December 2003 - cypherpunks-legacy

How trustworthy is *Skipjack*?
by grendel＠netaxs.com 17 Dec '03

17 Dec '03

It seems to me, that with the NSA being the devious bastards that they are, might not just make the key escrow scheme the only weak point in the Clipper encryption system. The escrow scheme (supposedly) allows only legal wiretaps, as you need a warrant or wiretap order in able to obtain the escrowed key parts. HOWEVER: The NSA known to be damn secretive about what exactly it is doing and investigating. If they were snooping on something, especially something they're not supposed to be involved in (domestic surveillance, for instance), they wouldn't want to go to the trouble of legally obtaining a wiretap order. They take so long, and there's always the trouble of that potentially incriminating paperwork... In short, is it possible that the NSA built in a backdoor to Skipjack so they can stay ahead of the game like they used to in the pre-Clipper days? Also, at the risk of starting a "computing power increases vs. difficulty of factoring vs. potential cost" flamewar: Is it possible that the NSA can brute force Skipjack? _Applied Crypto_ says that Skipjack only has an 80-bit key. How easily could the NSA break an 80-bit key? -- ========================================================================== | Michael Brandt Handler | Philadelphia, PA | <grendel(a)netaxs.com> | | PGP 2.3a public key available via server / mail / finger | ==========================================================================

1 0

--who is doing the stego faq?--
by rarachel＠prism.poly.edu 17 Dec '03

17 Dec '03

Hi guys, I'm an idiot. :-) A few days ago, I saw a message by someone on sci.crypt about a stego faq he was doing. I contacted him and told him about a program I'm working on, but lost his address and the news daemon already purged his message. :-( So if you're the one, or know who is, please email me. The next message will be some info on WNSTORM which is the program I'm writing...

1 0

Re: cpunks quiz
by jim＠bilbo.suite.com 17 Dec '03

17 Dec '03

It came from one of the documents that John Gilmore received as a result of one of his FOIA requests. Here the relevant section from John's post... From: gnu(a)toad.com (John Gilmore) Return-Path: <gnu> Received: from localhost by toad.com id AA19157; Thu, 30 Dec 93 02:21:27 PST Message-Id: <9312301021.AA19157(a)toad.com> To: cypherpunks(a)toad.com Subject: Revised Clipper FOIA results from Asst Secretary of Defense Date: Thu, 30 Dec 93 02:21:27 -0800 We sent in an administrative appeal on June 17th, 1993, of various things that were withheld in the response to our FOIA request. The Office of the Secretary of Defense responded on December 21, 1993 -- six months later. (By law, agencies have twenty business days to respond to an administrative appeal. However, agencies regularly violate all FOIA time limits because the courts have largely refused to censure agencies for breaking the law, and have refused to force agencies to follow the law. I will point this out each time it happens, largely to educate you -- the general public -- about how pervasive a problem this is.) We did an administrative appeal of the parts they withheld and other documents they did not provide. The result is that one more doc came out (a cover sheet for a review copy of the President's actual directive, which is still classified and has been referred back to the National Security Council for processing), and the previously withheld paragraph of the last two memos below is now only blacked out for a sentence or two. The newly released text is highlighted with XXXX's and explanation. John Gilmore [first few letters deteled -jm] [This page originally XXXXXXXX SECRET; now UNCLASSIFIED] OFFICE OF THE ASSISTANT SECRETARY OF DEFENSE WASHINGTON DC 20301-3040 COMMAND, CONTROL, COMMUNICATIONS AND INTELLIGENCE 30 APR 1993 (stamped) MEMORANDUM FOR THE ACTING ASSISTANT SECRETARY OF DEFENSE (C3I) Subject: PRD/NSC-27 Advanced Telecommunications and Encryption (U) [first six paragraphs deleted -jm] (U) Despite these concerns, the President has directed that the Attorney General request that manufacturers of communications hardware use the trapdoor chip, and at least AT&T has been reported willing to do so (having been suitably incentivised by promises of Government purchases). The Attorney General has also been directed to create a system for escrow of key material. The Secretary of Commerce has been directed to produce standards based on the use of the trapdoor chip. [remainder of letter deleted] (signed) Ray Pollari Acting DASD (CI & SCM)

