A few weeks ago my wife and I visited Bletchley Park, home of Britain'sfamous World War II code breakers. I thought the pictures might beinteresting to computer folk.
Allan Turing and many otherothers, both famous and not, worked at BletchleyPark to crack a number of codes used by the Axis powers. Themost famous of these was the German Enigma. Their naval Enigma codes were the mostvaluableof all due to the amount of shipping being lost to the U-boats. Thenumbers were staggering, and the allies read intercepted messages, whenthey could, to route convoys around the wolf packs.
A couple of weeks before going to England Marybeth and I stumbledinto the Museum of Science and Industry in Chicago. The museum has U-505 on display, which wascapturedand provided hints used in the always on-going process of crackingEnigma's latest encryption settings.
It's well worth a visit, but do read “Enigma” by Hugh Sebag-Montefiore beforegoingto the museum for a far different take on the military politics of thetaking of U-505 than you'll get in Chicago.
The museum is almost worshipful in its admiration of CaptainGallery, who bravelyeffected U-505's capture. But according to the book, Admiral Kingconsideredcourt-martialing Gallery since his failure to sink the U boat couldhave lead to the German's getting wind of Bletchley's successes.
(Mr. Sebag-Montefiore's familyactually owned Bletchley Park just before the government acquired itfor their code-breaking purposes .)
The Enigma machines used three, and later four, “rotors” plus aplugboard to encipher plaintext. To read these messages the allies hadto figure out rotor settings. In the 30s a small team of Poles, usinginformation leaked by a German to French intelligence officers, workedout the Enigma machine's design and some of the critical algorithmsneeded to crack the codes. The Polish team gets scant respect in moderncoverage of the Enigma saga, but the Park has a monument dedicated tothem.
|Figure1: Monument to the Polish cryptographers.|
At Bletchley, Turing and others built the so-called “Bombe” machine to help work outEnigma rotor positions. Hundreds were built and deployed around Britainand the US. Few survived the war (though rumors persist several wereused through the 50s on Soviet codes before being destroyed). One hasbeen reconstructed and works. Here are a few pictures:
|Figure2: Front view of the reconstructed Bombe|
|Figure3: Rear Bombe view.|
|Figure4: Another Rear Bombe view|
Why the name “Bombe?” There are several theories; the one thatappeals most to me is that the Polish geniuses who initially crackedEnigma invented the basic ideas behind the machine while eating an icecream called “Bomba,” which the British Anglicized to “Bombe.”
For Americans, at least, Bletchley Park is just about impossible tofind. We counted ten roundabouts between the M40 and the Park – get onewrong and you're spun off into hyperspace into some seeminglypoorly-mapped little village. A couple of years ago I tried to visitbut got hopelessly lost. When we finally made it there I told theticket person that one had to be a cryptographer just to find theplace. So it's not surprising that tour groups, at least during theweek, are tiny.
Despite the small group, imagine our surprise – and delight – tofind that one member of our tour was Joyce Winston, née Meyer.Joyce had worked in the Park during the war as a Bombe operator! Shewas one of the 6000 or so WRENS assignedto the project, and now, 60+ years later, escorted by her son anddaughter-in-law, visited a scene of her youth. Her memory had some biggaps but after the tour she talked to me excitedly about her work andliving quarters nearby.
|Figure5: Joyce Winston, a WWII Bombe operator, posing in front of a Bombe.|
The Bombe is not electronic; it's electromechanical. That's not todemean the machine at all, but my biggest interest in visiting the Parkwas to make a pilgrimage to Colossus . The British claim it is the first electronic digital computer. I nolonger make any claims about historical firsts as there's always someexpert who unveils some heretofore barely-recognized first. Maybe theGerman Konrad Zuse had someearlier technology (for computer buffs a visit to the Deutsches Museum,which shows off some fascinating Zuse machines, is a must-see).
Conventional history posits ENIACas the first electronic digital computer. Note that “programmable”isn't an included adjective, since the ENIAC originally wasn'tparticularly programmable, though later modifications changed that.Colossus, though, was a heavily-guarded secret whose existence wasn'trevealed till the 70s, and which wasn't acknowledged till 2000 by theBritish government. So ENIAC priority claims were made while the truthremained classified.
Colossus was not built to break Enigma. It was meant to helpdecipher the German High Command's Lorenz code, which was much harderto crack than Enigma. Lorenz encryption machines had 12 encryptingrotors compared to Enigma's 3 or 4. Since each rotor gave 26 possiblescrambles, the permutations defy imagination.
No Colossus survives. Yet Tony Sale managed to build one from scratch based on incomplete andfragmentaryinformation he garnered from a meager partial schematic, plusinterviews with some of the Colossus engineers. The reconstructedmachine actually works, and was running while we were there. It'ssimply breathtaking.
|Figure6: Colossus – the tape reader.|
|Figure7: Colossus – front view|
|Figure8: Colossus ” Document Output?|
|Figure9: Colossus ” User Interface?|
The Colossus prototype was so successful that practically before itran follow-on versions were ordered. Version 2 had some 2400 “valves,”or vacuum tubes. Most were single-pentodes. Now we can't design atelephone with less than a couple of million active elements, sobuilding a machine with 2400 transistor-equivalents was quite the feat.
I hung around after the tour wandered off and chatted up theColossus docent, an ex-Honeywell engineer. He showed me hand-drawnsketches of Colossus circuits for basic logic elements, which usedextremely clever suppressor-grid arrangements to get more work out of asimple circuit. I wish I'd taken pictures of the schematics. Tommy Flowers designed themachine,and he must have been an engineering genius. It's a shame his creationwent unrecognized for so long.
The rear of the machine is off-limits yet houses the mostinteresting electronics. But the docent was responsive to anengineering discussion and let me behind the ropes. Here are a fewpictures:
|Figure10: Colossus – the rear of the machine. These tubes are mostlythyrotrons.|
|Figure11: Colossus rear side – wider view|
|Figure12: Colossus – rear side close up|
|Figure13: Colossus ” another rear side view|
At least one of the rebuilt Colossus's vacuum tubes (uh, “valves”)is 64 years old and still going strong. When Colossus was firstproposed skeptics felt it would be so unreliable, due to tube failures,that the machine wouldn't be practical.
Flowers realized most tube failures occurred at power-on, so leftthe machine on 24/7. The cost of electricity prohibits that today, sothe replica uses an automated variable transformer to slowly, over thecourse of minutes, bring the filaments up to rated voltage and therebyavoid thermal shock.
|Figure14: The arrow points to a 64 year old tube. See the glowing filament bythe cap with the red wire?|
Finally, I was struck by the “Engineer's Lament,” postedinconspicuously in a display case:
|Figure15: The ballard of Bletchley Park|
Jack G. Ganssle is a lecturer and consultant on embeddeddevelopment issues. He conducts seminars on embedded systems and helpscompanies with their embedded challenges. Contact him at . His website is .