Colossus the code maker

Colossus was the first programmable computer, and was built to crack the Lorenz cipher.

What if Colossus, rather than being destroyed, had been used again as a code making computer?
 
Last edited:
Who says it wasn't?
Seriously, Bletchley Park being closed down didn't mark the end of British codebreaking - the whole thing moved to Cheltenham and became GCHQ. The reason for the destruction was that ex-German Enigma machines and ones working on the same principle were being used everywhere after the war, and the British government was desperate to ensure that nobody realised they could read it. Concealing that they can read codes like that isn't much use if they lose the ability to read those codes.

As for using Colossus to set codes, it isn't at all clear that it is any better than the mechanical alternatives - and realistically doing so would force you to have valve-driven computers at both ends of the coded message, which at the time would probably have stretched technology too much.
 

Deleted member 94680

As for using Colossus to set codes, it isn't at all clear that it is any better than the mechanical alternatives - and realistically doing so would force you to have valve-driven computers at both ends of the coded message, which at the time would probably have stretched technology too much.

Use it for a special set of coding for Main Base to Main Base communications then - Embassy to Embassy for instance?
 
If it has enough computing power to crack the Lorenz cipher, then surely it has enough to encode and decode a more powerful cipher than the Lorenz.
 
Use it for a special set of coding for Main Base to Main Base communications then - Embassy to Embassy for instance?
That leads to a proliferation of the technology, however, and also begs the question why a mechanical coding machine isn't good enough. The problems with enigma were largely in how it was used, not the design of the machine (with the exception of being unable to encipher a letter as itself).

If it has enough computing power to crack the Lorenz cipher, then surely it has enough to encode and decode a more powerful cipher than the Lorenz.
Ummm... no. Doesn't work like that.
 
Point is that the encoding process takes trivial amounts of computing power - and throwing more computing power at only really helps you find longer and more obscure keys, rather than apply more complex codes.
 
I'm not sure what "trivial amounts of computing power" is supposed to mean. Code breaking takes more power than encoding. Yes, Colossus could have found longer and more obscure keys, and the fact is that the Germans could not have cracked them because they had no computers at that time.
 
I'm not sure what "trivial amounts of computing power" is supposed to mean. Code breaking takes more power than encoding. Yes, Colossus could have found longer and more obscure keys, and the fact is that the Germans could not have cracked them because they had no computers at that time.
But neither could the Brits. No point in having a code the receiver can't decode, is there?
 

Deleted member 1487

Colossus was the first programmable computer, and was built to crack the Lorenz cipher.

What if Colossus, rather than being destroyed, had been used again as a code making computer?
Well, it was actually the 2nd, 1st Allied one. The Zuse computers were earlier:
https://en.wikipedia.org/wiki/Konrad_Zuse#The_Z2.2C_Z3.2C_and_Z4

I'm not sure what "trivial amounts of computing power" is supposed to mean. Code breaking takes more power than encoding. Yes, Colossus could have found longer and more obscure keys, and the fact is that the Germans could not have cracked them because they had no computers at that time.
That is grossly misinformed, they had more advanced models than the Colossus:
https://en.wikipedia.org/wiki/Z3_(computer)
It was Turing complete too.

Arguably the Z1 version in the late 1930s was doing the same thing as the Colossus, but less efficiently:
https://en.wikipedia.org/wiki/Z1_(computer)

Not sure if the Colossus was capable of code making.
 
But Colossus was certainly the first programmable electronic computer. Even the Z3 was electromechanical. How could a code breaking computer be incapable of code-making?
 
I'm wondering if the Brits already had the next generation machine being installed at Chetlenham? New machines were being built in the US, it seems illogical such were not ordered up for Britain.
 

Deleted member 1487

But Colossus was certainly the first programmable electronic computer. Even the Z3 was electromechanical. How could a code breaking computer be incapable of code-making?
Generating new codes on it's own? I'm not sure it had the ability to generate it on it's own.

As to the Z3:
The Z3 was an electromechanical computer designed by Konrad Zuse. It was the world's first working programmable, fully automatic digital computer.[1]
It was the first programmable automatic digital computer. Not sure if being fully electronic functionally mattered to use. The Colossus was not Turing complete, the Z3 was.
 

marathag

Banned
I'm wondering if the Brits already had the next generation machine being installed at Chetlenham? New machines were being built in the US, it seems illogical such were not ordered up for Britain.

That's about the time IBM introduced their 701 computer, that was used in both commercial and military application.
 

CalBear

Moderator
Donor
Monthly Donor
The critical element (SO wanted to use "key" there) in establishing and maintaining codes, at least before mega-flop computers, was/is operator discipline. Enigma would never have been broken had operators not failed in proper procedure. A close second is regular replacement, followed by volume. The Reich made all THREE errors and did so on an ongoing basis. Using the same header/footer, sending the same text twice, non random settings, all greatly aided the code breakers. The decision to use plain text under Enigma was also a weakness. To give an opposite example: American "Code talkers" spoke Navajo, a language that was dying even among the tribes, making it almost impossible to understand, this was however, only the first layer. Code Talkers also used code words that only someone who had been trained to interpret the word from its code to actual meaning (this was aided since Navajo lacked words for things like machine gun, artillery, mortars, etc. so the USMC communications department had to come up with work around, some obvious, other utterly impenetrable). There were documented case of the Japanese capturing non Code Talkers, putting them on a tap, and ordering them to translate. What came out was such gibberish (sentences like "house hawk stream sun canyon flower") that the poor bastard who had the bayonet to his throat couldn't tell the Japanese even if he wanted to.

The last error, and it was a huge one, was to send almost everything using the same code setting. Weather report, attack order, location of super secret supply dump, all went out using the same code. There were a few messages, usually straight from Hitler, that didn't. That code was not broken (just like the IJN Flag Officer code resisted penetration due to lack of traffic to analyze).
 
Last edited:

Deleted member 94680

The critical element (SO wanted to use "key" there) in establishing and maintaining codes, at least before mega-flop computers, was/is operator discipline. Enigma would never have been broken had operators not failed in proper procedure. A close second is regular replacement, followed by volume. The Reich made all THREE errors and did so on an ongoing basis. Using the same header/footer, sending the same text twice, non random settings, all greatly aided the code breakers. The decision to use plain text under Enigma was also a weakness. To give an opposite example: American "Code talkers" spoke Navajo, a language that was dying even among the tribes, making it almost impossible to understand, this was however, only the first layer. Code Talkers also used code words that only someone who had been trained to interpret the word from its code to actual meaning (this was aided since Navajo lacked words for things like machine gun, artillery, mortars, etc. so the USMC communications department had to come up with work around, some obvious, other utterly impenetrable). There were documented case of the Japanese capturing non Code Talkers, putting them on a tap, and ordering them to translate. What came out was such gibberish (sentences like "house hawk stream sun canyon flower") that the poor bastard who had the bayonet to his throat couldn't tell the Japanese even if he wanted to.

The last error, and it was a huge one, was to send almost everything using the same code setting. Weather report, attack order, location of super secret supply dump, all went out using the same code. There were a few messages, usually straight from Hitler, that didn't. That code was not broken (just like the IJN Flag Officer code resisted penetration due to lack of traffic to analyze).

Spot on. Wasn't there something about sending birthday greeting to/from Hitler to all U-boats at sea at the time that aided cracking of the code?
 
Top