- cypherpunks-legacy - lists.cpunks.org

PGP on soda.berkely.edu
by Hal 01 May '93

01 May '93

-----BEGIN ??? SIGNED MESSAGE----- I want to apologize to Eric and Tim for intimating that their motives in exploring the possibility of closer ties to RSADSI were due to pressure from that company and its officers. I had apparently misinterpreted Eric's statements about the popularity of the Cypherpunks FTP directories to conclude that it was a major distribution site for PGP. I am glad to hear that Eric and Tim are not facing any immediate legal problems due to their support for this software. Turning to the issue under discussion, I do have a couple of other thoughts. First, I don't see that the interests of RSADSI are fully aligned with ours regarding Clipper. Despite PKP's success in accumulating patents, Clipper per se does not appear to infringe, being based on a new symmetric cryptosystem. So they don't have any direct leverage over the use of Clipper. Now, Clipper-based phones presumably need some way to exchange keys, and here PKP's patents are likely to be relevant. But I was under the impression that AT&T, at least, was already producing secure telephones. I don't see why whatever arrangements they made for key exchange under their previous technology would have to be changed with Clipper. In fact, Clipper in some ways represents a major market opportunity for PKP. To the extent that the publicity leads to increased sales of encrypting phones, PKP may benefit from the success of the Clipper. (The follow-on Capstone project does appear to pose a greater threat to PKP, since it will use DSS (for key exchange???).) Furthermore, in any future government prohibition on non-Clipper cryptography, our greatest nightmare, it is plausible that the government would "take care" of PKP by making sure that they get a nice piece of the pie. I could easily imagine a situation in which non-Clipper crypto is banned, Clipper is widely distributed, and PKP is doing very well financially with a slice of the profits from every sale. Even if Jim Bidzos were personally committed to widespread, strong, public cryptography, and opposed Clipper for fundamental philosophical reasons (just like us), he would be faced with a conflict of interest. As several people have pointed out here, Bidzos has a fiduciary responsibility to his shareholders to maximize profits for his twin companies. If it comes down to a choice between opposing Clipper on principle and accepting it along with guaranteed profits, he may be forced (in the same sense in which he is forced to send threats to Stanton McCandlish) to back Clipper. So, even if Bidzos is personally a nice guy I think we need to remember that his company may not be a natural ally of ours. One final point, for now. I like Tim's .sig and all it represents. But frankly, it is hard for me to square a commitment to radical change with the proposed alliance with PKP. Part of the trouble is that I still don't understand exactly what our relationship with RSADSI is proposed to become. But at a minimum it sounds like we would avoid supporting activities which would infringe on their patents. That means that when we want to start working on some of those things in Tim's .sig, we are in many cases going to have to get Jim Bidzos's permission. Can you imagine asking something like this: "Dear Jim: We request permission to use the RSA algorithm for an implementation of digital cash which we will distribute in an underground way among BBS's all over the world, with the goal being the support of "information markets, black markets, [and] smashing of governments" (to quote Tim's excellent .sig). "Please sign on the dotted line below. Yours truly, an anonymous Cypherpunk." Obviously there is no way Bidzos could give such approval. Even if he personally were a card-carrying member of the Anarchist Party he could not bear the legal liability that someone in his position would take if he granted this request. How, exactly, are we supposed to progress towards Crypto Anarchy if we have to be sure not to step on PKP's toes? Do we just not ask him for permission (in which case we are in PGP's boat)? Do we ask for permission without revealing the full scope of the project (in which case it may be rescinded later)? I am not being facetious here. I honestly don't see how you can carry out Cypherpunk activities with a corporate sponsor. I guess that's enough for now... Hal 74076.1041(a)compuserve.com -----BEGIN ??? SIGNATURE----- Version: 2.2 iQCVAgUBK+HUu6gTA69YIUw3AQF9hAP+K6HXxXxjpK2qmjtFmj6LnWFW10KG09P+ o09BpbCJsiXTulv85XEtDfTyqus+T9o2dp01xaJaj0T/En3nKPs7NjKlgNciLmhV 3gzAAuv3VedheUR4cLuZOKxk6MkcwywRB4T/PHPomJ411FeYHI1DgBxZEbpM25e0 Y5mk4vQP+oo= =zKde -----END ??? SIGNATURE-----

1 0

Re: Introduction...
by Sy Verpunc 01 May '93

01 May '93

I wrote: : To whoever it was at Ohio who just fingered this account... (whether it's : one of us or the obligatory NSA mole :-) [oops, no, the NSA mole works out : of uunet.ca, right?]) - I'll save you the bother - it's not a cunning alias, What a nosey bunch... I've had this machine on the net for almost a year and not been fingered by as many strangers as over the last three days. Anyway, here's a little experiment in paranoia... which of these people *aren't* cypherpunks...? I've always wondered if joining a list like this would bring unwelcome attention... I've marked the people who I know are on the cypherpunks list with a *. If any of the rest of them are 'one of us', let me know. By elimination we'll work out where the gummint spies hide out :) ... (The machines that reject finger requests and telnet VRFYs are pretty suspicious...) And I'm also particularly suspicious of the finger from ohio-state - it *was* someone who reads this list (the svp ID was created especially for it) but there are *no* ohio-state sites in the mailing list at all. If this is considered a waste of the list's bandwidth, flame me privately... In fact, reply privately whatever you have to say... (gtoal(a)gtoal.com) Apr 27 20:03:05 finger from serdlc21.essex.ac.uk [155.245.11.40] Apr 27 20:03:18 guest Guest p3 Wed 20:00 Apr 27 20:03:18 millph P H Mills p5 Wed 19:44 Apr 27 20:03:18 guest Guest p6 Wed 20:25 Apr 27 20:03:18 guest Guest p7 Wed 20:31 Apr 27 23:10:19 finger from PANAM1.PANAM.EDU [129.113.1.2] Apr 27 23:10:44 GMJ2393B Jimenez, George M 20401FE1 FINGER LTA6549DSCC08/PORT4 Apr 27 23:13:08 finger gtoal from uunet.ca [142.77.1.1] Apr 28 19:48:53 finger from harrip(a)serdlc23.essex.ac.uk [155.245.11.42] Apr 28 19:49:13 harrip P Harrington p2 Thu 19:47 Apr 29 18:58:05 finger svp from mathserv.mps.ohio-state.edu [128.146.110.31] Apr 29 19:32:17 finger from mathserv.mps.ohio-state.edu [128.146.110.31] [! denotes no idle time during both fingers] akos alden aparson baker bkm bloch bogdan cao carlson chohan cthomas! davis dijen edgar falkner fcarroll fiedorow! forest ggelder! goedde goss haar hamilton! haradako harmon henri holbrook hpallen huneke jocha kappeler lguo ling! lingshu march morlet! neumann! nevai ogle overman patmac! ponomar rld robertso! root siegel singhi! sinha! sinnott stanton tanveer terman wang Apr 29 19:38:15 *finger gtoal from ee92jks(a)monge.brunel.ac.uk [134.83.72.1] Apr 29 19:38:20 ee92jks Jonathan K Saville *p1 Fri 20:03 cc-02.brunel.ac. Apr 29 19:44:11 finger svp from seneca.SED.Provo.Novell.COM [137.65.96.1] Apr 29 19:45:11 finger gtoal from seneca.SED.Provo.Novell.COM [137.65.96.1] ?*<Tim_Myers(a)Novell.COM (Myers, Tim)> Apr 29 23:23:21 finger gtoal from sck(a)naucse.cse.nau.edu [134.114.64.1] Apr 29 23:23:27 sck Sean Koontz *p1 Fri 13:41 =*<koontz(a)raslon.iac.honeywell.com> Apr 30 00:01:39 finger gtoal from dent.uchicago.edu [128.135.72.13] PS Someone mailed me to ask how I knew if I'd been fingered - it's a combination of a home-hacked fingerd and the log_tcp wrapper package.

1 0

Re: 800 numbers
by Sy Verpunc 01 May '93

01 May '93

From: Nickey MacDonald <i6t4(a)jupiter.sun.csd.unb.ca> Of course, only Americans can dial those numbers.. which is a shame... (A few Canadians could have skewed the voting... :-) On a note similar to remailers, has anyone ever given consideration to creating an "outdial" service? If I could dial long distance into the States, and then be able to dial a 1-800 number that I cannot dial direct from Canada (which is most) this would extremely useful... Could be taken mail me your pgp key... G

1 0

unsubsribe
by cplai＠csie.nctu.edu.tw 01 May '93

01 May '93

Please, unsubsribe me now. Please, unsubsribe me now.

1 0

PGP on soda.berkeley.edu
by Hal 01 May '93

01 May '93

The fact that PGP is available on the Cypherpunks FTP site, soda.berkely.edu, was publicized in sci.crypt last week (among other places and times). Eric said that the Cypherpunks FTP site has become among the top FTP access points in the world. When Stanton McCandlish posted on sci.crypt that he had PGP available on his system, he quickly received threats from Jim Bidzos demanding that he remove it. It follows, doesn't it, that Jim Bidzos would be forced to apply similar legal pressure to the Cypherpunks leadership to get them to remove PGP from their site, a site which may be one of the leaders worldwide in distributing PGP. Now we see Tim suggesting (in the spirit of "pancritical rationalism"? I'm not so sure...) that we move away from PGP, and Eric offering arguments showing why Bidzos has to do what he is doing. Is there a connection? Are Tim and Eric under pressure from Bidzos to get PGP off their site? I don't know, but they would certainly be likely targets for Bidzos' efforts, and so it is understandable that they may be feeling personal discomfort. But I'd like to hear more about the relationship with Bidzos. Hal

