Home » Tuck Talk » Chapter discussion of Tuck Season... » Episode ten plus bonus
Episode ten plus bonus [message #6415] Mon, 05 July 2010 09:34 Go to next message
Ellen Hayes  is currently offline Ellen Hayes
Messages: 684
Registered: September 2002
Senior Member
Episode ten is up this morning. Also included is a bonus piece involving the coding scheme Mike and Tuck are using this summer. Rather math-heavy, but something Tuck can do with pencil and paper; and likely to keep even the newspaper-cryptanalysts at bay.

Not entirely dead yet!


Ellen
nosig
Re: Episode ten plus bonus [message #6416] Mon, 05 July 2010 10:22 Go to previous messageGo to next message
mkemp  is currently offline mkemp
Messages: 421
Registered: April 2006
Senior Member
Thanks, Ellen.
Re: Episode ten plus bonus [message #6417] Mon, 05 July 2010 16:07 Go to previous messageGo to next message
mkemp  is currently offline mkemp
Messages: 421
Registered: April 2006
Senior Member
*What could I do to change her?*
I almost snorted coffee on my keyboard at that.

At this point it seems as if Tuck's messages should be giving Jane a faint glimmer of a Clue of who they're dealing with.

I hope Tuck realizes that the trip to Jane's place is a scouting mission.
Re: Episode ten plus bonus [message #6418] Tue, 06 July 2010 00:15 Go to previous messageGo to next message
Erin Halfelven  is currently offline Erin Halfelven
Messages: 712
Registered: September 2002
Location: Surf City, USA
Senior Member
Administrator
Cryptanalysis is something I've actually done for pay, back in my spying days. Smile

Anything not a simple cipher pretty much needs a professional to break and a professional with a computer to break in a timely manner. Tuck/Mike's is a good cipher-rotor comparable to the ones used in WWII. English and Americans broke almost every German and Japanese cipher and cipher-rotor used during the war, even one's more complex than this.

That's why the Allies used codes instead of ciphers for the important stuff. Smile Or even better, codes enciphered with one time pads.

The big easy way to break codes and ciphers before powerful computers: taking advantage of operator error.

Also, the story of ciphers and codes in WWII is the story of why a more free society has advantages over a more tyrannical one. In Germany, cryptographers were told to come up with an unbreakable cipher. They were mathematicians, they knew it could not be done with the current technology but they came up with one they hoped would be good enough and told their tyrannical bosses that yes, it was unbreakable.

Among the Allies, no one much feared what would happen if they told their bosses that something was impossible so their bosses ended up being well enough informed to use codes and one time ciphers for really important information. Look up "one time pad" in Wikipedia.

In Germany, people knew that no mechanical (predictable) cipher could be completely secure but how could you tell a Nazi boss that? The situation in Japan was just as bad if not worse. End effect, the Allies read much of the Axis communications all through the war and gained real advantages from it.

Not always in time: raids on Pearl Harbor, USA and Coventry, England were described in ciphered messages that were not correctly interpreted and communicated in time to prevent the destruction. But those happened early in the war.
Re: Episode ten plus bonus [message #6419] Wed, 07 July 2010 06:30 Go to previous messageGo to next message
stanman  is currently offline stanman
Messages: 292
Registered: May 2008
Senior Member
thanks
Re: Episode ten plus bonus [message #6420] Thu, 08 July 2010 10:48 Go to previous messageGo to next message
stanman  is currently offline stanman
Messages: 292
Registered: May 2008
Senior Member
I wonder if Ellen will allow for her new Tucky Seasons story to be posted at Big Closet and at other sites?
Re: Episode ten plus bonus [message #6421] Thu, 08 July 2010 17:23 Go to previous messageGo to next message
Brooke  is currently offline Brooke
Messages: 695
Registered: August 2004
Location: Portland, OR
Senior Member
Erin Halfelven wrote on Mon, 05 July 2010 21:15

Cryptanalysis is something I've actually done for pay, back in my spying days. Smile

Anything not a simple cipher pretty much needs a professional to break and a professional with a computer to break in a timely manner. Tuck/Mike's is a good cipher-rotor comparable to the ones used in WWII. English and Americans broke almost every German and Japanese cipher and cipher-rotor used during the war, even one's more complex than this.

That's why the Allies used codes instead of ciphers for the important stuff. Smile Or even better, codes enciphered with one time pads.

The big easy way to break codes and ciphers before powerful computers: taking advantage of operator error.


