AvailableCoins sets raw pubkey outputs with an OutputType::UNKNOWN type instead of OutputType::LEGACY.
Gladly, no type unknown outputs are skipped during coin selection, nor anything different is done with them. It is just an inconsistency.
AvailableCoins sets raw pubkey outputs with an OutputType::UNKNOWN type instead of OutputType::LEGACY.
Gladly, no type unknown outputs are skipped during coin selection, nor anything different is done with them. It is just an inconsistency.
Raw pubkey outputs are `OutputType::LEGACY` not `OutputType::UNKNOWN`.
<!--e57a25ab6845829454e8d69fc972939a-->
The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.
<!--021abf342d371248e50ceaed478a90ca-->
See the guideline for information on the review process. A summary of reviews will appear here.
I'm not sure if that's really a bug so long as we aren't excluding them from coin selection. I don't think we actually classify p2pk outputs as being legacy since they don't have an address. Really legacy is talking about address type.
Additionally, in descriptors, pk() descriptors are treated as having no output type, rather than unknown or legacy.
I'm not sure if that's really a bug so long as we aren't excluding them from coin selection. I don't think we actually classify p2pk outputs as being legacy since they don't have an address. Really legacy is talking about address type.
Yeah, so far we aren't excluding them from coin selection but hmm, the more I think about this, the odder is turns for me. E.g. we mix the "legacy" output type usage for the address encoding and also to create and handle the legacy spkm. Because, in parallel to the legacy spkm (which could return two different address encodings), we can also have single p2pkh or p2sh descriptors that returns the same "legacy" output type.
But ok, that can be left aside. It should get better once we deprecate the legacy spkm.
Additionally, in descriptors, pk() descriptors are treated as having no output type, rather than unknown or legacy.
It's odd to be classifying a known spendable output type as unknown. I wouldn't be surprised if someone proposes to skip them during coin selection thinking that they are non-spendable in the future (or thinking that the wallet wasn't able to recognize them so it will not be able to create an input that spends them etc).
We should document this stuff better somewhere.
closing since it's not a bug, just an odd classification.