2

What are the trade-offs involved in:

  1. Buyer and seller signing the transaction. HodlHodl signs the transaction only in case of dispute.

Vs

  1. Seller and HodlHodl signs the transaction. Buyer signs the transaction only when dispute.

According to https://gitlab.com/hodlhodl-public/hodl-client-js/-/blob/master/multisig-spec.md

Server generates release transaction with output address taken from user's trading settings. Server signs it's part of the release transaction input with it's private key.

Server signature is always goes first expecting client signature as next chunk. Client decrypts it's seed with payment password. Client signs the transaction with it's private key derived with index and sends back to server. Service broadcasts provided transaction.

NOTE: In both examples below transaction is signed by seller. Buyer has to sign transaction only if contract is disputed and resloved in favor of buyer.

Will the trade-offs remain same for any other similar bitcoin projects using 2 of 3 multisig?

2

To state the obvious in a 2-of-3 multisig escrow if two parties collude they can cheat the honest third party.

Buyer and seller signing the transaction. HodlHodl signs the transaction only in case of dispute.

In this scenario it is less work for HodlHodl and less trust/reliance on HodlHodl. You would choose this if you don't trust HodlHodl any more than the buyer and would rather only rely on HodlHodl in the case of the buyer becoming unresponsive.

Seller and HodlHodl signs the transaction. Buyer signs the transaction only when dispute.

In this scenario it is more work for HodlHodl and more trust/reliance on HodlHodl. You would choose this if you trust HodlHodl more than you do the buyer and HodlHodl are willing to get involved in a greater number of escrow releases. If they can avoid it HodlHodl would rather not get involved as when they do there is the possibility of mistakenly or maliciously releasing the money from the escrow when it shouldn't yet be released or releasing it to the wrong party.

Will the trade-offs remain same for any other similar bitcoin projects using 2 of 3 multisig?

Yes. As long as there are three keys being distributed to three parties there is this concern of two parties colluding to cheat the honest party. Note that some 2-of-3 multisig use cases distribute all three keys to the same individual so the individual can simultaneously protect their funds and offer redundancy in case the individual loses one of the keys. Other 2-of-3 multisig use cases (e.g. BitGo) give two keys to the same individual with BitGo keeping the third key. In this case BitGo can't steal the funds without obtaining an additional key from the individual.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.