Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

That photo of Colossus

tnmoc.org

Post reply on HN
Print view View original

31–37 of 37 posts

Re: That photo of Colossus

#31

Post by josst »

mathieuh wrote: post #25

I'm not sure of the prior history, which may affect my appreciation for it, but as far as I know, what happened with Colossus was that they were aware the cipher seeds changed each day, and they obtained an Enigma machine so they could decipher Enigma codes, it was the fact there was a reliance on intercepting the day’s seeds, or having a spy/double agent leak them which meant cracking Enigma messages was usefully im…

Small correction, Colossus was designed to break the Lorenz cypher (which was considerably more complex than Enigma). The machine itself was designed by Tommy Flowers (not Turing), who was a post office worker who had previously worked with vacuum tubes and thought it would be possible to build a programmable computer with them. After the war Flowers applied for funding to build a similar machine (Colossus being top secret) and was denied on account of the idea being impractical.

The machine Turing is famous for his work on was Bombe, which was designed to crack Enigma.

Re: That photo of Colossus

#32

Post by skolsuper »

mathieuh wrote: post #29
Waterluvian wrote: post #27

Earlier quoted context omitted.

Wasn't it that they could feed in certain known possible letters to cut down on the number of options? Something like, "this encrypted segment is usually the weather so the s is probably the word 'sunny'".

I'm shamefully unsure of how it works exactly, but it's possible they had subroutines that looked for phrases like "weather", "movement", "troops" etc. to try and speed up the brute force. https://www.youtube.com/watch?v=9HH-asvLAj4 Here's a really good video I've watched before and since forgotten the details of which explains Colossus and its context

Some enigma operators had flourishes that the code-breakers could pick up on. Starting or ending a lot of messages with "Heil Hitler" was also a factor IIRC

Re: That photo of Colossus

#33

Post by someguydave »

skolsuper wrote: post #32
mathieuh wrote: post #29

Earlier quoted context omitted.

I'm shamefully unsure of how it works exactly, but it's possible they had subroutines that looked for phrases like "weather", "movement", "troops" etc. to try and speed up the brute force. https://www.youtube.com/watch?v=9HH-asvLAj4 Here's a really good video I've watched before and since forgotten the details of which explains Colossus and its context

Some enigma operators had flourishes that the code-breakers could pick up on. Starting or ending a lot of messages with "Heil Hitler" was also a factor IIRC

Yes this is correct. The resulting ciphertext from Enigma was sent by manual Morse transmission. Morse operators can identify others by thier “hand” which is the Morse equivalent of the sound of your voice in speech.

Earlier in the war the Germans typically sent their call signs in plaintext. Many of the intercept operators in the “Y stations” in the UK became familiar with the “hands” of various German operators. Later in the war the Germans started encrypting their call signs, but Bletchley was generally able to distinguish them by the intercept operator’s familiarity with the “hands” of the German Morse operators.

Re: That photo of Colossus

#34

Post by mathieuh »

josst wrote: post #31
mathieuh wrote: post #25

I'm not sure of the prior history, which may affect my appreciation for it, but as far as I know, what happened with Colossus was that they were aware the cipher seeds changed each day, and they obtained an Enigma machine so they could decipher Enigma codes, it was the fact there was a reliance on intercepting the day’s seeds, or having a spy/double agent leak them which meant cracking Enigma messages was usefully im…

Small correction, Colossus was designed to break the Lorenz cypher (which was considerably more complex than Enigma). The machine itself was designed by Tommy Flowers (not Turing), who was a post office worker who had previously worked with vacuum tubes and thought it would be possible to build a programmable computer with them. After the war Flowers applied for funding to build a similar machine (Colossus being top…

Can't edit my comment now that it has children (I think that's how the logic works) so upvoting you for visibility.

Re: That photo of Colossus

#35

Post by Animats »

lb1lf wrote: post #30
mirimir wrote: post #23

Earlier quoted context omitted.

Ah. So they could have many copies of those tapes, and quickly replace ones that broke, right?

I seem to recall reading in either Kahn's or Sebag-Montefiore's book on Bletchley Park that one early failure mode of the tape was that it was run through the reader at a fast enough pace to occasionally catch fire. Oops.

Basic problem of early computing: no good memory devices. IBM had electronic arithmetic before WWII, as an experimental tube multiplier. But nobody had a memory device that didn't require lots of parts per digit or bit. Hence cards and paper tape.

Re: That photo of Colossus

#36

Post by mirimir »

Animats wrote: post #35
lb1lf wrote: post #30

Earlier quoted context omitted.

I seem to recall reading in either Kahn's or Sebag-Montefiore's book on Bletchley Park that one early failure mode of the tape was that it was run through the reader at a fast enough pace to occasionally catch fire. Oops.

Basic problem of early computing: no good memory devices. IBM had electronic arithmetic before WWII, as an experimental tube multiplier. But nobody had a memory device that didn't require lots of parts per digit or bit. Hence cards and paper tape.

I had an old friend who worked with 60s era mainframes. You had a decent enough CPU, but no memory. And very slow storage.

So you worked out some set of instructions with total execution time comparable to storage I/O. And put that into ~firmware. Then you read data from one device, and wrote data to another device. And kept repeating that, with other instruction sets, until you were done.

That's not very different from current usage. Except that most intermediate results go into various sorts of memory. Or /tmp devices, if there's too much data for memory. And at the CPU level, instruction sets are just about as small.

But the point is that one could process ~huge amounts of data. Many GiB, as I recall. Just very slowly, with lots of thrashing tape.

And now one can use the same approach, except with clusters. And process humongous amounts of data.

Post reply on HN