Back in the mid-80s some of us were doing a multi-author "spy" sytory on a rather odd BBS (it was essentially an online *line* editor, so there were onl;y seperate messages to the extent that you put in dividers.

Someone posted a message in a simple Caesar cipher. I cracked it and replied oin something harder. Then we started using progressively more complex substitution ciphers (details were arranged at weekly get togethers). And one guy kept cracking them.

Turned out he was using a few really simple cryptananlysis programs from a book (Cryptanalysis with Microcomputers by Caxton). The code was originally for a Commodore PET. He'd translated it (in spite of *major* typos by the publisher in transcribing the listings) to run on his Apple II.

It'd give you key lengths and a few other things in seconds.

We finally got a cipher he couldn't crack by shifting to bitwise encryption (technically it wasn't *really* encryption. Just convert ASCII into a stream of bits and then split that into 5bit blocks which got turned into printable characters. But letter frequencies went right out the window as a tool)
Re: Episode ten plus bonus [message #6422] Wed, 14 July 2010 10:45 Go to previous messageGo to next message
Ellen Hayes  is currently offline Ellen Hayes
Messages: 684
Registered: September 2002
Senior Member
You're presenting a somewhat overoptimistic and lopsided picture of Axis vs. Allies cryptography here.

By late 1941, a tremendous effort by William Friedman and others had cracked the highest level DIPLOMATIC code of the Japanese; but its message traffic did not include military planning. Pearl Harbor was not mentioned in the diplomatic military traffic; nor were any other specific targets.

Much British merchant ship traffic was found and sunk due to an early German capture of the BAMS code (which IIRC stands for British Allied Merchant Shipping); much fun was had by all as British codebreakers would pick up an Enigma order going to submarines, and radio a warning in BAMS to the endangered vessels; which would be picked up and decoded by the German codebreakers, and a revised order sent to the submarines via Enigma, which would be intercepted and decoded by British codebreakers... I'm hearing Yakety Sax...

Seemingly many if not most of Rommel's successes in Africa can be traced to the diligent efforts of one man, Bonner Fellers, who was, unfortunately, a US military attache. The British told him everything, which he then analyzed, coded, and had radioed to Washington - but at the same time, the Germans had a superb field radio interception unit that was armed with a good crack of BLACK (the code Fellers was using), and were getting the British strengths, plans, concerns, etc., at least as fast as Washington was. Much easier to defend and attack successfully if you know what your opponent is going to do at the same time he knows it. The German radio-intercept unit was overrun on 10 July 1942, and the capability was lost; Rommel never had such successes again.

One of "the miracles of Midway" was that the Imperial Japanese Navy was supposed to change its main high-level Navy code (called by the US 'JN25') on 1 April 1942, but the far-flung nature of the Navy delayed the changeover until June. Most of the American codebreaking was done in these last two months, including the oft-repeated story of confirming that AF meant Midway.

Very few military forces - none that I can find - used one-time pads; they present tremendous difficulties in keeping synchronized, especially when dealing with more than one recipient. Also, they use up key letters at the rate of one key letter per letter of message; this would present a large logistical burden to keep up with the megabytes of traffic issued and received by one division in one day - at a time when most of the messages had to be encrypted and decrypted by humans using pencils.

A cipher concept invented by freakin' Thomas Jefferson (3rd President of the US) was used often for quick local encryption as the M-94 and M-138-A 'machines'.

Later, a joint Allied project called VENONA searched through Russian secret-agent radio traffic that had been picked up during the war years, because the Soviet Union had reused some of its one-time pads in different areas. Very computer heavy, especially during a time when punch cards were the data entry format, but it opened a surprising number of messages. VENONA was finally stopped (or so it's claimed) in 1980, over thirty years after the messages were sent.

For a tolerably comprehensive overview, I'd suggest picking up (carefully, it's heavy) "The Codebreakers" by David Kahn. While it doesn't have much of the last 50 years (the digital revolution), it covers the two thousand years before 1960 pretty well.


Ellen
who did the research
Re: Episode ten plus bonus [message #6423] Wed, 14 July 2010 10:49 Go to previous messageGo to next message
Ellen Hayes  is currently offline Ellen Hayes
Messages: 684
Registered: September 2002
Senior Member
stanman wrote on Thu, 08 July 2010 15:48

I wonder if Ellen will allow for her new Tucky Seasons story to be posted at Big Closet and at other sites?