1 0

WNSTORM UPDATE (STEGO RND CRYPT)
by rarachel＠prism.poly.edu 17 Dec '03

17 Dec '03

The following describes the a crypto program I've written. This is an update to WNSTORM which is on csua.berkeley.edu (?) was soda... Mucho improvements have been made. Unfortunately, it will probably be a while before this winds up in a readable directory... so until then you can request it from me, and as long as you're in the USA I'll send you a copy... Here's the blurb... (send requests to rarachel(a)photon.poly.edy rather than replying to this message if you want me to get to them quicker...) WNSTORM(tm) 2.10 Copyright (C) 1994 by Ray (Arsen) Arachelian All Rights Reserved, GTDR. DO NOT EXPORT!!! USA AND CANADA USE ONLY!!!!!!!!!!!!!!!!!!!!!! WNSTORM (White Nouse Storm) is a new cryptography and steganography software package which I'm developing. You can use WNSTORM to encrypt your files to keep prying eyes from invading your privacy. You can use WNSTORM to hide files, be they text, or otherwise within PCX images (other image formats will be supported in the future.) WNSTORM's method of hiding files in PCX images is very secure. With the right kind of image, you won't even see any changes made to the picture unless you look very very closely, and even then you are unlikely to recognize that there is hidden data in the image. (Some images don't work well, others work very well.) You can then take the PCX image containing the hidden data and send it to your friend, or post it up publicly on the alt.binary.pictures.* or whatever. Only you and anyone you share the password with can get at the hidden data file! More importantly, the intended recipient of the picture does not have to have a previous copy of the image to get at the hidden data. WNSTORM is also an encryption system. Unlike conventional systems, WNSTORM uses random numbers to hide and encrypt your file. (see documentation for technical details.) WNSTORM's encryption uses elements of frequency hopping (also known as spread spectrum transmissions) but instead of using it over radio waves, it uses it on your files. It does this by breaking up your file into eight separate channels and mixing in anywhere between 16 to 240 other bogus random noise channels with your eight data channels. The number of bogus channels varies with each byte taken from your file. The position (similar to a frequency of a channel used by spread spectrum communications) depends on the values in the bogus channels as well as your password. The password can be upto 150 characters (that's 1200 bits) however this can easily be changed to longer passwords. All the source code for WNSTORM and its supporting programs is also included. It's there for you to examine and see that there are no back doors, and to modify so as to improve it. WNSTORM is >FREE<, but not public domain. You may freely use it and give copies away to anyone you like. (see documentation for license information.) Enjoy. -- Ray (Arsen) Arachelian rarachel(a)photon.poly.edu

1 0

Re: D-H key exchange - how does it work?
by Hal 17 Dec '03

17 Dec '03

The problem with "strong" primes, primes for which (p-1)/2 is prime, is that they are hard to find. It takes hours and hours of searching to find a 1024 bit strong prime on a workstation. Granted, you don't need to change very often perhaps, but some people would like to change every day. They may need a dedicated prime-searching machine to do that. (The best way I know to find strong primes is to find a prime q and then check 2q+1 for primality. Finding 1024 bit primes takes a long time, and the chances that 2q+1 is prime is very low.) It's much easier to find a "strongish" prime, one for which (p-1)/k is prime, where k is on the order of 100 or so. Take your prime q in the above and try kq+1 for k=2,4,6,.... This only takes a few minutes after you find q. The question is, how good are strongish primes? What fraction of elements of the group will have short periods, given that p-1 has a pretty small number of prime factors? Also, given a strong or strongish prime, are the chances that g^x has a small period good enough that it makes sense to check for that case? Any event whose chances are smaller than your computer making a mistake is generally not worth checking for. Hal

4 6

Crypto-history
by CHRISTOPHER JAY LEONARD 17 Dec '03

17 Dec '03

Crypto-history buffs, Apropos the recent discussion of the Bletchley Park and NSA exhibits of WWII rotor-type cipher machines on sci.crypt, there is an excellent article by Glenn Zorpette (of IEEE Spectrum) in the Summer 1994 issue of American Heritage of Invention and Technology, Volume 10/Number 1. This is a Forbes publication that is put out by their American Heritage Magazine group. The title of the article is "The Edison of Secret Codes" and it is about Edward Hugh Hebern who contributed much to the development of rotor-type cipher machines from 1912 until after WWII. It features quotes from David Kahn (The Codebreakers) and some terrific photos of early cryptographic equipment, as well as a nice description of the design/function of the machines and vulnerabilities to cryptanalysis by the Army's William Friedman in the 1920's. There lawsuit by Hebern against patent infringement by the government serves as an an eerie premonition of the Inslaw case. There is a sidebar regarding the National Cryptologic Museum. Last but not least it features this marvelous stanza of an ode to Hebern's crypto machine, an equally fitting a description of public key systems (e.g. PGP). Marvelous invention comes out of the West Triumph of patience, long years without rest Solved problem of ages, deeper than thought A code of perfection, a wonder is wrought. I got my copy directly from someone who works for American Heritage, so I can't vouch for it's availabilty on the news stands. Share and Enjoy :-) cjl

1 0

Sternlight "kill" file
by Tom Allard 17 Dec '03

17 Dec '03

Well, Sternlight has threatened to talk to my superiors (no one is superior to me :) if I don't apoligize to the world for forwarding beker(a)netcom.com's message posted to this list to alt.security.pgp. He told me that he knows the Chairman [of the Federal Reserve Board]. Would Alan Greenspan get involved? Sure, he's kinda busy with that inflation thing... He told me his message to me was copyrighted and I couldn't publish it. I'm going to take that advice while I ponder this. Any advice would be welcome. My pgp key is on the servers. PS: I'm considering pleading "forgery". rgds-- TA (tallard(a)frb.gov) I don't speak for the Federal Reserve Board, they don't speak for me. pgp fingerprint: 10 49 F5 24 F1 D9 A7 D6 DE 14 25 C8 C0 E2 57 9D

