You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It's not yet entirely clear to me how this should like in the end.
Will we include the full code of the package in the BIP? Or just refer to it, possibly pinned to some commit. The latter sounds cleaner.
The package should at least have docstrings for all objects we import, to clarify the operations we rely on (like in BIP340, where we define stuff like lift_x). But then, if we don't include the full contents of the package, then the BIP is technically speaking not fully self-contained... Is that a problem? In theory, we could also run pydoc on the package and include the output in the BIP. But that sounds a bit like overkill.
One additional insight is that we could also have a git subtree with the package. That would make the BIP self-contained (i.e., no external resources in external repos).
We think it's best to discuss this with the BIP editors before we take a decision.
We discussed offline that requiring an external dependency is fine if secp256k1proto becomes a package. For now, secp256k1proto could also be a submodule instead of a package. Whatever is easier until we have a final deisionon this.
"secp256k1-proto" sounds good to me
The text was updated successfully, but these errors were encountered: