December 2003 - cypherpunks-legacy

Kids and Computer Privacy Was Re: No more stupid gun thread ...
by Scott Schryvers 17 Dec '03

17 Dec '03

At 06:24 PM 7/22/96 -0700, you wrote: >Ok. This thread has gone on long enough and covered just about every >point except the one which I originally made in my first response, >which is that I abhor the idea that kids should carry weapons (of any >sort) to school as standard equipment. In fact, I abhor the idea >that kids should carry weapons at school for any reason. > >Enough said. > >I do not care to discuss: > >1. Should kids have any weapons at any time? > >2. Should kids have guns (specifically guns)? > >3. Should kids know how to operate weapons of any sort? > >If any of you really really have to discuss this issue, let's spare >the rest of the list and send me E-Mail directly. > >Thanks! > >Ern > Should kids have crypto? [weapons] Should parents teach their kids crypto at an early age? Should kids have a right to privacy? If yes, should there be a limit to the privacy they hold? Under Itar crypto is a weapon. If a kid were to bring a disk containing pgp to school could they be expelled for carrying a weapon? Should children's legal rights be taught at school at an early age?

1 0

emscrypt 0.01 ALPHA
by Matt Carpenter 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- I finally managed to find some time to do a little testing of emscrypt and make some of the changes that were suggested here earlier. It is still rather ugly and I haven't added many of the features/capabilities I hope to eventually implement. But I'm making it available so that people can play with it if they want. It looks like I'm going to be very busy for at least the next month trying to finish up the work for my master's degree, so I probably won't be making any major changes for a while. emscrypt's purpose is to automatically run PGP signed scripts received by e-mail and return the results to the submitter. emscrypt is a heavily mutated version of morepgp, originally by Jason Steiner and modified by Greg Spencer. Please realize that this program has NOT undergone extensive testing, so you may encounter strange behaviour. Make sure you read the documentation that exists. Using this program may make your system insecure, especially if you don't follow the installation procedure carefully. Use at your own risk. You should be able to find emscrypt and some documentation at: http://www.bmen.tulane.edu/~carpente Look under the "Other random projects or possibly useful stuff" heading. In order to try out emscrypt you need the following: * Perl -- emscrypt is written in Perl. * PGP 2.6.2 -- Other versions may work, but I haven't tested them and emscrypt relies strongly on knowing the format of the PGP output messages. * Procmail -- For passing incoming mail to emscrypt. This can be accomplished in other ways, but at this point I haven't tried any of them. If you don't have procmail you can still play with emscrypt by piping messages to it manually. I strongly suggest you do this anyway, to make sure you trust emscrypt to answer incoming mail. Here are the major things that have changed with emscrypt since I described it here a while ago: There are now two required headers which must be included in the body of the signed message: 'Reply-To:' and 'PGP-Key-Fingerprint:'. Both must appear before the beginning of the script. The beginning of the script is considered to be anything other than the above headers, blank lines, or lines that begin with '::' but that are otherwise empty. Duplicate headers will generate an error and prevent the script from executing. Case within the headers is not important. (You are free to use eLiTE d00dz capitalization techniques, just don't try to use kRe8yv sP3lliNg5.) 'Reply-To:' must be followed by a valid e-mail address, otherwise you won't get your results. (emscrypt will also recognize a 'Request-Remailing-To:' header in place of 'Reply-To:'. I added this so that emscrypt could be treated as a "remailer". The idea is that you could use software such as premail to automate the generation of messages going to emscrypt, so that you don't have to do the signing/encrypting manually. Then you could just send a script to yourself, "chaining" it through your emscrypt "remailer". But I don't think this will work without slight modifications to the remailer message generating programs, since emscrypt requires the 'Request-Remailing-To:' header within the signed body of a message (it would also be nice if the 'PGP-Key-Fingerprint' header could be automagically included)). The 'PGP-Key-Fingerprint:' header gives the fingerprint of the key used to sign the message and which will also be used to encrypt the results. It must also be included within the signed body of the message. Most of the other changes involved minor debugging and general clean-up (not that I consider it clean now). I also improved the error handling a little, but it needs more work, too. Right now emscrypt generally tries to generate, encrypt, and mail an error report if things go wrong. If that doesn't work, then it saves an error message to a log file. If you find any problems, please let me know. Comments, suggestions, etc. are also welcome. Thanks, - --Matt - -- mcarpent(a)mailhost.tcs.tulane.edu -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQEVAwUBMfXlaijtJAMyBnp9AQHAdAf+MQ/ZroKoLeYyQDYabVrIq1eLSQB6vpr+ 2tXu63wDbcUeFoeSFNx6Sar7DNtAJyJUlwVcVKlb5SOuYR/8aFDvAnIYuQLPfdXd xXjC4iv+Hh3hNx4ibeyAB4xbFmDYAMB19zEf6nhmJdxR03oFXP+Qfx2m/aN/LDKZ zVSjtOs/ujTa6ltP6r/9x1vdiqmSNuSCNLvL/f4YulfdzR8frF0uLyLmiDH6mUpm etKxSpIg4ZI+iy1YvvSd+FtA0F3XSziaLEepx4X8gYjZKP0YXPTEMGhTeWj1o6bP yl06KUIRTL5k56P0xnW2MezGN5c0Cz2W9TEa9NBbHxY1DgSTDWa3sg== =DTpY -----END PGP SIGNATURE-----

