December 2003 - cypherpunks-legacy

minor address correction
by bart＠netcom.com 17 Dec '03

17 Dec '03

Forwarded message: > Date: Tue, 21 Sep 93 21:35:16 -0600 > From: "L. Detweiler" <ld231782(a)longs.lance.colostate.edu> ... > > note: the new sci.crypt FAQ will have a much-improved section on public > key cryptography. watch for it on the newsgroup or > rtfm.mit.edu:/pub/usenet/news-answers/cryptography-faq if you want it right away. > > You'll do better in: /pub/usenet-by-group/news.answers/cryptography-faq" is current directory. ^^^^^^^^^^^^^^^ ^

1 0

Re: Master Key: A Clipper Story
by George A. Gleason 17 Dec '03

17 Dec '03

This is odd,... I never got the original posting; if anyone can email it to me I'd be thrilled (gg(a)well.sf.ca.us) thanks- -gg

1 0

New York Times: Federal Inquiry on Software...
by gtoal＠an-teallach.com 17 Dec '03

17 Dec '03

In article <199309220117.AA11227(a)bongo.cc.utexas.edu> zeek(a)bongo.cc.utexas.edu writes: > ---------------------------- > The New York Times > Tuesday, September 21, 1993 > Business Day > ---------------------------- > > Federal Inquiry on Software Examines Privacy Programs > > By John Markoff John Markoff regularly writes good and informed articles that he has clearly researched (if that's the right word) by reading usenet and possibly lists like this one. Yet I've *never* seen the guy post. John, if you're reading this, drop me a line to say hello just so I have your email address, in case I ever want to mail you something. Graham -- Personal mail to gtoal(a)gtoal.com (I read it in the evenings) Business mail to gtoal(a)an-teallach.com (Be careful with the spelling!) Faxes to An Teallach Limited: +44 31 662 4678 Voice: +44 31 668 1550 x212

1 0

a horrible conspiracy revealed!
by gtoal＠an-teallach.com 17 Dec '03

17 Dec '03

In article <9309220341.AA23923(a)snark.lehman.com> pmetzger(a)lehman.com writes: > I wonder what the significance of L. Detweiler's name is? Obviously > its a Germanic name. Perhaps he's a Nazi... in fact, he might be the I sincerely *hope* this article was a forgery, otherwise Perry you're digging yourself a big credibility hole. Drop it, please. G -- Personal mail to gtoal(a)gtoal.com (I read it in the evenings) Business mail to gtoal(a)an-teallach.com (Be careful with the spelling!) Faxes to An Teallach Limited: +44 31 662 4678 Voice: +44 31 668 1550 x212

2 1

Re: a horrible conspiracy revealed
by Perry E. Metzger 17 Dec '03

17 Dec '03

Timothy Newsham says: > > > > > > "L. Detweiler" says: > > > I was thinking. What is the significance that PRZ was not actually > > > cited in any subpoena so far? This is very puzzling. It seems to > > > > I wonder what the significance of L. Detweiler's name is? Obviously > > its a Germanic name. Perhaps he's a Nazi... in fact, he might be the > [...] > > What was the point of this letter? Is this example output of a > new stegonagrphy program you just wrote? where can I get it? > Do you know mr. subliminal from saturday night live? Can I meet > him? Can I have a porsche ? If I told you what the point was, there would be no point in having written it. Perry