Only if Ellen doesn't have to go to the hassle of actually uploading and writing summaries for each one. Also, I'm no longer allowing removal of the PGP 'envelope'; so any site that has a desperate need to pretty-print Tuck will either have to suck it up and live with non-pretty-print, or will be in violation of my licensing and will be persecuted until they comply.


Ellen
nosig
Re: Episode ten plus bonus [message #6424] Wed, 14 July 2010 10:56 Go to previous messageGo to next message
Ellen Hayes  is currently offline Ellen Hayes
Messages: 684
Registered: September 2002
Senior Member
Brooke wrote on Thu, 08 July 2010 22:23

Turned out he was using a few really simple cryptananlysis programs from a book (Cryptanalysis with Microcomputers by Caxton).


*snip snip* This sounds right up Bill's alley... hand the book to Tuck, start encrypting messages, and paying lil' Eugene per crack. And possibly per program, or per new algorithm.


Ellen
nosig
Re: Episode ten plus bonus [message #6427] Wed, 14 July 2010 21:39 Go to previous messageGo to next message
Brooke  is currently offline Brooke
Messages: 695
Registered: August 2004
Location: Portland, OR
Senior Member
Ellen Hayes wrote on Wed, 14 July 2010 07:56

Brooke wrote on Thu, 08 July 2010 22:23

Turned out he was using a few really simple cryptananlysis programs from a book (Cryptanalysis with Microcomputers by Caxton).


*snip snip* This sounds right up Bill's alley... hand the book to Tuck, start encrypting messages, and paying lil' Eugene per crack. And possibly per program, or per new algorithm.



Some day I hope to unearth my copy of the book (buried in storage for more years than I want to think about) and get all the algorithms converted to at least more modern versions of BASIC.

The ones I'd converted back in the 80s are one of the reasons I'l holding onto my old Model 100 and the disk interface for it. I need to get the data & programs off the disks and onto more modern media.

Hmm. Found several links for the book, including this one:
https://www.amazon.com/Cryptanalysis-Microcomputers-Caxton-Foster/dp/081 0451743

In any case, cryptanalysis seems like a toss-up as to whether Mike or Tuck would be more into it.

But the algorithms in the book aren't all that great by current standards. Even by 1997 standards. Wasn't PGP already out then?
Re: Episode ten plus bonus [message #6428] Thu, 15 July 2010 09:50 Go to previous messageGo to next message
Ellen Hayes  is currently offline Ellen Hayes
Messages: 684
Registered: September 2002
Senior Member
Brooke wrote on Thu, 15 July 2010 02:39

But the algorithms in the book aren't all that great by current standards. Even by 1997 standards. Wasn't PGP already out then?


PGP was definitely out; note that all Tuck episodes are signed with it, and you can check the dates.

However, if anyone ever bothered to do what Mike and Tucker are doing in this episode - which I'll admit is unlikely; those that want security would use programs like PGP, and The Rest Of Them wouldn't know encryption if it ate their genitals - then such programs would be useful.

I suggest if you are interested in the code-cracking stuff, that you definitely pick up David Kahn's "The Codebreakers" and check out his descriptions on how to solve the various methods. Don't even bother with trying to convert some godawful hardware-specific ancient BASIC programs to something modern; start over. The book, I think, gives more general algorithms that could "easily" be translated into code.

Hey, maybe you could come up with a Debian library - libcryptanalysis or something. =-)


Ellen
nosig

Re: Episode ten plus bonus [message #6429] Sun, 18 July 2010 14:27 Go to previous message
Brooke  is currently offline Brooke
Messages: 695
Registered: August 2004
Location: Portland, OR
Senior Member
I'd had a paperback copy of The Codebreakers for years before I got pointed at that other book. And I've had the full hardcover for about 10 years or so.

The Caxyon book is worth converting vecause the only "hardeware specific" bits where the characters used for a few math functions (having a copy of a book on how to interconvert Appe Copmmodore and TRS-80 programs was useful but not necessary).

The algorithms for breaking codes are useful. The ones for encryption are nice examples, but not all that new. Though I'd never encountered the "rail fence" cipher before.

Previous Topic:Number Nine... Number Nine...
Next Topic:Surprise!
Goto Forum:
  


Current Time: Tue Dec 18 15:59:26 EST 2018

Total time taken to generate the page: 0.02074 seconds
.:: Contact :: Home ::.

Powered by: FUDforum 2.7.7.
Copyright ©2001-2007 FUD Forum Bulletin Board Software