Message-ID: <21c0cdea-f929-b9a8-baa6-e33eb2cee80f@riseup.net>
Date: Mon, 16 May 2022 11:26:45 +0100
MIME-Version: 1.0
Content-Language: en-US
From: Chris Belcher <belcher@riseup.net>
To: bitcoin-dev@lists.linuxfoundation.org
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Subject: [bitcoin-dev] Improving chaumian ecash and sidechains with fidelity
 bond federations
Precedence: list

Hello list,

Fidelity bonds could be used to help create trust-minimized federations 
that are needed for things like chaumian ecash servers or sidechains.

 From what I've seen until now, people working on chaumian ecash or 
sidechains say that the federation controlling the multisig keys will be 
based on some kind of reputation. Perhaps it will be some pseudonymous 
nyms that have built up a good reputation over a long time. I suggest 
another option is to use fidelity bonds to decide who gets to control 
the multisig keys.

Fidelity bonds are a way to deliberately sacrifice bitcoin value in a 
way that can be proven to a third party. In practice this is done by 
sending bitcoins to an address which is time-locked using the 
OP_CHECKTIMELOCKVERIFY opcode. The redeemscript and UTXO, along with a 
signature, can be shown to anyone to prove that the sacrifice happened. 
This system has already been deployed in JoinMarket since August 2021, 
and at the time of writing about 600 btc have been locked up, some for 
several years. The whole scheme is similar in some ways to PoW that 
bitcoin itself uses to avoid sybil attacks when solving the double spend 
problem.

It's important to understand what is the value-add of fidelity bonds and 
what it isn't. Fidelity bonds don't solve the trust issue, as someone 
with a big fidelity bond could still steal funds from the ecash server 
or sidechain using multisig keys they control. Such systems will always 
be custodial.

Rather, fidelity bonds strongly incentivize that the different fidelity 
bond owners are actually different people. That might be exactly the 
kind of thing needed for distributing the keys of big multisigs, 
especially now that taproot allows us to create very big multisig 
schemes. This happens because the value of a fidelity bond is calculated 
as a greater-than-linear power of the bitcoin sacrifice. So for example 
if the power was 2, and someone sacrificed 5 bitcoins of value, their 
fidelity bond would be worth 5 x 5 = 25. If instead they sacrificed 6 
bitcoins their fidelity bond would be worth 6 x 6 = 36. This superlinear 
power is what creates a strong incentive for the different fidelity 
bonds to actually be controlled by different people, because anyone 
behaving rationally will put all their bitcoins into just one fidelity, 
not split them up over many bonds. As a sybil attacker needs to 
distribute their bitcoins over many different bonds, they are 
mathematically punished. The fidelity bond system achieves this without 
revealing anything much about those people's identities.

Another value-add of fidelity bonds is they are very much in keeping 
with the cypherpunk ethos, as anyone can create a fidelity bond and 
advertise it in the market. As the bitcoins can be mixed with coinjoin 
before and after sending to the timelocked address, the scheme doesn't 
have to be linked to any identity. Only money talks; not reputation, 
political power or geographical power.

I don't know yet exactly the details of how such a scheme would work, 
maybe something like each fidelity bond owner creates a key in the 
multisig scheme, and transaction fees from the sidechain or ecash server 
are divided amongst the fidelity bonds in proportion to their fidelity 
bond value.

Regards
CB