View Single Post
  #234  
Old July 2nd 04, 07:12 AM
Geoffrey Sinclair
external usenet poster
 
Posts: n/a
Default

Denyav wrote in message ...
Of the 50,000 code groups used only 3800 had been cracked


Nope,by December almost 1000 were cracked.


Wrong, the December 1941 figure for JN-25B code groups
assigned meanings was 3,800.

ope the Japanese imtroduced a new 50,000 group additive
book in August 1941 which set them back months


Question was about April not August.


The reality is Denyav is unable to cope with the fact the number
in April 1941 was 300 JN-25B code groups assigned meanings.

They also had low priority, Joe Rochefort recorded that his group had
little
equipment, the tabulating machines and teletypes needed had been
diverted to the efforts to break the German and Italian codes and
the Japanese diplomatic cipher


After Pearl Harbor suddenly everything changed and US (and British)code
breakers broke all codes that they were unable to crack in 12 months prior to
Pearl Harbor within a few months .
I wonder how this miracle happened,maybe your famous outer space aliens ,this
time masquerading as US (and British) code breakers made it happen.


Translation the massive increases in resources thrown at the problem,
Hawaii had all those unemployed sailors from the sunken ships for
example, add the greater use of radio by the IJN giving more text, then
the fact that the code breakers started from knowing 3,800 meanings
in December 1941, not zero in December 1940, (and as everyone
knows it is easier to figure out a missing word in a sentence than a
missing sentence with only one word). The fact the code had uppercase
(2 meanings per code group) and auxiliary tables was known, the fact
the "next x groups uppercase" markers were well known and used as
ways to crack the messages (the difference table) in December 1941,
versus not being known in December 1940 and so on.

JN-25 was the sort of code that the more you knew the quicker you
could learn the missing parts.

Yes folks, Denyav tries to announce that the code was compromised
in 2 months in December 1940 and January 1941, but now has to
turn around and say the allies, with more resources, starting from a
much higher level of knowledge, could not compromise the code in
6 months. You see for Denyav to be right the allies went backwards
in 1942 when it came to reading JN-25. Simple really, just invent facts
and delete the problems.

By the way one of the signals used as "proof" the allies could read
JN-25 when required is the following, British dd/mm/yy format.

"XXXXX NAVAL MESSAGE T.O.O.2339A/12 IN Received: 12.2.42

Addressed: B.A.D. Washington 234
From: Admiralty
Japanese aircraft carrier SHOKAKU intends
to leave Yokosuka and pick up aircraft 9th February

It is intended preparations of First and
Second aircraft Squadrons were to complete 7th
February, sail and embark aircraft 8th. E.T.A.
Palao 1200/12th

All above from special intelligence. 2339A/12

This is now mid February 1942, with two extra
months of work, more resources, with more messages to
work on, with forward feedback in action (the fewer missing
words the easier it is to figure out the missing words, with
more of the random number encypherments known the
easier to strip them out). Note it took until the 12th to decrypt
the message which talks about the 7th in future tense, in other
words about a week to crack a simple message.

This has been used to claim real time decodings of much
larger messages in 1941.

So in mid February 1942 the allies could crack relatively
small, standard, messages after about a week.

He didnt even have a secure comms line to the radio intercept
station. All intercepts had to be delivered by courier. It wasnt
until Febuary 1942 that the group began to read even a small
portion of messages in JN-25


Even according to Safford 7000 codes were enough to read the most
JN25B coded messages and many messages could be read with 1500
codes.


Safford never wrote the work being quoted, and the example
appears to be about a special case in 1943.

Lets remember US had recovered almost 10000 codes by Dec.4


No, the 10,000 claim is straight fiction. The real number was 3,800


Geoffrey Sinclair
Remove the nb for email.