1 0

PKZ and NON-RSA
by David Colston 17 Dec '03

17 Dec '03

When I posted the offer for source code to a non-RSA public key yesterday, little did I know that I would be beseiged by requests. I will therefore send the paper on the matter to cypherpunks and ask that anyone wanting a copy of the source send me a floppy. Any size or density is okay. As a preamble to this paper, I'd like to note why and how it was developed. After the first release of PGP, Phil told me he was worried about using RSA. I offered to come up with another scheme and originally designed a version of Rabin. I asked Phil to post a copy of that paper on sci.crypt since I do not have internet access. There were few responces, none of which attacked the safety of the method, but mainly stated that it was a version of Rabin. Nothing was ever done with that proposal. A year latter, 2.0 PGP was due to be released and I came up with the current public key system. Again, Phil ignored it. Posts on sci.crypt were totally ignored by anyone with the ability to judge the math. Phil was then approached by the customs folks and I recieved a call from him in December of last year. By that time I was tired of waiting around for Phil and had coded QPK (Quick Public Key) in PDS 7.1. At any rate, I suggested that Phil foward a copy of the program to RSA and see if they would challenge that program in court. He declined. It became apparant that Phil wanted to stick with the original RSA concept and would refuse to do anything with another approach, although he did stick an bite in 2.1 to flag an encryption as being non-RSA. At any rate, I believe that RSA would not attempt to defend the original public key patent in court, since it smacks of over kill. They attempt in the RSA pattent to claim any encryption method involving a polynomial modulo another number. This is clearly prior art, since Ceaser used m+1 mod n as an encrytion method. The program I generated has been posted on Compuserve, Bix, and GEnie for over a year with no feed back from anyone, even though I am resonably sure that RSA is aware of its existance. They have not challenged it because there has been no publicity. I would welcome a court challenge on the whole idea of crypto patents. The customs thing is meerly a way for RSA to challenge Phil and scare off others while avoiding a test of their patents. The "paper" follows. FOR THE MATHEMATICALLY ORIENTED - HOW A QUICK PUBLIC KEY WORKS. Math notation: + plus - minus +- plus or minus * multiplication / division ^ exponent <> unequal = equals == congruent < less than > greater than INT truncated integer round SQR square root ( open expression ) close expression x^-1 modulo p the multiplicative inverse of x in the field of p x... y the range of values XOR exclusive or N a number equal to P times Q, where P and Q are prime Variables in capital letters are permanent and those in small letters are temporary. BACKGROUND Because the "secret key" function of many public keys methods are so slow, most cryptographers use these functions only to "boot strap" into a conventional key system, which is faster to send the actual message. Most of the conventional key systems use comparatively small numbers in relation to the size of the public N as a "random seed number". The holder of the secret key may actually have a larger amount of computer time to decipher the starting point of the conventional algorithm than to decipher the actual message. It would seem to be a good idea, if a public key function could take advantage of the actual message size required to speed up the public key process. The range of message sizes is described below, but generally speaking we a discussing messages less than SQR(Q) in actual size. Imagine a series of related equations modulo a prime, P. These equations have the formula ((e * e + e)/2 * L + C) modulo P. The value, C, is a constant determined by the rule (L - 1) * (1 / (2 * L)) modulo P == C. For L = 1, C = 0. Therefore, if L is known, the expression ((e * e + e)/2) modulo P may be determined, even if e is unknown. Each value of L has an area or series of areas, in which the value of e becomes discoverable, WITHOUT resorting to a modular square root. ie. Let r == ((e * e + e)/2 * L + C) modulo P. If e is in the correct range relative to L, then (r * L * 8 + (L - 2) * (L - 2)) will have an integer square root and the value, e, may be determined with ease. The range of values of e, for any value of L, which have this property and the location of those values vary greatly. The following illustrates a public key approach for L = 12, but other values of L may also be used. Perhaps I should also note that the particular L, which a secret key holder uses need not be public knowledge, but is not all that sensitive. ESTABLISHING A QUICK PUBLIC KEY (Q.P.K.) BASED ON L = 12 A person wishing to receive public messages, which he/she alone can decrypt calculates N = P * Q. Where P and Q are a randomly selected prime numbers, Q being the larger. A == (11 * 24^-1 ) modulo Q B == (2 * A) modulo Q D = Q - B If D > (Q - 1)/2 then set D = Q - D - 1. F = (Q - 1)/2 - D NOTE: We may chose to use the F for all of the following calculations instead of D. This applies to no value of L lower than 12 and F is not valid for most sequences higher than 12. F may be an even safer value to use, for reasons which are too long to discuss here. L is NOT super-sensitive information. Let Y1 ... Y2 be a range of numbers with in the limits: (D - k) and (D + k), where k = INT(SQR(2 * Q / 12)). Y1 may be randomly selected from any point in this range, but Y2 may not be larger than (D + k), and Y2 - Y1 is the maximum message size. A message range for N, public information, is then created by using Chinese remainder theorem to find the modular intersection Q == Y1 and P == x, x being a random number in the range x > 0, x < P. This intersection is called S. A check is made to verify the following: A' = (11 * 24^-1) modulo N B' == (2 * A') modulo N D' = N - B' If D' > (N - 1)/2 then Set D' = N - D' - 1 NOTE: P, Q, D, F, Y1, Y2, k, A, AND B are secret values. Q.P.K. ENCRYPTION A public key for short messages consists of S and N. To send a Message the sender calculates: e = (S + Message) ((e * e + e)/2) modulo N == Cipher Q.P.K. DECRYPTION t == Cipher modulo Q f == (t * 12 + A) modulo Q g = SQR(f * 8 * 12 + 100) NOTE: If g is NOT an integer value, the message is rejected as invalid. If g modulo (2 * L) <> (L - 2) then Q is repetitively added to g until g modulo (2 * L) == (L - 2). z = (g - 10)/24 e == ((B - 1) + z) modulo Q If e > (Q - 1)/2 then set e = (Q - 1) - e. Message = e - Y1 For other values of L: A == ((L - 1) * 2^-1* L^-1)) modulo Q k = INT(SQR(2 * Q / L)) NOTE: If L = 1 then D = 0 and the message range is 1... k. If L = 2 then the message range is D... (Q - 1)/2 and these values modulo Q are already perfect squares < Q. f == (t * L + A) modulo Q g = SQR(f * 8 * L + (L - 2) ^ 2) z = (g - (L - 2)/(L * 2)) If anyone wants the source code for this drop a disk to: Colston & Associates 5111 Rogers Ave. Suite 507 Fort Smith, AR 72903

