Hi,
when a new extended key is derived, the depth can grow over 255. Such an extended key is later not serializable, because python-hdwallet's code raises an exception ValueError: byte must be in range(0, 256). This behavior is in compliance with BIP32's serialization format, however, I wonder if keys with greater depths should be allowed to be derived in the first place at all. See also this bitcoin/bitcoin#32201 for occurrences in other libraries.
The MWE can be quickly created by using the example from the README.md and using "m" + "/0" * 256 and dumping the wallet.
(The first link also hints that Cardano derivation variant might also be affected.)
Is it intentional to allow deriving keys with depth over 255 that cannot be serialized?
Hi,
when a new extended key is derived, the
depthcan grow over 255. Such an extended key is later not serializable, because python-hdwallet's code raises an exceptionValueError: byte must be in range(0, 256). This behavior is in compliance with BIP32's serialization format, however, I wonder if keys with greater depths should be allowed to be derived in the first place at all. See also this bitcoin/bitcoin#32201 for occurrences in other libraries.The MWE can be quickly created by using the example from the
README.mdand using"m" + "/0" * 256and dumping the wallet.(The first link also hints that Cardano derivation variant might also be affected.)
Is it intentional to allow deriving keys with depth over 255 that cannot be serialized?