We should add a test where we generate a signature for each of the 36 parameter sets, and compare that to known good signatures we have harvested from the reference code.
Of course, it'd be just as good (and far more practical) if we compare the hashes, rather than the entire signatures.
We should add a test where we generate a signature for each of the 36 parameter sets, and compare that to known good signatures we have harvested from the reference code.
Of course, it'd be just as good (and far more practical) if we compare the hashes, rather than the entire signatures.