1 0

Re: a horrible conspiracy revealed!
by gtoal＠an-teallach.com 17 Dec '03

17 Dec '03

> I wonder what the significance of L. Detweiler's name is? Obviously > its a Germanic name. Perhaps he's a Nazi... in fact, he might be the Graham> I sincerely *hope* this article was a forgery, Graham> otherwise Perry you're digging yourself a big Graham> credibility hole. Drop it, please. I guess you didn't get it huh? I'm really sick of the flame wars on here, but this one had me laughing on the floor. No, I did get the joke, I just didn't think it was funny. I *do* remember the thread some time ago of Perry being accused of being a Nazi because a well- known Illinois Nazi has the same surname. I meant he was losing credibility for indulging in childish games, not for meaning whatever he said literally. I'd rather the pair of them just let it drop, instead of trying to outsmart each other. G

3 2

Public-Key Crypto Toolkit
by Mike Ingle 17 Dec '03

17 Dec '03

What we need is a toolkit which makes it easy to write public-key applications. This would be an easy interface to the routines in PGP. I was thinking of doing this in Clisp under Linux, since Lisp makes it easy to put together and take apart complex objects, and makes it easy to kick around objects of arbitrary size. Unfortunately Clisp does not seem to support calling C functions! Perhaps it could be hacked right into the Clisp source code, giving it a fast modmult, idea, md5, etc. Any ideas as to what language/method would be best for implementing a PK toolbox? This could really get the "Cypherpunks write code" ideal moving. Anyone want to help? It should not contain any cryptography in itself - you should link it with the crypto from pgp, optionally applying diffs to the source files first, so that there is no fear of distributing crypto or violating patents. --- MikeIngle(a)delphi.com Brought to you by AT&T ClipperPhones - reach out and tap someone!

4 4

Former S.O.F. Editor Supports PGP in Newspaper
by James Still 17 Dec '03

17 Dec '03

Local boy makes good at home. The Federal Grand Jury's PGP/RSA test case, as reported in the NYT has hit full circle here in Boulder. Boulder's own Paul Danish (former editor and writer for Soldier of Fortune magazine, based in Boulder) has come out in support of PGP/privacy issues with an excellent op-ed piece in today's Colorado Daily. The Colorado Daily is read by over 50,000 students, fac/staff and residents of Boulder and the Univ. of Colorado. Danish's piece does an amazing job of backgrounding the issues, presenting them in layman's terms, and he even goes into the Clipper chip proposal. Could it be that the NSA/U.S. Customs investigation is causing a huge backlash of public support in favor of PGP?

1 0

META: Re: a horrible conspiracy revealed!
by jrk＠sys.uea.ac.uk 17 Dec '03

17 Dec '03

I was about to do an ::exclude on this thread, when I realised that it was the cypherpunks list and not the extropians. Any chance of cypherpunks running the extropian list software? Or maybe it does already? -- ____ Richard Kennaway __\_ / School of Information Systems Internet: jrk(a)sys.uea.ac.uk \ X/ University of East Anglia uucp: ...mcsun!ukc!uea-sys!jrk \/ Norwich NR4 7TJ, U.K

3 2