1 0

Re: DES brute force? (was: Re: Borders *are* transparent)
by Dan Bailey 17 Dec '03

17 Dec '03

If we choose a plaintext/ciphertext pair carefully, we can easily save ourselves some work (50%) while still making the attack a credible demonstration. The idea is to use each trial encryption twice. In _Differential Cryptanalysis of the Data Encryption Standard_, Biham and Shamir note the following, known as the complementation property of DES: if T = DES(P, K) where T is ciphertext, P plaintext, and k key, then: T' = DES(P', K') where T', P', and K' are the bitwise complement of the above. Now the interesting part. If two pairs (P1,T1) and (P2,T2) are available with P1 = P2' or T1 = T2', then an attacker can restrict his search to only the keys with LSB = 0. The attacker runs through the remaining 2^55 keys (with LSB = 0) and tests the results against both T1 and T2'. Since testing for equality is much faster than performing the actual encryption, time savings is on the order of 50%. Just a thought on how to save some cycles. Dan

1 0

Re: Digital Watermarks (long, getting off-topic)
by tcmay＠got.net 17 Dec '03

17 Dec '03

(My comments are really minor quibbles, based on my longtime use of DAT machines, which I now have three of.) At 2:19 PM 7/23/96, Alex F wrote: >The Ent. Ind. got what they wanted though. There are taxes, etc. >(some sort of import restrictions anyway) that keep DAT player/recorder >devices at around $700 per unit. By this time normally the prices >*should* be down to like $200 (using the CD industry as a guide) As I said, I have three DAT machines. They are complicated machines, having lots of moving parts and precise tolerances. (And they are prone to break!) CD players are vastly simpler. I'm not at all surprised that prices have remained at about the $400 level for DAT decks, and about the same for DAT portables. After all, camcorders, which use much the same technology, have also remained at about the same price. And I don't think the SCMS code had too much to do with mass-acceptance. Most comsumers, according to available figures, *buy* C-90 cassette tapes, and do not make their own. (That _you_, the CP list reader, may use your cassette deck to make tapes has little to do with the vast numbers of cassette users out there do...most don't know how to record with their cassette decks.) Pre-recorded DAT tapes were available for a while...they did not sell. I believe this was because DAT machine purchasers were sophisticated and new how to make CD-to-DAT copies, with or without SCMS. Thus, the failure of DAT as a consumer medium (not to mention the much-hyped MD and DCC formats) probably is due to other reasons, including the mechanical issues, the lack of a real need for consumer DAT, and the confusion over new emerging formats. --Tim May Boycott "Big Brother Inside" software! We got computers, we're tapping phone lines, we know that that ain't allowed. ---------:---------:---------:---------:---------:---------:---------:---- Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay(a)got.net 408-728-0152 | anonymous networks, digital pseudonyms, zero W.A.S.T.E.: Corralitos, CA | knowledge, reputations, information markets, Licensed Ontologist | black markets, collapse of governments. "National borders aren't even speed bumps on the information superhighway."

