Hash.WriteAny should employ domain separation
cronokirby opened this issue · comments
Right now, it just hashes things based on their bytes, but this could theoretically cause problems, because a Curve Point might affect state in the same way as a byte slice with the same pattern. It would be a good idea to add in some extra bytes when hashing each of these, so that different sequences of these objects can't be confused with each-other.
I don't think this presents a concrete issue, but it's a good best practice to follow, and pretty simple to implement.