. Well, not really. As a couple of examples, Banburismus and many of the techniques used against the German teletype ciphers were novel applications of maths to cryptanalysis developed at Bletchley.
A Polish patriotic myth, I'm afraid. While you're correct to say that Enigma was broken by the Polish, it was *also* broken by the British.
Enigma wasn't an egg: one broken, always broken. On the contrary, you had to find the keys every time they changed, which was at least daily. Worse, the Germans upgraded Enigma with a series of security features between 1932 and 1945 that severely complicated cryptanalysis. By mid-1940, the Polish techniques were obsolete, and so the British invented new ones.
All that's just a long way of saying that there's plenty of glory to go around.
A question is, is this a problem? Are those "obvious" steps in proofs ever wrong? Given that, as you say, they are "well beyond most talented undergraduates", aren't there lots of slip-ups?
Some people's attitude is "just change the code, no need to think too hard about correctness, your tests will catch it if you get it wrong." That's a bit foolhardy. As you point out, there's every chance the tests won't catch all the problems you might introduce. You always have to use your brain. Still, unit tests can catch mistakes and give you evidence that you haven't changed the semantics.
There's a whole school of thought that advocates refactoring as one of the primary activities in writing software. The purpose of unit tests is then to give you confidence that you haven't changed the semantics.
The thing about unit testing is that it's subject to the law of diminishing returns. A simple test of the basic functionality gets you a lot for minimal effort. Writing dozens of carefully chosen tests to examine boundary conditions etc. gives you a little bit more, but for a great deal more effort. Whether or not it's worth it depends very much on the situation and the nature of the code you're writing.
"They present it as an encyclopedia," Brandt said Friday.
Well, yes. Not that odd, really, given that it is an encyclopedia.
"They go around claiming it's almost as good as Britannica."
Actually, Wikipedians don't, in my experience. Most are quite sober when it comes to comparisons with Britannica. Brandt may be referring to the journal Nature, which did make such a claim for science articles.
They are trying to be mainstream respectable. Wikipedia is already pretty darn mainstream, and if by "respectable" Brandt means "free of plagiarised material", then he's correct.
I'm all for getting information out to the greatest amount of people, I just don't think that rebuilding a section of Wikipedia in a separate location (and from scratch) is a particularly good way of doing that.
I agree, neither Wikipedia's crypto stuff or CryptoDox are particularly newsworthy at this point. Or ever.
I'd just like to point out that when you said, "with (apparently) fewer articles than even the Wikipedia crypto collection", that pretty much every comparable work has fewer articles than the Wikipedia crypto collection; Wikipedia has over a thousand articles on the subject. Now, I'll grant you that the majority are pretty poor, but that's twice as much any other crypto encyclopedia (not that there's a huge choice!)
CryptoDox seems to be doing the same thing: creating an encyclopedia about cryptography using MediaWiki. To be honest, I don't really understand why this guy wants to do it outside of Wikipedia -- I've asked him, but he's never given any reason for it . Still, he's quite welcome to do what he likes, of course, and since he's now using the GFDL (he was using Creative Commons Non-Commercial last week), we can copy material back and forth between Wikipedia and CryptoDox. So, if you're into crypto and fancy helping out, feel free get involved with either project -- it's a win for both.
"After World War II, some fifty bombes were retained at Eastcote, while the rest were destroyed. The surviving bombes were put to work, possibly on Eastern bloc ciphers (Smith, 1998). The official history of the bombe states that "some of these machines were to be stored away but others were required to run new jobs and sixteen machines were kept comparatively busy on menus. It is interesting to note that most of the jobs came up and the operating, checking and other times maintained were faster than the best times during the war periods."
And:
"Two Colossus computers, along with two replica Tunny machines, were retained, moving to GCHQ's new headquarters at Eastcote in April 1946, and moving again with GCHQ to Cheltenham between 1952 and 1954[3]. One of the Colossi, known as Colossus Blue, was dismantled in 1959; the other in 1960[3]. In their later years, the Colossi were used for training, but before that, there had been attempts to adapt them, with varying success, to other purposes[4]. Jack Good relates how he was the first to use it after the war, persuading the NSA that Colossus could be used to perform a function for which they were planning to build a special purpose machine[3]. Colossus was also used to perform character counts on one-time pad tape to ensure their randomness[3]."
