page.

Listen to Lixin Liu on Spotify, Apple Podcasts, and More

Time stamps:

  1. 00:03:24– Head of hardware, not CEO
  2. 00:04:58– Human error and big screens
  3. 00:06:19– Three advantages of QR codes
  4. 00:08:33– Multisig with Coldcard
  5. 00:10:14– Miners and traders in China
  6. 00:11:39– Pivoting to the West
  7. 00:15:16– PSBT wallets
  8. 00:17:41– Relative security
  9. 00:19:13– When QR codes get huge
  10. 00:20:57– Lead times over 90 days
  11. 00:21:44– Costs three times higher
  12. 00:26:00– Why the chip makers are hidden
  13. 00:26:55– Build your own Cobo Vault
  14. 00:29:12– Open sourcing the secure element
  15. 00:34:29– Not the most open source
  16. 00:36:55– Why no altcoins
  17. 00:37:37– Time is the real advantage
  18. 00:42:35– Pro vs Essential
  19. 00:44:40– Self-destruct
  20. 00:49:31– Android and monoculture
  21. 00:52:02– Respect for Trezor
  22. 00:57:06– Strong passwords need typing
  23. 01:15:03– A defense line against hackers
  24. 01:17:45– Don't make people anxious
  25. 01:21:06– Verifying with a cloud HSM
  26. 01:24:40– No xpubs on the server
  27. 01:26:13– Shamir and multisig together
  28. 01:28:56– A Shamir honeypot
  29. 01:34:06– Don't overcomplicate it
  30. 01:41:43– Deleting shipping data
  31. 01:48:01– Buying with cash at conferences
  32. 01:51:30– A Bitcoin-only device
  33. 01:52:25– Keeping an eye on Taproot

Verify this episode

This is the SHA-256 fingerprint of the MP3 as published here, timestamped in the Bitcoin blockchain with OpenTimestamps. If one byte of the file changes, the fingerprint changes.

SHA-256
65c85ab663109db4417473703448a197c49818e20a3bea00172e72b224fc2002
Proof
2021-04-23_btctkvr_s8_e7_lixin_liu_on_cobo_vault_hardware_wallet.mp3.ots · all episode hashes (proof)
Status
Anchored in Bitcoin block 974,312
Timestamped
After release. The episode came out on 23 Apr 2021, but this proof was made on 3 Nov 2026 and anchored in block 974,312, so it proves this exact file existed by 3 Nov 2026, not on the release date.
How to check it yourself
  1. Download the MP3 and the proof into the same folder.
  2. Run sha256sum 2021-04-23_btctkvr_s8_e7_lixin_liu_on_cobo_vault_hardware_wallet.mp3 (on a Mac: shasum -a 256 2021-04-23_btctkvr_s8_e7_lixin_liu_on_cobo_vault_hardware_wallet.mp3). The result must match the SHA-256 above.
  3. Install the OpenTimestamps client (pip install opentimestamps-client) and run ots verify 2021-04-23_btctkvr_s8_e7_lixin_liu_on_cobo_vault_hardware_wallet.mp3.ots. It hashes the MP3 again and checks the proof against your own Bitcoin node, then prints the block and the date the file existed by. Without a node, ots info 2021-04-23_btctkvr_s8_e7_lixin_liu_on_cobo_vault_hardware_wallet.mp3.ots shows the block height to look up in any block explorer.

Share: X · Email · Link

Related

Comments

Comments run on my own server. No trackers, no tracking cookies.

0:000:00