Today at 07:24 GMT Nuller[NDS] from Russia Team returned machine settings which decoded the following plaintext:

dievonkolonneabteilungyfenf
neundreikommandiertendiolawtkr
aetwagenfuenutonnxtslcbygvhayc
taqtakdayywlypo

Check out the forum thread to read more details.

---

M4 Project homepage
M4 Project wiki

The message starting with KEJNQ was taken from this page:

http://cryptocellar.web.cern.ch/cryptocellar/bgac/GArmy_messages.html

First 5 letters were stripped and the message was split into two parts: letters 1-102 (workunits kejnq102_1 and kejnq102_3) and 51-153 (workunits kejnq102_2 and kejnq102_4).

kejnq102_1 and _2 were assigned to standard '1941' dictionary, _3 and _4 were assigned to Stefan Krah's naval dictionary.

The broken part was kejnq102_3 so it looks like Stefan's dictionary was better for this message.

The message needs cleaning up - the last part is garbled and of course there are still missing letters 103-153, but that shouldn't be too hard.

---

M4 Project homepage
M4 Project wiki

Machine settings for this message:

UKW: B
Wheel order: 253
Rings: MC
Start position: IWX
Stecker: AD BH FG IJ KN LZ MR OS PW QV

I have disabled all the KEJNQ workunits, however I did not remove them from the server queue, as a result the validator would grand 0 credits for the remaining workunits and not all hosts are connected 24/7.
Feel free to abort any remaining KEJNQ tasks in the client queue, however continuing the work on kejnq102_4 might help to clean the message (the server will accept all results, just won't generate new WU's).

Screenshot from the server's top scores log:



and the lucky task: http://www.enigmaathome.net/result.php?resultid=14585224

---

M4 Project homepage
M4 Project wiki

6.5 hours later at 14:08 GMT SKB@P - BOINC@Poland's 'supercomputer' returned another result with different rings and start position, this time the ciphertext (first 102 letters of message) is fully decoded with minor garbles:

UKW: B
Wheel order: 253
Rings: LE
Start position: IVZ
Stecker: AD BH FG IJ KN LZ MR OS PW QV
dievonkolonneabteilungfuenfneundreikommandiertenzwolawtkraetwagenfuenutonnensindsofortzumpakzgruppennj

---

M4 Project homepage
M4 Project wiki

Cool! :) thanks for the update TJM!

I have disabled all the KEJNQ workunits, however I did not remove them from the server queue, as a result the validator would grand 0 credits for the remaining workunits and not all hosts are connected 24/7.
Feel free to abort any remaining KEJNQ tasks in the client queue,...

How many was in the queue? Seems like I've canceled over 200 now but more keep showing up.

The server will keep resending them until they are removed from the queue, so -unfortunately- you'll have to bear with it for a while.

I'll block the resends too, but I need some time to find out how to leave the small part of workunits that's still needed. Unfortunately they all share the same name.

---

M4 Project homepage
M4 Project wiki

This is the closest I could get to clear decrypt of full message (153 letters); machine settings were manually adjusted because the left wheel turn near the end of the message garbled the remaining part.

UKW: B
Wheel order: 253
Stecker: AD BH FG IJ KN LZ MR OS PW QV
Rings: AJE
Message key: ITZ

DIEVONKOLONNEABTEILUNGFUENFNEUNDREIKOMMANDIERTENZWOLAWTKRAETWAGENFUENUTONNENSINDSOFORTZUMPAKZGRUPPENNAQSQUBFUEHRERVIERNAQXPLYUSSAXPLJUSSAXINMARSQZUSETZEN

DIEVO NKOLO NNEAB TEILU NGFUE NFNEU NDREI KOMMA NDIER TENZW OLAWT KRAET WAGEN FUENU TONNE NSIND SOFOR TZUMP AKZGR UPPEN NAQSQ UBFUE HRERV IERNA QXPLY USSAX PLJUS SAXIN MARSQ ZUSET ZEN

---

M4 Project homepage
M4 Project wiki

So if I paste this into Google Translate and guess at where the spaces go, I should be able to see the message?

---

Cool! I think this is the 1st message I've seen broken since I've joined. I was wondering if this contraption does what it says it would do, now we all know....excellent!

Mike D

---

I hope you folks don't mind an outsider butting in; anyway, congratulations on your latest break.

Correcting garbled letters, replacing the Q's with "CH", and adding spaces, I get:

DIE VON KOLONNE ABTEILUNG FUENF NEUN DREI KOMMANDIERTEN ZWO LASTKRAFTWAGEN FUENF TONNEN SIND SOFORT ZUM PANZ GRUPPEN NACHSCHUBFUEHRER VIER NACH X PLYUSSA X PLYUSSA X IN MARSCH ZUSETZEN

Which I translate into English as:

THE TWO 5-TON TRUCKS COMMANDED BY COLUMN DETACHMENT 593 ARE TO BE IMMEDIATELY DISPATCHED ON THE MARCH TOWARDS PLYUSSA TO THE SUPPLY-OFFICER OF PANZER GROUP 4.

It seems that Plyussa is a river in the vicinity of Leningrad (St. Petersburg.)


Dan Girard

That indeed seems to be the correct decryption. Thank you for the insight.

Wikipedia ( http://en.wikipedia.org/wiki/Plyussa_River ) confirms that Plyussa is a russian river located in the extreme west part of Russia, that joins the Narva River at the border of Estonia and Russia, passing through the twin cities of Narva/Ivangorod and ending in the Finnish Gulf.

Awesome work!

Someone asked me about the sum of CPU time used to break this message, I'll post the answer here.

The second attempt (message split to two parts and assigned to two sets of dictionaries) would take 450 days (10800 hours) on Core 2 Duo E6400 running both cores 24/7 at 1.86GHz; the combined power of attached hosts did all this in just few days.

---

M4 Project homepage
M4 Project wiki

Great !!
Now let's break another one !!!

Yeah!! Gotta break 'em all :)

---

Wow! just signed up on this site and see that you guys just broke a code! This is pretty amazing. Ive always had an interest in WWII crypto and here it is just falling into my CPU! Congratulations.

Interesting. I'm new here, longtime interest in the tech of WWII.

So given the huge number of cycles required by this approach, how did the folks at Bletchley do it with tubes and relays in a meaningful fashion?

Seems like the enigma should be crackable in seconds on a modern PC.

FWIW: I did an implementation of an Enigma machine in ASM on an Atmel AVR a couple years ago, just for kicks.