Re: Should be in crypto for John E. Hadstate Re: just stupid?

From: Joe Peschel (
Date: 07/29/05

Date: Fri, 29 Jul 2005 20:02:24 -0000

"Terry Ritter" <> wrote in

> Joe Peschel wrote:
>> No! Ritter was cited an as expert by this CryptoSMS fellow. His was
>> an "Appeal to Authority" argument, and was in this case, a logical
>> fallacy.
> Is every statement which gives a name an appeal to
> authority error? Clearly not, and the distinction
> is obvious:

You're right, not every statement that gives a name is an appeal to
authority. You were cited, however, as an authority. I believe that such an
appeal is fallacious in that you are clearly not an authority, and you even
admit it by describing yourself as an engineer.

> As I see it, my name was cited for my particular
> conclusions on Multiple Encryption, all of which
> were developed from facts and reasoning, all of
> which is extensively documented on my pages,
> starting with my Glossary.

> That is not Authority,

Nonsense! You or you work was cited as being authoritative. Clearly, you do
not have credentials as a cryptologist, so your opinions lack the weight of

> that is shorthand for an entire body of available
> work.

I'm afraind most of your work is unpublished, and much of your opinion on
layering ciphers is, frankly, just that, opinion.

>> In other words, if an authority is cited as an expert on the subject,
>> he ought to actually be an expert on the subject. In this case, the
>> issue is cryptology, and Mr. Ritter is an engineer and,
>> unfortunately, not a cryptologist.
> I doubt Shannon would have called himself "a
> cryptologist."

Irrelevant. We're talking about your credentials. On the other hand, I
doubt that Shannon would call himself a cipher designer.
> Vernam actually was an engineer when he came
> up with mechanical stream ciphering which then
> led to the academically-loved one-time-pad.

Irrelevant. We're talking about your credentials.

> I do in fact claim expertise in particular areas
> of cryptography, based on over a decade of work
> and publication, including refereed articles and
> issued fundamental patents in the field.

Do you claim expertise on the concatenation of ciphers?

>> When we encrypt with multiple different ciphers, the resulting
>> encryption needs to analyzed as if it was one (not a cascade of
>> ciphers) cipher.
> I disagree: We do not insist on "analyzing" a
> cipher with each exact key that may be used.
> Yet a bad key can be a real issue: With some
> conventional block ciphers, some keys could
> possibly create the identity permutation, or
> some other simple permutation, or a substantial
> part thereof. Very few ciphers are proven to
> produce no such result.

We are not talking about analyzing each key; we're talking about analyzing
the cipher that results from the concatenation of different ciphers.

> Sometimes known weak keys are identified and
> eliminated, but then that becomes a bias to
> the randomness of the keys. Other times, the
> probability of weak keys is known to be very
> small and is then ignored. More often, the
> probability of key weakness is simply unknown.
> And then we still do not "analyze" the cipher
> under each new key.

This is just more irrelevant talk about keys.

> A general system in which keying selected a stack
> arbitrary ciphers on a message-by-message basis
> could easily test adjacent cipher keys to assure
> they differ, if that was thought worthwhile.

I am not talking about analyzing individual messages or individual keys. I
am talking about analyzing the cipher ABC that is the result of
concatenating the individual ciphers A, B, and C.

So far, you have presented no good reason to avoid thoroughly analyzing the
concatenated cipher. If you are going to propose that such a cipher
combination be used, the onus is upon you and upon peer review to analyze
the beast.


Joe Peschel 
D.O.E. SysWorks                        

Relevant Pages