Re: [bitcoin-dev] Proposal: Bitcoin Secure Multisig Setup

From: Hugo Nguyen <hugo@nunchuk•io>
To: Pavol Rusnak <stick@satoshilabs•com>
Cc: Bitcoin Protocol Discussion <bitcoin-dev@lists•linuxfoundation.org>
Subject: Re: [bitcoin-dev] Proposal: Bitcoin Secure Multisig Setup
Date: Thu, 11 Feb 2021 05:45:33 -0800	[thread overview]
Message-ID: <CAPKmR9vg1BMDQWNDk41N4i4cJ8J6K9GuqSpstFpMFwyiVBYw-w@mail.gmail.com> (raw)
In-Reply-To: <CAF90AvkeG53o5H2dZsdsG_c4PxxooMgx-Fv47RWpNNwm_su-hg@mail.gmail.com>

[-- Attachment #1: Type: text/plain, Size: 3169 bytes --]

Hi Pavol,

On Thu, Feb 11, 2021 at 5:25 AM Pavol Rusnak <stick@satoshilabs•com> wrote:

> > ENCRYPTION_KEY = SHA256(SHA256(TOKEN))
>
> This scheme might be vulnerable to rainbow table attack.
>

Thank you for pointing this out! Incidentally, Dmitry Petukhov also told me
the same privately.

>
> The following scheme might be more secure:
>
> DESCRIPTION = ASCII description provided by user
> NONCE = 256-bit random number
> ENCRYPTION_KEY = hmac-sha256(key=NONCE, msg=DESCRIPTION)
>
> Coordinator distributes DESCRIPTION (fka TOKEN) together with NONCE to
> the signers.
>

This does seem to add a lot more entropy. The challenge is to balance the
security requirement with UX. In the absence of some handshake protocol to
exchange the shared secrets (DESCRIPTION / NONCE) , the user will have to
enter these manually on the devices. I'll think about this some more.

>
> Also, is there any reason why you'd want to disable encryption? Why not
> keep that as mandatory?
>

Making it mandatory would be nice, but IMHO not all use cases might require
encryption. For example, if you are setting up the multisig locally under a
safe environment you control, encryption might be an overkill.

Best,
Hugo

>
>
> On Tue, 9 Feb 2021 at 12:39, Hugo Nguyen via bitcoin-dev <
> bitcoin-dev@lists•linuxfoundation.org> wrote:
>
>>
>>
>> On Tue, Feb 9, 2021 at 2:19 AM Christopher Allen <
>> ChristopherA@lifewithalacrity•com> wrote:
>>
>>>
>>>
>>> On Tue, Feb 9, 2021 at 2:06 AM Hugo Nguyen <hugo@nunchuk•io> wrote:
>>>
>>>>
>>>> I don't think reusing XPUBs inside different multisig wallets is a good
>>>> idea... For starters, loss of privacy in one wallet will immediately affect
>>>> privacy of other wallets. I think multisig wallets should be completely
>>>> firewalled from each other. That means one unique XPUB per wallet. This is
>>>> what we have been doing with the Nunchuk wallet.
>>>>
>>>
>>> To be clear, I have stated repeatedly that xpub reuse into multisig is a
>>> poor practice. However, finding a trustless solution when a wallet is
>>> airgapped with no network, or is stateless like Trezor, is quite hard.
>>>
>>> The challenge also includes how does an airgapped or stateless wallet
>>> know that it is talking to the same process on the other side that that it
>>> gave the xpub to in the first place. Without state to allow for a
>>> commitment, or at least a TOFU, a cosigner who thought he was part of a 3
>>> of 5 could discover that he instead is in a 2 of 3, or in a script with an
>>> OR, as some form of scam.
>>>
>>
>> The shared secret approach that I mentioned in the proposal actually can
>> help you here. The TOKEN doubles as a session ID - thereby establishing a
>> common state on both sides.
>>
>> Best,
>> Hugo
>>
>>
>>>
>>> — Christopher Allen
>>>
>>>> _______________________________________________
>> bitcoin-dev mailing list
>> bitcoin-dev@lists•linuxfoundation.org
>> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
>>
>
>
> --
> Best Regards / S pozdravom,
>
> Pavol "stick" Rusnak
> CTO, SatoshiLabs
>
>

[-- Attachment #2: Type: text/html, Size: 6239 bytes --]