> umm NO, despite British Government propaganda, it was POLES who cracked enigma, who captured first Enigmas, and who > deciphered Enigma codes by HAND on a daily basis until Germany invaded us:(
Please avoid SHOUTING, particularly when don't know what you're talking about.
The Poles, like the British and Americans, did indeed crack Enigma messages. However, I don't believe the Poles ever performed a naval or other military capture of an Enigma, nor did they ever get hold of the standard "Wehrmacht" Enigma. (Apparently, they managed to have a peek at a commercial Enigma over a weekend at some point before 1930, and they built their own Enigma copies, of course, after Rejewski solved the rotor wiring in December 1932.)
Far from using only "HAND" methods, the Poles used machinery (bomby and cyclometer) and deciphering aids (perforated sheets) aplenty in their work.
> Turing did NOT crack Enigma! > England _exploited_ Polish work and took all the fame and glory for it:(. Turing merely implemented Rejewski algorythms.
There's two errors you can make when discussing history of Enigma codebreaking. The first is to fail to mention the early Polish work. This happened a lot in the 1970s and 1980s, and was chiefly a result of ignorance, rather than a deliberate attempt to swipe the glory. Happily this oversight occurs less often now.
The second error, which is becoming more common, is to say what you're saying: that the British simply copied Polish work. It's presumably an overreaction to the first error. The truth is that there's plenty of credit to go around, as the British and Polish *both* cracked Enigma using remarkably innovative and clever -- yet for the most part different -- techniques.
After May 1940, the Germans changed their procedures and the Polish techniques no longer worked. The British had to come up with their own ideas, the most significant was the Turing-Welchman Bombe (completely different to the Polish bomba).
You underestimate the complication added by the plugboard; there's no simplistic way it can be ignored. Remeber the path of the current: it passes first through the plugboard, forward and back through the rotor stack, then out via the plugboard again. Your attack would only work if the current didn't pass through the plugboard a second time.
Lots of errors in your post: All of the methods used in World War II were short-cut algorithms. For example, bombes would search systematically over all starting positions and orderings of the rotors, but not over the billions of plugboard settings. Colossus wasn't used for Enigma. It would seem Alan Turing never used Colossus. Colossus was not faster than a P4 (unless you program in Javascript, as Tony Sale does). Sometimes operators did not randomize their rotors between messages, but this would not lead to a message in depth (overlayed messages).
The IWM website is mistaken. Unfortunately, it's a very common mix-up. Colossus was a very specialised piece of hardware, and was not used on Enigma traffic. You can easily find confirmation if you care to dig into the various references hanging off of the Wikipedia Colossus computer article:
They've already factored one with 200 digits last May (RSA-200).
Not the largest RSA number factored to date
on
RSA-640 Factored
·
· Score: 2, Interesting
A nice result! Interestingly, the same team factored RSA-200 last May, which is 663 bits long (confusingly, there's two series of RSA challenge numbers with different naming conventions) but for which no prize was given for its solution. RSA-640 is shorter, at 640 bits, but carries a US$20,000 prize. It's not entirely clear why the team went for the larger, prizeless number first.
Maybe there's other factors at work here besides prize money? (ROFL etc).
Slashdot Mirror
Bletchley Park is well worth a visit for some codebreaking + early computing. ~45 minute train journey from Euston.
http://www.bletchleypark.org.uk/
Is this the "perfectly accurate" which is also hideously biased, or the "perfectly accurate" which just happens to be pure original research?
The mathematics was done in Poland
. Well, not really. As a couple of examples, Banburismus and many of the techniques used against the German teletype ciphers were novel applications of maths to cryptanalysis developed at Bletchley.
A Polish patriotic myth, I'm afraid. While you're correct to say that Enigma was broken by the Polish, it was *also* broken by the British.
Enigma wasn't an egg: one broken, always broken. On the contrary, you had to find the keys every time they changed, which was at least daily. Worse, the Germans upgraded Enigma with a series of security features between 1932 and 1945 that severely complicated cryptanalysis. By mid-1940, the Polish techniques were obsolete, and so the British invented new ones.
