[Feature Request] Factory-reset PIN

[snufon]

Optionally set a trick PIN, which factory resets the device if inputted.

If the user is under duress, an attacker can force them to input the right PIN. Right now, the factory reset only happens after 10 wrong PINs. The attacker would take action long before that.

Coldcard has this feature, with the exception that it bricks the device instead of simply wiping the memory. The factory-reset seems much better.

[NickeZ]

A few thoughts:

• If you reset the device in front of an attacker they will force you to get the backup.
• You will never remember the pin unless you practice it monthly (weekly?).
• During the threat of physical violence you will probably just open the device correctly anyway to avoid losing body parts / family members.

What do you think?

[snufon]

Thanks for your reply.

• Good security would prevent that. A good backup is m of n and is not all accessible from the same place (e.g. one part in a bank vault)
• That trick pin does not have to be complex since we don't care it is bruteforced
• Wiping the device's memory means the attacker has no more incentive to do anything. Plus some people would rather die than to give their money to evil people. A good backup strategy also guarantees your closed ones get your BTC after your death, you don't want that to be stolen

[NickeZ]

How about hitting the device with a hammer or throwing it in the toilet, fireplace?

BB02 doesn't have battery, so you would need to plug it into something to use the factory-reset pin.

If you think you can handle the torture, why would you give them the password in the first place? An option is also to not give them the password, or tell them it isn't your device.

Still, how will you remember if you picked 1111 or 9999 or whatever, 5 years after initializing the device?

[snufon]

The BB02 aims to provide and markets itself as having top-tier security, suitable for paranoid users who might be securing millions worth of Bitcoin on the device. From what I know it fulfills the promise outside of this single feature

• The user is under duress from potentially professional attackers, those actions are not possible
• The trick PIN is inputted instead of the real one, so the BB02 is necessarily plugged-in
• On the contrary, the point is that you don't handle the torture. This wipes the device while under duress to make sure the money cannot be stolen no matter what happens next
• That simple PIN is part of of the user's security strategy which potentially secures all of their life-savings. Why would remembering this PIN be a problem? You must remember your true PIN anyway, so remembering the simpler trick PIN is trivial.

I checked and the Blockstream Jade has this feature too (not only Coldcard). That's probably the case of other hardware wallets I don't know yet. The feature was not implemented without reason. Even if actually needing to use this trick PIN is unlikely, this can definitely be a deciding factor when choosing which wallet to buy. Why take the risk when the different wallets basically cost the same?

[NickeZ]

I'm personally not convinced that it is useful. If you have a lot of wealth and your threat model includes physical violence then I think you should think about physical protection like home security or personal alarms so that you can avoid getting hurt at all. If you are storing generational wealth you should probably have a setup requiring more than one bitbox (multisig or MPC).

As you note other hardware wallets implement it and that could be reason enough to do it, basically check the feature checkbox. We have a lot in our backlog, so even if we want to do it we cannot promise by when.

[benma]

I don't have a strong opinion on the feature, but wanted to also add @snufon that if you want to protect life savings, it would be wise to avoid a single point of failure and to look into multisig or Liana. That way, a need for a duress PIN is gone, as there is nothing for the attacker to get right away anyway. There are additional benefits, such as avoiding a single point of failure in the backup.

[NickeZ]

Trezor calls it "Wipe code": https://trezor.io/learn/security-privacy/personal-security-standards/create-wipe-code-to-erase-device

[snufon]

Yeah I definitely get that it's not trivial to implement and not high priority. Something that would maybe be simpler to implement would be to allow the user to customise the number of PIN retries allowed. A user could set it to lower than the default 10 tries and this kinda gives the same utility as a trick PIN (set it to 0 or 1).

I personally do not want to deal with multi-sig for multiple reasons. I use the hardware wallet often to spend/receive, so only needing one hardware wallet and a simple PIN is very practical. A trick PIN easily removes the only single point of failure I have (my backup is M of N so no worries there). I got a Coldcard I will use in the meantime (no Taproot support tho, you guys are better on that front).

I will take a look a Liana, thanks for the quick replies!

[benma]

Thanks for your feedback. We will definitely consider it.

I use the hardware wallet often to spend/receive,

Just fyi, you can safely receive on a multisig wallet using one BitBox only, as the descriptor is registered on it. If you spend often, then yes multisig is definitely not convenient. There are also solutions where one has a "spending" single-sig wallet with less funds on it and the multisig containing the majority of the funds.

my backup is M of N so no worries there

Out of curiosity, how did you achieve that, seeing that the BitBox currently does not offer splitting backups like that?

[snufon]

I'm not using Shamir Secret Sharing because of an annoying lack of standard across wallets last I checked, plus it's more complex to explain to potential heirs. Instead, the backup is a 2 of 3 where each location only contains 16/24 words. While the fact that the shares are not encrypted seems scary, it's fine since 8 words is impossible to brute-force (and would cost more money than what can be stolen anyway).

Interesting note: since the last 8 words are theoretically the easiest to brute-force because of the checksum in the last word, a small optimization I did was to hide the share which lacks those words in my most secure location

[cedwies]

I can see some advantage of having such a feature. However, even though "the BB02 aims to provide and markets itself as having top-tier security", it also markets itself as being easy to use while providing maximum security. That being said, there are numerous features that can be implemented for "advanced" users. From a "wipe code", over a fake wallet to having a custom number of tries before the device resets. My personal opinion is that it's easy to increase complexity with little to benefit, and one has to be careful to not unnecessarily confuse others. I think if we do more of such advanced features they have to be very clearly separated from the basics. Very clearly.

[NickeZ]

We have created an issue in our internal issue tracker for this. But for now it will remain low prio in our backlog.

[NickeZ]

@snufon The cold card has like 14 different behaviors of their trick pin. Any behavior you are specifically thinking about?

https://coldcard.com/docs/pins/#trick-pins

[snufon]

Ah interesting, their main page (coldcard.com) advertised only some of those. To me, the only interesting one is simply to wipe the seed when the trick PIN is inputted. (Or let the user customise the number of wrong tries before the seed is wiped)

The others seem redundant, because the only other strategy under duress is to trick the attacker into stealing a secondary wallet, and the BB02 can already do that with the optional passphrase feature. Bricking the device seems 100% worse than wiping the seed