- cypherpunks-legacy - lists.cpunks.org

Prime Numbers
by rjc＠gnu.ai.mit.edu 17 Dec '03

17 Dec '03

Eric Hughes: > It was first claimed that if (2^n-2)/n was an integer, then n was > prime. That's false. I thought he said "if p prime, then p|(2^p-2)" which is why I stated the converse isn't true. > then: > > This is fermat's little theorem. What you have written basically > >says 2^N - 2 = 0 (mod N) or 2^(N-1) = 1 (mod N). Note, the converse > >doesn't apply. If (2^N-2)/N is an integer, N isn't neccessarily > >prime. For example, take N=561=(3*11*37) > > 561 is the first Carmichael number. If you replace 2 by any other > number relatively prime to 561, then the congruence still holds. (The > second Carmichael number is 1729, if I remember right.) It was Which is why I chose it. Carmichael numbers are pseudoprime in any valid base so when coming up with a counterexample to the converse of fermat's little theorem, just memorize a few Carmichael numbers. The key property of them is if n is a Carmichael number and n=p*q*r, then (p-1), (q-1), and (r-1) divide (n-1). I wonder if Carmichael numbers always have some small factors. If true, PGP's sieve test probably eliminates the very very rare case that you actually choose one. -Ray -- Ray Cromwell | Engineering is the implementation of science; -- -- rjc(a)gnu.ai.mit.edu | politics is the implementation of faith. --

1 0

Cyber PsyOps
by Jeff Davis 17 Dec '03

17 Dec '03

Having read Black Unicorn, Thank you. Would you please critique my _Paradigms_of_Revolution? I am prone to action on instinct. I'm still thinking about your questions relative to function. For one thing, I suggest that everyone have a refferance with the Constitution with in arms reach of their terminals. Its despicalbly under read by law enforcement types, and may give us an edge up somewhere down the road. -- PGP PUBLIC KEY via finger! JAFEFFM Speaking & Thinking For Myself! * eagle(a)deeptht.armory.com email info(a)eff.org * *** O U T L A W S On The E L E C T R O N I C F R O N T I E R **** ***** Committed to Free Public Internet Access for World Peace *****

1 0

Cyber PsyOps & Media
by Jeff Davis 17 Dec '03

17 Dec '03

I just got off the phone with David Wilson with the Chronicle of Higher Education. Sarah Simpson confirmed my membership in EFF to him, and said I'd better talk to him myself. I told him the document was posted to Usenet, I just accelerated the news of its leak. I told him it was my statement that a wake up call was needed to the American public. They need to take the initiative themselves to attain internet access by their own means. I told him in my opinion it was authentic and I told him why. I assured him that if the document contained classified information I would not have blasted it all over the globe. I told him I had no idea of the origin of the leak. He asked for a copy of my essay, _Paradigms_of_ Revolution_ and I sent it to him. That is a complete report. If it isn't in the above paragraph, I didn't tell him. Are we all clear? Thought so. -- PGP PUBLIC KEY via finger! JAFEFFM Speaking & Thinking For Myself! * eagle(a)deeptht.armory.com email info(a)eff.org * *** O U T L A W S On The E L E C T R O N I C F R O N T I E R **** ***** Committed to Free Public Internet Access for World Peace *****

2 1

Re: Crypto, satellites, & China
by koontzd＠lrcs.loral.com 17 Dec '03

17 Dec '03

>From: Lyle_Seaman(a)transarc.com > >paul(a)poboy.b17c.ingr.com (Paul Robichaux) writes: >> I never knew that encryption fell under MTCR in addition to ITAR. >> Presumably any crypto hardware or software which can be used to >> protect missile or satellite telemetry or commanding can be >> restricted. > >It's my understanding that, according to international missile >treaties (probably the MTCR), missile telemetry data must be >transmitted in the clear. This is ostensibly so that the various >world powers can be assured that a "test" missile launch really is a >test and is not going to wipe out someone's capital. It's also so >that various intelligence agencies can have accurate estimates of >everyone's capabilities. This knowledge is widely viewed as >stabilising. The space shuttle has provision for receive only secure voice for the same reason. On military missions they carry up codelists such as one would use over insecure medium. One wonders if they install a locking door on one of the lockers for securing classified.

1 0

RE: Cyber PsyOps (Re: Shot)
by Blanc Weber 17 Dec '03

17 Dec '03

From: Jeff Davis "Psychological Warfare is duofold. It harrases the enemy, and hopefully wins the hearts and minds of the population." And then all you have to do is maintain your control over the situation. Once you have persuaded everyone to come over to your psychological point of view and have won their sympathies, all you have to do is leave them to their own imagination and they will, like "herds", self-regulate and organize into a new paradigm of confluence according to the tribal provisions written into the Constitution by Thomas Jefferson, thus giving weight to progressive theory. Then for sure you would have a society based on consensual normalities, the horde for whom you would have prepared the way: aware, informed, and actively participating in your aim. Definitely, this would prove that Marbury v. Madison (1803) was a *not* a fuck up. The Libertarian platform would not reverse that decision, because the Constitution has provided amending provisions, and we can always rectify the errors of shortsightedness. I mean, the FOIA may be useless, but we *could* always sit here and read the damn thing. Psychological Warfare is truly the only way to compete with the NSA on this issue. Blanc

1 0

Crypto, satellites, & China
by paul＠poboy.b17c.ingr.com 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- A story in the 3/14/94 issue of _Aviation Week_ reports that Hughes and Martin Marietta have satisfied the State Department and will now be able to launch the Optus B3 comsat. All the quotes below are from the article; it didn't have any detail on the cryptostuff. State had held up the launch because the Optus B3 payload included a "sensitive decryption chip". The USG made trade with China subject to the Missile Technology Control Regime (MTCR) after China was suspected of selling missiles to Pakistan. Hughes agreed to remove the chip, which "guards the satellite's steering commands"; once removed, State lifted the export restriction. "By removing the chip, authority for approving the launch was shifted to the Commerce Dept. from an interagency review commission headed by the State Dept. that had far less leeway to act." I never knew that encryption fell under MTCR in addition to ITAR. Presumably any crypto hardware or software which can be used to protect missile or satellite telemetry or commanding can be restricted. If anyone has more details on the MTCR and its application, I'd love to hear it. - -Paul Robichaux - -- Paul Robichaux, KD4JZG | Out the 10Base-T port, through the router, perobich(a)ingr.com | over the leased line, off the bridge, past Intergraph Federal Systems | the firewall... nothing but net. Of course I don't speak for Intergraph. -----BEGIN PGP SIGNATURE----- Version: 2.3a iQCVAgUBLamroCA78To+806NAQEK0wP/SD8692ZaMl71c57dSKAG7vmYVcyvIJi5 CNuWb4u3kL7WlfEtzrZ42Dfx4avgyzcjyoIyBDE5r+7n/nSMnbBGZGYR2OyTBQ5b GaK8RrrdJQoKujEvzR8i+XhtBin2doK5hzBHTY2VpYIexMH34MLZeYNST4nA7ek9 5EGWhbZ5mxw= =avDx -----END PGP SIGNATURE-----

2 2

encryption evaluater software
by Sheedy, E.C.C. 17 Dec '03

17 Dec '03

>July 10, 1997 > >---------- > > >Nikkei English News via Individual Inc. : (Nikkei Industrial Daily, July 8, >1997) > >Mitsubishi Electric Corp. said it has developed a software program which >can be used to evaluate the security of various symmetric-key encryption >algorithms, where the sender and recipient share the same key. > >The program analyzes a given encryption technology and displays the amount >of computing power required to deduce the key and decrypt the message. > >Secure and practical data encryption is a necessary technology for the >popularization of electronic money and electronic commerce. Mitsubishi's >program could end up becoming an index by which to objectively evaluate the >many new symmetric-key encryption methods being developed. > >The software combines the differential decryption method developed by >Israeli encryption expert Adi Shamir with Mitsubishi's proprietary linear >decryption method. By attempting to deduce the key with simple >approximations of the encryption algorithm, the program determines the >minimal volume of computations needed to crack the code, and also shows how >long the key should be in order to make decryption nearly impossible. >Mitsubishi said it has also finished development of a software program >which uncovers the key based on the pattern of random numbers generated in >the course of encrypting a message. >The next step for the company is to combine all three decryption methods >into a comprehensive program for evaluation. Working with support from the >Ministry of International Trade and Industry, the company hopes to finish >and publicize its program by next spring, providing it free of charge to >users. (More) ________________________________ kpn research E.C.C. Sheedy n m m m m m n________________________ Security & Card systems P.O. Box 15000 Telephone +31 50 582 17 92 9700 CD Groningen Telefax +31 50 312 24 15 The Netherlands e-mail e.c.c.sheedy(a)research.kpn.com ________________________________________________________________

