Germans efforts to break allied codes.

[ljadw]

How long would it take for the Germans to decode an Allied message (Afaics,it took BP 2 days to decode an Enigma message)? If it took the Germans also 2 days,what would be the usefulness of such message ?

[paspartoo]

How long would it take for the Germans to decode an Allied message (Afaics,it took BP 2 days to decode an Enigma message)? If it took the Germans also 2 days,what would be the usefulness of such message ?

Naval Cypher No3 (convoy cipher) was read in 1943 often '10-20 hours in advance'
Regarding the Enigma it depends on the 'key' and the time period. Rommel's 'key' was initially broken after a week.From June'42 after 24h on average.

For the U-boat key (by the way the stats are for the general u-boat key,not the officers):


Source: ibibllio

[Larry D.]

If they are used correctly it doesn't matter if you have cray supercomputers or quantum computers or whole planets covered with machinery. It is not solvable.

With a statement like that, you would never be hired by NSA. They do not believe in the word "impossible". The word is a function of time, time is continuous, therefore "impossible" is impossible. That's what they believe. Given enough time and resources, any man-made system can be broken, even if it is the product of random generation.

[Larry D.]

If they are used correctly it doesn't matter if you have cray supercomputers or quantum computers or whole planets covered with machinery. It is not solvable.

With a statement like that, you would never be hired by NSA. They do not believe in the word "impossible". The word is a function of time, time is continuous, therefore "impossible" is impossible. That's what they believe. Given enough time and resources, any man-made system can be broken, even if it is the product of random generation.

[paspartoo]

If they are used correctly it doesn't matter if you have cray supercomputers or quantum computers or whole planets covered with machinery. It is not solvable.

With a statement like that, you would never be hired by NSA. They do not believe in the word "impossible". The word is a function of time, time is continuous, therefore "impossible" is impossible. That's what they believe. Given enough time and resources, any man-made system can be broken, even if it is the product of random generation.

Really? You mean my application will be turned down?
Seriously though just google one time pad ,it will explain why it is unbreakable.

[Carl Schwamberger]

Several items in the Wiki article that run directly against my own experience. Also the text about Shannons proofs are clear in that the "hypothetical" unbreakability of a one time code depends on a hypothetical near perfection in the pattern of the code segment it self.

I also have to wonder if the writer of the Wiki article was depending on only open source information or was knowledgeable about actual practices within any agencies of organizations which kept their methods 'secret'.

[Carl Schwamberger]

How long would it take for the Germans to decode an Allied message (Afaics,it took BP 2 days to decode an Enigma message)? If it took the Germans also 2 days,what would be the usefulness of such message ?

'Enigma. Breaking the Code' by Hugh Seabag-Montifiore and 'Battle of Wits' by Stephen Budinsky have some descriptions of the time required. From memory it varied from 'to long' in 1940-41 to very fast in 1944-45. The introduction of the large British "Bombes" were a order of magnitude more capable than the early Polish and Brit machines, later the IBM machines with far more and higher speed rotors carried the process another order higher, The speeds allowed a larger quantity of material to be decrypted as well, as did mechanical decryption vs the tedious manual copying from a replica of a German machine.

The German organization sending the message made a difference. Those using the older three rotor machines were easier than those using four or five rotors. Some organizations had less well trained operators and sloppier practice which made them more vulnerable. By late 1942 the Brits had a fairly well established system for identifying which messages might be more vulnerable and other cribs which made the use of the machines more efficient. In 1941 they were still very slow or hit and miss in finding cribs & other aids.

Perhaps later I can dig through the books and find some specfic examples of time required. Both books are published in the past two decades and reasonablly easy to locate.

[paspartoo]

Several items in the Wiki article that run directly against my own experience. Also the text about Shannons proofs are clear in that the "hypothetical" unbreakability of a one time code depends on a hypothetical near perfection in the pattern of the code segment it self.

I also have to wonder if the writer of the Wiki article was depending on only open source information or was knowledgeable about actual practices within any agencies of organizations which kept their methods 'secret'.

Ehm the fact that otp is unbreakable is as i said before mathematically proven. In fact for a long time all major powers used teleprinters running otp tape as their secure means of communications.
Even back in WWII the Germans had the T43 otp teleprinter , the Americans the SIGTOT and the Brits the Rockex.
Of course in order for otp to be true unbreakable it means that the stream of numbers comprising the additive must be random. So using a computer program to create a stream of otp tables could be broken since it would not create a truly random sequence.
This is pretty elementary stuff...

From memory it varied from 'to long' in 1940-41 to very fast in 1944-45.

It depended on the 'key' . Keep in mind each Enigma key had three setting : General , Officers , Staff. When you read in books that Bletchley Park read a code 99% it means the general .Officers was much harder. Staff more or less unbreakable.

[paspartoo]

A correction. Rommel's 'keys' where broken in second half '42 on average half of the time and of these half within 48 hours,the rest within a week.
Statistics provided by Ralph Erskine:

[Urmel]

How long would it take for the Germans to decode an Allied message (Afaics,it took BP 2 days to decode an Enigma message)? If it took the Germans also 2 days,what would be the usefulness of such message ?

That may have been an average, but it could be much quicker than that.

[Uncle Fritz]

I heard that in 1945 Germans had a center in Niederschlesien, that was able to read messages to Moscow from soviet spies operating in USA, and that they learned a lot about Manhattan Project that way. I believe Wołoszański wrote an interesting book on this subject.

[paspartoo]

I heard that in 1945 Germans had a center in Niederschlesien, that was able to read messages to Moscow from soviet spies operating in USA, and that they learned a lot about Manhattan Project that way. I believe Wołoszański wrote an interesting book on this subject.

Interesting story. Do you remember more details? What was the organisation that decoded these messages ? What sources did this author use ?

[Carl Schwamberger]

How long would it take for the Germans to decode an Allied message (Afaics,it took BP 2 days to decode an Enigma message)? If it took the Germans also 2 days,what would be the usefulness of such message ?

The content of the message may or may not have perished by or after two days. If it has become dated the content is usually still useful for establishing patterns & identifying/eliminating long term possibilities. identifying past patterns for better conjecture of future intent is a critical part of intel analysis.

[Urmel]

How long would it take for the Germans to decode an Allied message (Afaics,it took BP 2 days to decode an Enigma message)? If it took the Germans also 2 days,what would be the usefulness of such message ?

That may have been an average, but it could be much quicker than that.

Example, message of 'late' 19 November 1941 was decoded and circulated by 0915 GMT 20 November.

[Carl Schwamberger]

I heard that in 1945 Germans had a center in Niederschlesien, that was able to read messages to Moscow from soviet spies operating in USA, and that they learned a lot about Manhattan Project that way. I believe Wołoszański wrote an interesting book on this subject.

Reading some on the US "Verona " operation I saw the claim the Soviet communications from the US used one time pads exclusively. That renews my interest in Woloszanskis book. What was the title?

Thanks