Skip to content
This repository has been archived by the owner on Apr 6, 2019. It is now read-only.

Account abstraction and anonymity notes #3

Open
HarryR opened this issue Jul 6, 2018 · 1 comment
Open

Account abstraction and anonymity notes #3

HarryR opened this issue Jul 6, 2018 · 1 comment

Comments

@HarryR
Copy link

HarryR commented Jul 6, 2018

See:

When account abstraction is ready it will be possible to avoid the 'transaction payee undermines anonymity' problem because, in the example of a ring-signature based token or something using stealth addresses, the contract can pay for gas upon providing proof of a correct signature (hopefully, anyway).

However, I am afraid, having payments of different denoms would enable network analysis and leak valuable information about the recipient of a payment...

Yup... that's why occluded values exist in RingCT (as used by Monero), and why the privacy offered by Dash is worse than most others with 'real anonymity'.

Probably better to keep the mixer bounded to a standard denomination (although network/timing analysis could leak info if users are not careful enough when using the mixer for odd denominations that need to be spit into multiple deposits).

Yup... this is used by Dash, in lieu of having a real mechanism of providing anonymity.

An interesting project is https://github.com/barryWhiteHat/miximus

However, this has the same problem of who pays gas for the withdrawal (they're waiting on account abstraction too), they also have all of the problems of transaction amounts revealing the sender / recipient. And an even bigger problem ontop of that - it's not fully trustless as there's a trusted setup phase.

The https://github.com/solidblu1992/ethereum project supports RingCT signatures where the total amount of each payment can be obscured, so somebody can pay you 1 ETH, then you can (at a later date) withdraw 0.25 ETH without being linked to the original 1 ETH deposit. It still suffers from the lack of account abstraction revealing sender/recipient though.

IMO a solution to this problem would be an off-chain relay service, accessible via Tor, I2P or other anonymity layer (which could just be promising to not log the transactions), which allows people to publish their signed transactions along with an incentive that will be drawn from the amount being withdrawn or moved.

@AntoineRondelet
Copy link
Owner

Thanks for the resources! I'll dig a bit into account abstractions when I have a moment! ;)

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants