Join GitHub today
GitHub is home to over 20 million developers working together to host and review code, manage projects, and build software together.
Unified partial signature format #10674
Comments
|
Wouldn't this be a BIP? |
|
@dcousens Once completed, I'd say yes. But here I'm just asking for some brainstorming about the approach. |
|
Why would the full previous transaction being spent from required for non-witness transactions? Isn't just the UTXO enough? |
|
@achow101 The transaction creator could lie about the UTXO amount, resulting in you signing over a huge amount of money into fees. You need the full transaction to know that the fee is what is claimed. This is the reason why hardware wallets require seeing the full transactions being spent from. In BIP143 that is no longer necessary, as the full UTXO being spent is included in the sighash. If you lie about UTXO amount, the signature won't be valid. |
sipa commentedJun 27, 2017
The current
createrawtransaction+fundrawtransaction+signrawtransactionworkflow is complicated by the fact that there is some information that needs to be passed out of band:These are currently handled transparently by the fact that
signrawtransactionhas access to the wallet transactions, scripts, and node UTXO set. However, third party signers do not necessarily have access to this information.Given that it is known in advance what signers will need when the unsigned transaction is constructed, I wonder if the easiest approach isn't creating a more generic partial transaction format that can store multiple extra records for each input (which don't go into the final transaction):
This would also simplify extending to more complicated script types in the future.