1 0

number theory
by rperkins-remailer＠nyx.cs.du.edu 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- > I should point out that the standard argument that picking 'k' > different values for 'a' and then calculating the probability as > (1/2)^k is fallacious. This would be true if the probabilities were > independent, but they aren't. There was a paper on this about five > years ago whose awareness has not been yet widespread. I no longer > have the reference. Okay, my memory has been jogged... is this a paper by Pomerance, "On the distribution of pseudoprimes"? He gave more precise estimates for the number of base-2 pseudoprimes. With his more precise estimates, the chance of a 100 digit number passing the base-2 pseudoprime test is about 1/10^13... I think his work applies only to base-2 pseudoprimes, so my statement concerning the error rate of Miller-Rabin is still correct: for s iterations, the chance of failure is 2^-s. -----BEGIN PGP SIGNATURE----- Version: 2.3a iQCUAgUBLar8xIOA7OpLWtYzAQEAmgP2NQx7a3woaZMgT5CeqOFrhqyRcYt3mAPd 9bnf+f19E4Il42e0xw9vQjOMyowB/IkATQf+//ADIFxhE9p+2MOpD8eDr9saGYOV bVwV2/bWtzsHqjsbWRH27/5lEwFXerGfJNSc1ITkZFwp1QwpzmVvn6gkOZ2lf0AJ /q3QneS7iw== =2XH+ -----END PGP SIGNATURE-----

1 0

Re: alias in phone book
by dmandl＠lehman.com 17 Dec '03

17 Dec '03

> From: hughes(a)ah.com (Eric Hughes) > > >A professor of communications has brought up the fact that it is illegal > >to use a pseudo in the white pages of the phone book. Is this the case? > > I would suggest first, to ask this professor to make a legal citation, > and if one is not forthcoming, to ask for a retraction of the claim. > > Eric It's definitely false, at least here in NYC. A well-known alternative to having an unlisted phone number (a privilege that the phone company charges extra for, the crooks) is to have your phone listed under a different name. So, if I'm listed in the phone book as "Nick Drake," I'm still listed, and so I don't get charged for being unlisted. However, anyone looking for D. Mandl in the phone book won't find me, so I'm _effectively_ unlisted. Strange but true. --Dave.

1 0

Yet more number theory
by Anonymous 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- Well, I'm the person posting all the number theory stuff anonymously. Well, not too anonymously since I am signing each post... ;) I thought I'd try out Matt Ghio's service. I'm not sure exactly what will happen, but hopefully you will able to reply to this message and reach me! Anyway, I got my copy of "Elementary Number Theory and its Applications" by Kenneth Rosen just now, and checked Miller-Rabin primality testing, and pseudoprime primality testing. Eric pointed out some recent work (by Pomerance I presume) and it does indeed junk the notion that for pseudoprime testing, the failure rate is 2^-n, n being the number of trials. However, Miller-Rabin isn't susceptible (it uses strong pseudoprime testing) - and what it even better is the latest bound is 4^-k! That is, if you pick k integers and perform M-R on n for each, the chance a composite will pass is less than (1/4)^k. And, there is no analogy of a Carmichael number for strong pseudoprimes. So I guess the bottom line is M-R is the way to go. -----BEGIN PGP SIGNATURE----- Version: 2.3a iQCVAgUBLas39YOA7OpLWtYzAQETVQP/YzHMudKp/ehgcG0MkBeoyhQsItAlAvXL VVj2VN2ac7KjlqtyP/Frjq+6s/T0ai4MhojboaWKBJfuUvZT1hBj0c0PvkaHVeiQ H1eJpEXEqbFoouRX/M7ZYLmwfeJenKn0th408gJBf6yDHwdv9dyo7//Hhd/GreWJ K+9nHl4k3kU= =9zRl -----END PGP SIGNATURE-----

1 0