2 1

Re: Sternlight "kill" file
by Richard Johnson 17 Dec '03

17 Dec '03

From the keyboard of: Tom Allard <m1tca00(a)FRB.GOV> > Well, Sternlight has threatened to talk to my superiors (no > one is superior to me :) if I don't apoligize to the world for > forwarding beker(a)netcom.com's message posted to this list to > alt.security.pgp. Heh. Sternlight is really trying to throw his weight around here, isn't he. How sad, but not surprising. His job will be much easier if he can silence some of the opposition to his half-truths and innuendo. As Jim Thomas notes in the following post to alt.security.pgp, M. Sternlight is somewhat ignorant of Copyright law and precedent, including such things as fair use, implicit assignment of rights, etc. Note that Mr. Thomas posted his missive to a usenet newsgroup that receives posts from cypherpunks. The fact that we have to use manual means to make the gateway go in two directions makes it no less valid a gateway than those between other mailing lists and usenet newsgroups (like comp.society.cu-digest, for example :-). --------- From: jthomas(a)well.sf.ca.us (Jim Thomas) Newsgroups: alt.fan.david-sternlight,alt.security.pgp Subject: Re: David Sternlight's Slurs About Folks With "2.3a" Keys Date: 20 May 1994 04:02:28 GMT Message-ID: <2rhcok$sbq(a)nkosi.well.com> In article <strnlghtCq2Eqs.2xB(a)netcom.com>, David Sternlight <david(a)sternlight.com> wrote: >By publishing a private e-mail, both beker and Johnson are in violation of >my copyright in that message, since it was a private communication and no >permission was given to republish. Not surprisingly, Sternlight reveals his ignorance of copyright law. >Thus beker has now committed a new violation of his netcom agreement, and >Johnson shows he has bad judgement. What's more, by trying to make what >should be a private matter, to be decided by netcom based on their own >rules, into a public cause celebre, and by writing to netcom as Johnson did >to intervene in a private matter, he has made the situation worse for beker, >not better. Sternlight has failed to demonstrate: 1) That the first alleged violation was, in fact, a violation, and not simply a labeling ruse. Sternlight has rushed to judgment and tried and convicted without evidence. An honorable person would first ascertain facts prior to taking action 2) Sternlight claims a second violation of netcom's agreement without demonstrating what the violation is. An honorable person would reproduce the relevant text of the agreement and then make the corresponding case. Instead, Sternlight asserts. This is consistent with his style in which he defames others and then complains that he is defamed when others hoist him by his own petard. 3) That Sternlight snoops through others' stuff and then leaps to judgment on the bases of superficial cues is, indeed, a matter of public concern. If Sternlight snooped through my system files and found titles such as suckme.gif, jailbait.gif, and 69riders.exe, would he have complained to our university officials that I am in violation of school anti-porn policies? If the facts of Sternlight's latest escapades are accurate, and Sternlight's post seems to confirm them, then it is fully appropriate to alert the public that a demonstrable defamer is actively perusing accounts and notifying sysads of what he finds. Perhaps Sternlight should look up "honor" in his dictionary. Jim Thomas ... >David

1 0

Re: Sternlight "kill" file
by Peter Murphy 17 Dec '03

17 Dec '03

I was aware that David Sternlight had a dubious reputation among the cypherpunks community, but uptil now I was never really sure why. Now I know. Could anyone please tell me, for my benifit (and possibly for other newbies' benefit) what other "escapedes" he has pulled in the past that have been shonky. I would also like to know why he has an instinctive dislike of PGP. (Much of it may have happened before my time.) ======================================================= | Peter Murphy. <pkm(a)maths.uq.oz.au>. Department of | | Mathematics - University of Queensland, Australia. | ------------------------------------------------------- | "What will you do? What will you do? When a hundred | | thousand Morriseys come rushing over the hill?" | | - Mr. Floppy. | =======================================================

2 1

Why PGP 2.5 sucks...
by ER CRAMER 17 Dec '03

17 Dec '03

PGP 2.5 really sucks... Why? - You can't use your old secret key. So you have to build a new one. Why? I don't know. But PGP 2.5 don't recognize your pass phrase anymore. Maybe that new RSA algoritme is easier to crack. So they force everybody to create a new key... - A new RSA algorithm??? What's wrong with the old one??? - Why not bigger keys than 1024 bits??? - Why can't PGP 2.3a read PGP 2.6 messages anymore after 1 september 1994? The only good thing about 2.5 is that it's legal in de USA and Canada? But why doesn't someone gets a license for PGP 2.3a so that's a legal version too! (or is that inpossible???) ... If you outlaw Privacy, only the Outlaws will have Privacy! Eelco Cramer <crame001(a)hio.tem.nhl.nl> ------ --------------------------------------------------

2 1