All that's just a long way of saying that there's plenty of glory to go around.
I'm about as far from being a gay-rights activist as it's possible to be
You're Fred Phelps?
A question is, is this a problem? Are those "obvious" steps in proofs ever wrong? Given that, as you say, they are "well beyond most talented undergraduates", aren't there lots of slip-ups?
Some people's attitude is "just change the code, no need to think too hard about correctness, your tests will catch it if you get it wrong." That's a bit foolhardy. As you point out, there's every chance the tests won't catch all the problems you might introduce. You always have to use your brain. Still, unit tests can catch mistakes and give you evidence that you haven't changed the semantics.
There's a whole school of thought that advocates refactoring as one of the primary activities in writing software. The purpose of unit tests is then to give you confidence that you haven't changed the semantics.
The thing about unit testing is that it's subject to the law of diminishing returns. A simple test of the basic functionality gets you a lot for minimal effort. Writing dozens of carefully chosen tests to examine boundary conditions etc. gives you a little bit more, but for a great deal more effort. Whether or not it's worth it depends very much on the situation and the nature of the code you're writing.
"They present it as an encyclopedia," Brandt said Friday.
Well, yes. Not that odd, really, given that it is an encyclopedia.
"They go around claiming it's almost as good as Britannica."
Actually, Wikipedians don't, in my experience. Most are quite sober when it comes to comparisons with Britannica. Brandt may be referring to the journal Nature, which did make such a claim for science articles.
They are trying to be mainstream respectable.
Wikipedia is already pretty darn mainstream, and if by "respectable" Brandt means "free of plagiarised material", then he's correct.
"you'd probably look for a website that was specifically about disseminating info about cryptography."
h y
Sure, but most people would start with a Google search:
http://www.google.co.uk/search?hl=en&q=cryptograp
First result is a Wikipedia article.
I'm all for getting information out to the greatest amount of people, I just don't think that rebuilding a section of Wikipedia in a separate location (and from scratch) is a particularly good way of doing that.
I agree, neither Wikipedia's crypto stuff or CryptoDox are particularly newsworthy at this point. Or ever.
I'd just like to point out that when you said, "with (apparently) fewer articles than even the Wikipedia crypto collection", that pretty much every comparable work has fewer articles than the Wikipedia crypto collection; Wikipedia has over a thousand articles on the subject. Now, I'll grant you that the majority are pretty poor, but that's twice as much any other crypto encyclopedia (not that there's a huge choice!)
Disclaimer: I started the English Wikipedia's Cryptography Project page:
t _Cryptography
http://en.wikipedia.org/wiki/Wikipedia:WikiProjec
CryptoDox seems to be doing the same thing: creating an encyclopedia about cryptography using MediaWiki. To be honest, I don't really understand why this guy wants to do it outside of Wikipedia -- I've asked him, but he's never given any reason for it . Still, he's quite welcome to do what he likes, of course, and since he's now using the GFDL (he was using Creative Commons Non-Commercial last week), we can copy material back and forth between Wikipedia and CryptoDox. So, if you're into crypto and fancy helping out, feel free get involved with either project -- it's a win for both.
Yeah. Wikipedia says:
"After World War II, some fifty bombes were retained at Eastcote, while the rest were destroyed. The surviving bombes were put to work, possibly on Eastern bloc ciphers (Smith, 1998). The official history of the bombe states that "some of these machines were to be stored away but others were required to run new jobs and sixteen machines were kept comparatively busy on menus. It is interesting to note that most of the jobs came up and the operating, checking and other times maintained were faster than the best times during the war periods."
And:
"Two Colossus computers, along with two replica Tunny machines, were retained, moving to GCHQ's new headquarters at Eastcote in April 1946, and moving again with GCHQ to Cheltenham between 1952 and 1954[3]. One of the Colossi, known as Colossus Blue, was dismantled in 1959; the other in 1960[3]. In their later years, the Colossi were used for training, but before that, there had been attempts to adapt them, with varying success, to other purposes[4]. Jack Good relates how he was the first to use it after the war, persuading the NSA that Colossus could be used to perform a function for which they were planning to build a special purpose machine[3]. Colossus was also used to perform character counts on one-time pad tape to ensure their randomness[3]."
