1
0
forked from cheng/wallet
wallet/docs/design/mixer.md

84 lines
3.7 KiB
Markdown
Raw Permalink Normal View History

---
title: >-
Bitcoin mixer.
sidebar: false
notmine: false
abstract: >-
I want to get early adopters to use my planned privacy
social net, as a small step to replacing SWIFT.
all existing bitcoin mixers have been shut down because
they relied on a centralized social net,
which depended on domain names, which resulted
in the fbi arriving at address of the owner of the
domain name
---
Samourai wallet migrated to centralisation, which directly led to them getting
busted. That codebase is poisoned with communications that the FBI
has flagged as actionable. And integrating anything into current
Bitmessage is impossible except for the original developer. No
big Python program is truly open source.
The correct design for a mixer is as follows. One has a social net,
on which anyone can offer to coordinate a single mixing transaction.
for a mix that will produce mixed coins (utxos )of a particular
round number, 10mBTC, 20mBTC, 50mBTC, or 100mBTC,
plus unmixed changed coins.
All the mixed coins are of equal value, for example all 100mBTC.
Not some funny value highly identifiable value like 99.9872384mBTC
People offer to contribute utxos to this mix transaction - revealing
to the coordinater the public keys, the address, of the utxos,and
revealing to the coordinator that these utxos have a common owner.
They also give him the blinded addresses of coins they want to
receive. He blindsigns those addresses. They then reveal the unblinded
addresses, and his unblinded signature, which proves he signed those
addresses, but does not reveal to him which of the addresses he blindsigned it
is -- he does not learn the relation between the utxos that will be contributed
to the mix transaction, and the mixed or the change utxos that it will replace
them. (Though he and anyone doing blockchain analysis can trace the
change coins by the sudoko attack. But the sudoko attack is irrelevant to
coins that are all the same round number of bitcoin, such as 10mBTC)
He then creates the transaction, and everyone signs it. If not everyone
signs, everyone can see what the missing utxos were, the ones that
were promised, and not delivered, and blacklist them,
then try again.
People contributing already mixed utxos do not have to pay transaction fees
so get back exactly what they contributed.
People contributing as yet unmixed bitcoin have to pay a portion of the transaction
fee proportional to the number of utxos contributed and received. This is good for
them because the free of charge remixed utxos are enlarging their anonymity pool.
making each mixing transaction part of one enormous anonymity pool instead of many
tiny anonymity pools.
If mixing does not work like this, then someone has fucked it up in order to profit from
it, their users will be traced, and *they* will be traced, then arrested.
> > The core of my plan has always been Web 3.0, a privacy social net,
> > and everything else is just monetization, because software never
> > gets done properly or properly maintained without someone making
> > money off it.
> I got what you mean now. Once you reach a point that's indeed a good
> strategy to reinforce value of the network. You offer those
> integrate your service to beat metcalfe's law, your network becomes
> much more stronger. Facebook and some other social networks all
> followed this path via. 'applications' within them.
> > I have been trying to do that, but it is hard to get to the front of
> > the line of all the people who want to tell blockstream why
> > blockstream should fund them and their projects.
> Have you really tried? Adam must have had some fidelity to
> cypherpunks.
I have not tried, but I have been looking for entry points, and
have come up empty. One has to have an in, and one has
to listen before one speaks.s