2 1

Re: A Snake-Oil FAQ
by David E. Smith 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- To: cypherpunks(a)toad.com Date: Wed Jul 24 00:10:54 1996 > Honestly, this boycott campaign looks out of place on Cypherpunks, at > least to me. I mean, we are for freedom of speech, aren't we? > Sternlight is talking about on-topic things. How come that renowned > defenders of freedom of speech resorted to name calling and attempts to > push their opponent out of the public forum? > (cheer) I don't agree with David Sternlight; frankly, I doubt very many of us agree with David Sternlight. But if we're going to start plonking just on the basis of a periodic rant, after he goes Perry goes, and Tim goes, and before long there's no list. Now to keep things interesting, somebody wanna subscribe Dorothy Denning to the list? Might be fun :) dave - ---- David E. Smith POB 324 Cape Girardeau MO USA 63702 dsmith(a)prairienet.org http://www.prairienet.org/~dsmith send mail of 'send pgp-key' subject for my PGP public key "I'm only a social smoker, just a few packs a day really" -----BEGIN PGP SIGNATURE----- Version: 2.6.2 Comment: Automagically signed with Pronto Secure for Windows. iQEVAwUBMfWwaTVTwUKWHSsJAQHmAwf/XkwZZtAbymiSni4L8N6D6qW75W2B/Hgv yuPX0uCoXnBmvUXzq6RS7zNN0A1vtgXhCgzgyYY+1/1PRAAoQAd8q7evX1K+rTVe 1MmlOfJWFHJG9q0tHc45LONagW1m0eAp9Z9fkO4NRhGpybfWDCZKwtClbaLiAdoA 86lJnNRacjEHf1dAAHc/1dibJz+6617nfgLkea0OVbLICIsSpkc3PJLkFE9jjK2a g+5CXJ8oXg4bKlEw7QoefagXYsM9o6PF+MHIDVT/pZnfoVmGoI9+3BYHkByXh/vQ 8M+w6hZu4X50zzxCx/6lDvzfd1zccmHvJAsx6GXf4VGUp6lIw+lfCg== =KXCp -----END PGP SIGNATURE-----

1 0

Re: Bare fibers
by Bill Stewart 17 Dec '03

17 Dec '03

h > > Doesn't that make it vulnerable (detectable) to Tempest attacks? h > > Harka rw> No. rw> Transmitting light via fiber doesn't emit EM. Light is electromagnetic :-) rw> Anyway, the original post, as I recall, was about keeping sensitive data rw> on a second hard drive, connected via (very thin, therefore harder to rw> notice) fiber. Tempest monitoring was not a factor. rw> -r.w. I assume that if you've got a fiber hanging from your PC, and thugs come in to steal/confiscate/forfeitize your PC, that they're bright enough to notice it and maybe to follow it. Thugs who want your information will follow it to find information; thugs who just want to resell your hardware will follow it to find more hardware. I had initially assumed that the mention of fiber optics was in the context of "Infrared transmitter on the PC, fiber optic sticking out of the wall to receive the IR and transmit it to a hidden detector." or something silly like that. If you want to hide a small PC in your attic/wall/etc and use the newer faster IR stuff for clandestine backups, and have a Real Operating System so you can run it in the background (since your disk drive is presumably much faster than IR), I suppose you could do that. The main use of TEMPEST here is to detect backup systems hidden in the attic that they hadn't noticed. Hiding it in your stereo system doesn't protect it from honest thieves, who might also want to resell your stereo, and info thieves have been known to seize anything even resembling computer equipment, such as Mozart CDs. You could also hide the computer in your kitchen cabinets like one Famous Cypherpunk, but you'd have to use a thinner bundle of cables than he does to connect the monitors and keyboards :-) # Thanks; Bill # Bill Stewart +1-415-442-2215 stewarts(a)ix.netcom.com # http://www.idiom.com/~wcs # Confuse Authority!

1 0

RE: Distributed DES crack
by koontz＠netapp.com 17 Dec '03

17 Dec '03

>I'd like to see a very general hardware processing power equivalence table. >For example, 1 MasPar equals how many Pentiums. One MasPar MP2 (4K processors) could run 300,000 crypt(3) crack attempts per second (a password checker). That should give you 6 or 7 million brute force key attempts per second. Unfortunately the only way to show an equivalance for a particular problem is to have both machines work on the same problem. (Perhaps we could manage to break a DES key in the course of executing a new benchmark?)

