Re: Encryption algorithm used for the IBMi OS passwords.

Again, I'm not arguing against published algorithms which are vetted "in
the open". I'm suggesting that using a private algorithm crafted by an
expert may be more secure than algorithms which have been published.

It may be, but you don't know until it has been tested. How many encryption algorithms go out for vetting that pass with flying colors the first time with no changes required? I don't know the answer to that. But unless these algorithms are supremely easy to create, I would suspect the answer is "not many", yet they are all built by cryptology experts. Up until the first time an algorithm is sent out for vetting, it is equivalent to a private algorithm crafted by an expert. So are you saying that the vetting process makes any algorithm necessarily less secure? I would argue that instead, the vetting process simply brings to light insecurities in the algorithm sooner, preferably before it is released for use. It follows that, more of the private algorithm's insecurities are necessarily not discovered until it is in use. I am not saying that open algorithms do not have insecurities that are discovered after being placed in widespread use, just that the most obvious bugs have already been found out. I guess it is just a matter of the quality and quantity of vetting an algorithm receives before distribution.