Closed DanGould closed 4 months ago
As a more long-term goal, it would be great to have integration tests across a couple of other popular wallets. It seems like relying exclusively on bitcoind
in tests is letting many little things slip through that other wallets can't handle (i.e. bitcoind doing its job a little too well).
it would be great to have integration tests across a couple of other popular wallets
The easiest route to test more integration support would be bitcoin dev kit since there are already integrations to copy in the wild. Automated e2e testing of the supported wallets in the readme would also be amazing, but is definitely a lift.
BIP 78 (Payjoin V1) does not follow BIP 174 (PSBT V1) because it removins UTXO data from PSBT inputs, which a sender needs to sign the augmented Payjoin PSBT. It only keeps the input OutPoints. Many PSBT signers will not sign PSBT Inputs with unspecified input UTXO data. Payjoin Dev Kit may not have noticed this at first because
bitcoind
does sign inputs by looking up OutPoints, but other wallets still have problems.The send module of the Dev Kit maintains the
original_psbt
including sender UTXO data and should add it back to the Payjoin PSBT before the sender signs to support as many wallets as possible. This would solve the protocol shortcoming for users of the Dev Kit without patching BIP 78.These wallets are confirmed unable to sign inputs without PSBT UTXO data
LND Sparrow BDK