1 0

the VTW---FBI Connection (fwd)
by Sharma 17 Dec '03

17 Dec '03

I am asking for some verification that you (cypherpunks) have something to do with this before I forward it anywhere. I looked at the vertexgroup and so-oregon webpages, and there is no investigative journalism there that I was able to find. Please get back to me soon. Thanks, sharma sharma(a)aa.net >X-URL: <URL:http://www.vertexgroup.com> >X-URL: <URL:http://www.program.com> >X-URL: <URL:http://www.so-oregon.com> > >>Date: Sun, 21 Jul 1996 20:51:23 -0600 >>To: cypherpunks(a)toad.com >>From: nobody(a)zifi.genetics.utah.edu (Anonymous) >> >>We have received information that VTW is run and supported by the FBI, >>which we have suspected for reasons listed here. >> >>I) They appear to have no financial support or funding source. They >>do not accept donations. They have no corporate funds. And yet they >>appear to be a thriving business. >> >>II) They are secretive about their location, and do not seem to have >>a headquarters. The address listed in the NIC is a vacant lot in of >>all places Brooklyn, NY. They do not have a listed telephone. >> >>III) On a tip from a "friend" we learned that the power leader behind VTW >>is a cleancut man with the unusual name "Shabbir M. Safdar." Not exactly >>a common name. Several people say they've met him. Our "friend" says >>that Mr. Safdar is in reality an agent of the FBI. >> >>IIII) We didn't believe this without external verification. So we called >>the Brooklyn office of the FBI and asked for Agent Safdar. No such person. >>I called the Washington office. No such person. Checked if there is >>any agent named Safdar. They don't give out this info. Then we tried to find >>ANY public records on a Safdar, with no luck. No driver's license in NY, >DC, NJ, >>etc. No phone, etc. Odd that such a person does not exist and yet runs a >>"human rights" organization? Then Alice called the NY FBI office. Asked for >>Agent Safdar. Guess what? "He's not in. Can I take a message". No message, >>thanks. He'll get the message all right. >> >>Now the big question: What is the FBI trying to do getting all these >>names? What else has "VTW" been doing? And what other organizations like >>them are there? Who else is in on it? What does this say about EPIC, CDT, EFF? >>Are Blaze and Schneier dupes, or willing participants? What about their >>ISP? I think we are all owed an explanation. This is serious. Maybe FOIA or >>a lawsuit before they burn the files. >> >>What do you want to bet "VTW" quietly fades away after a few prefunctory >>denials, and gets replaced by another organization in due course? >> >>Faithfully, >> >>Net reporter team Alice and Bob >>

2 1

RE: Distributed DES crack
by Bill Stewart 17 Dec '03

17 Dec '03

At 03:40 PM 7/23/96 PDT, koontz(a)netapp.com (Dave Koontz) wrote: > > I'd like to see a very general hardware processing power equivalence table. > > For example, 1 MasPar equals how many Pentiums. > >One MasPar MP2 (4K processors) could run 300,000 crypt(3) crack attempts >per second (a password checker). That should give you 6 or 7 million >brute force key attempts per second. Interesting - thanks for the result. Do you know how tightly tuned the crack implementation was (e.g. straight C with MasPar optimizer, hand-tuned assembler, etc.)? Since the MasPar has a large number of very small processors, I'd expect it to be better at bit-twiddling than conventional processors. # Thanks; Bill # Bill Stewart +1-415-442-2215 stewarts(a)ix.netcom.com # http://www.idiom.com/~wcs # Confuse Authority!

1 0

Re: DES-Busting Screen Savers?
by Steve Reid 17 Dec '03

17 Dec '03

On the subject of choosing keys randomly, rather than dividing up the keyspace... This seems like a very good idea to me. One potential problem is actually choosing the random keys. Have to be able to get entropy at a fast rate, and/or use a good PRNG. Both of these things would take time away from the actuall cracking (more time than just incrementing the key). And if a weak PRNG is used in order to save time, it's possible that it might favour certain keys and thus interfere with the attempt. Any ideas for a fast and good PRNG? ===================================================================== | Steve Reid - SysAdmin & Pres, EDM Web (http://www.edmweb.com/) | | Email: steve(a)edmweb.com Home Page: http://www.edmweb.com/steve/ | | PGP (2048/9F317269) Fingerprint: 11C89D1CD67287E68C09EC52443F8830 | | -- Disclaimer: JMHO, YMMV, TANSTAAFL, IANAL. -- | ===================================================================:)

1 1