Only if Turing helped create Colossus, but he didn't, or at least, he did only very indirectly.
> umm NO, despite British Government propaganda, it was POLES who cracked enigma, who captured first Enigmas, and who :(
> deciphered Enigma codes by HAND on a daily basis until Germany invaded us
Please avoid SHOUTING, particularly when don't know what you're talking about.
The Poles, like the British and Americans, did indeed crack Enigma messages. However, I don't believe the Poles ever performed a naval or other military capture of an Enigma, nor did they ever get hold of the standard "Wehrmacht" Enigma. (Apparently, they managed to have a peek at a commercial Enigma over a weekend at some point before 1930, and they built their own Enigma copies, of course, after Rejewski solved the rotor wiring in December 1932.)
Far from using only "HAND" methods, the Poles used machinery (bomby and cyclometer) and deciphering aids (perforated sheets) aplenty in their work.
> Is that the same Lorenz that Einstein mentioned in his theory of relativity? (Lorenz transforms.)
No, that was Hendrik Lorentz, a Dutch physicist. The company which made the Lorenz teleprinter cipher machine was founded by a Carl Lorenz in Berlin.
> Turing did NOT crack Enigma! :(. Turing merely implemented Rejewski algorythms.
> England _exploited_ Polish work and took all the fame and glory for it
There's two errors you can make when discussing history of Enigma codebreaking. The first is to fail to mention the early Polish work. This happened a lot in the 1970s and 1980s, and was chiefly a result of ignorance, rather than a deliberate attempt to swipe the glory. Happily this oversight occurs less often now.
The second error, which is becoming more common, is to say what you're saying: that the British simply copied Polish work. It's presumably an overreaction to the first error. The truth is that there's plenty of credit to go around, as the British and Polish *both* cracked Enigma using remarkably innovative and clever -- yet for the most part different -- techniques.
After May 1940, the Germans changed their procedures and the Polish techniques no longer worked. The British had to come up with their own ideas, the most significant was the Turing-Welchman Bombe (completely different to the Polish bomba).
You underestimate the complication added by the plugboard; there's no simplistic way it can be ignored. Remeber the path of the current: it passes first through the plugboard, forward and back through the rotor stack, then out via the plugboard again. Your attack would only work if the current didn't pass through the plugboard a second time.
This could well be because Tony Sale's simulators are written in JavaScript...
> The Enigma code was broken ...
Lots of errors in your post: All of the methods used in World War II were short-cut algorithms. For example, bombes would search systematically over all starting positions and orderings of the rotors, but not over the billions of plugboard settings. Colossus wasn't used for Enigma. It would seem Alan Turing never used Colossus. Colossus was not faster than a P4 (unless you program in Javascript, as Tony Sale does). Sometimes operators did not randomize their rotors between messages, but this would not lead to a message in depth (overlayed messages).
Actually, even AES-192 is considered sufficient for TOP SECRET by the NSA:
? MenuID=10.2.7
"CNSSP-15 correctly states that 192-bit AES keys are sufficient for protecting even TOPSECRET information. However, Suite B uses only 256-bit keys to enhance interoperability." -- http://www.nsa.gov/ia/industry/crypto_suite_b.cfm
The IWM website is mistaken. Unfortunately, it's a very common mix-up. Colossus was a very specialised piece of hardware, and was not used on Enigma traffic. You can easily find confirmation if you care to dig into the various references hanging off of the Wikipedia Colossus computer article:
t /t17/TR17-004.html
http://en.wikipedia.org/wiki/Colossus_computer
I'm afraid the Heath Robinson was indeed called the "Heath Robinson". Here's an actual contemporary document:
http://www.alanturing.net/turing_archive/archive/
There was also Super Robinson and Old Robinson.
They've already factored one with 200 digits last May (RSA-200).
A nice result! Interestingly, the same team factored RSA-200 last May, which is 663 bits long (confusingly, there's two series of RSA challenge numbers with different naming conventions) but for which no prize was given for its solution. RSA-640 is shorter, at 640 bits, but carries a US$20,000 prize. It's not entirely clear why the team went for the larger, prizeless number first.
Maybe there's other factors at work here besides prize money? (ROFL etc).