[Cryptography] FW: RFC 7253 on The OCB Authenticated-Encryption Algorithm

[Dave Horsfall]

On Mon, 2 Jun 2014, Richard Outerbridge wrote:

> No, no, no, no. The 1st Polish crack into early Enigma was due to 
> operator and procedural error, pretty much. But I think what you?re 
> thinking of is the crack of FISH, which was an offset by one of a 
> retransmission of a baudot stream. Could be wrong: I often am.

Hmmm...  Bruce mentions the Polish, so I'm not sure how I misremembered my 
version.  Trouble is, there are *many* books on Enigma...  The version I 
heard was indeed that a cipher clerk (presumably under stress - my guess 
only) forgot to change the rotor settings; pick your new schedule and send 
it under the old settings, then flip them for that message (think of it as 
a primitive CBC, in a way).  He forgot to flip the rotors to the new 
settings; as I said, this is what I recall reading somewhere (Google?).

As for the OP, I hadn't heard that it was Japanese-held Midway that got 
bombed, merely that it was Japanese.  Dunno where I read both the above.

Hmmm...  Could've been Simon Singh where I saw Enigma (he has a couple of 
chapters on it, and I misremembered as usual).  He doesn't mention FISH, 
though (but Bruce does).  Singh mentions Midway, but the Allies had 
already cracked it, and decoded a message telling the Japanese to retake 
Midway, so by then it was already broken.  I dunno...

We'll never know the whole story, of course.

-- Dave