3 2

Re: I don't take it personally...
by tien 01 May '93

01 May '93

Tim, I'm responding to your question about patents. I don't know of any reason why individual licenses to use the public-key patents in connection with specified authorized programs couldn't be granted under U.S. patent law. In general, a patent holder has the right to exclude unauthorized persons from making, using, or selling the patented invention. A patent holder can grant non- exclusive licenses to make, use, or sell. Legally speaking, a non-exclusive license can be thought of as a promise by the paent holder not to sue the licensee for acts within the scope of the license which would otherwise constitute patent infringement, in return for some kind of consideration. The major limits on a valid patent are "patent misuse" (where the patent holder overreaches) and antitrust (where the patent holder uses the patent to restrain trade). Patent licensing is a pretty flexible mechanism. I'm no expert, though, and I've never quite understood exactly what this patent protects, nor have I seen the actual claims of the issued patent. Lee

1 0

eurocrypt '93
by phr＠america.Telebit.COM 30 Apr '93

30 Apr '93

Is anyone thinking of going? It's in Norway, in late May. Pining for the fjords... Paul

1 0

clip.txt on NIST ftp
by wcs＠anchor.ho.att.com 30 Apr '93

30 Apr '93

csrc.ncsl.nist.gov has some new uploads in /pub/nistnews, including Krammer(NIST)'s testimony to Markey, and clip.txt, a more recent Clipper description. It's short enough I'm attaching it below; the interesting part in Section 5 is that it says the wiretap block has 80-bit session key crypted with unit key, 25-bit serial number!, and 23-bit "authentication pattern" making a total of 128 narc bits. It also gets transmitted "at least once" to the receiver. Also, chip prices include "programmed and unprogrammed" - I wonder if real people will be able to buy them unprogrammed? Bill Stewart ------ clip here ----- CLIPPER CHIP TECHNOLOGY CLIPPER is an NSA developed, hardware oriented, cryptographic device that implements a symmetric encryption/decryption algorithm and a law enforcement satisfying key escrow system. While the escrow management system design is not completely designed, the cryptographic algorithm (SKIPJACK) is completely specified (and classified SECRET). The cryptographic algorithm (called CA in this paper) has the following characteristics: 1. Symmetric, 80-bit key encryption/decryption algorithm; 2. Similar in function to DES (i.e., basically a 64-bit code book transformation that can be used in the same four modes of operation as specified for DES in FIPS 81); 3. 32 rounds of processing per single encrypt/decrypt operation; 4. Design started by NSA in 1985; evaluation completed in 1990. The CLIPPER CHIP is just one implementation of the CA. The CLIPPER CHIP designed for the AT&T commercial secure voice products has the following characteristics: 1. Functions specified by NSA; logic designed by MYKOTRONX; chip fabricated by VLSI, Inc.: manufactured chip programmed (made unique) by MYKOTRONX to security equipment manufacturers willing to follow proper security procedures for handling and storage of the programmed chip; equipment sold to customers; 2. Resistant to reverse engineering against a very sophisticated, well funded adversary; 3. 15-20 MB/S encryption/decryption constant throughout once cryptographic synchronization is established with distant CLIPPER Chip; 4. The chip programming equipment writes (one time) the following information into a special memory (called VROM or VIA-Link) on the chip: a. (unique) serial number b. (unique) unit key c. family key d. specialized control software 5. Upon generation (or entry) of a session key in the chip, the chip performs the following actions: a. Encrypts the 80-bit session key under the unit key producing an 80-bit intermediate result; b. Concatenates the 80-bit result with the 25-bit serial number and a 23-bit authentication pattern (total of 128 bits); c. Enciphers this 128 bits with family key to produce a 128-bit cipher block chain called the Law Enforcement Field (LEF); d. Transmits the LEF at least once to the intended receiving CLIPPER chip; e. The two communicating CLIPPER chips use this field together with a random IV to establish Cryptographic Synchronization. 6. Once synchronized, the CLIPPER chips use the session key to encrypt/decrypt data in both directions; 7. The chips can be programmed to not enter secure mode if the LEF field has been tampered with (e.g., modified, superencrypted, replaced); 8. CLIPPER chips will be available from a second source in the future; 9. CLIPPER chips will be modified and upgraded in the future; 10. CLIPPER chips presently cost $16.00 (unprogrammed) and $26.00 (programmed). 4/30/93

1 0

Re: PGP on soda.berkeley.edu
by John (FuzzFace/Fast-Eddie) McMahon 30 Apr '93

30 Apr '93

> What founding father said "If we all don't hang together we will surely hang > seperately" Ben Franklin, I believe. John McMahon

1 0

Re: npr report
by karn＠qualcomm.com 30 Apr '93

30 Apr '93

> This is a *wonderful* analogy. Imagine if one could not buy a set of > locks for their house w/o buying the government lock-box to go beside > the front door. It's an even better analogy (for me) because of the experience I had a year ago with a house I was renting. The owner had put it up for sale, so a lockbox was installed on the garage door. One day I came home and noticed that the box was open, and the key inside was missing. Gave me a real warm and fuzzy feeling. One of the really nice things about now owning a house (vs renting) is that I can change the locks whenever I please and I don't have to give a copy to *anybody*... Phil

1 0