- cypherpunks-legacy - lists.cpunks.org

Re: Remember, RC4 is now PC1
by jrochkin＠cs.oberlin.edu 17 Dec '03

17 Dec '03

At 4:31 AM 02/28/96, Mutant Rob wrote: >It'd be funny if the next time somebody hacks a proprietary code, if they >make some changes, redo the key schedule, perhaps in mind of >strengthening the algorithm, and then post it to sci.crypt as "hey, I got >this idea for a new crypto algorithm... what do you think?". If the >algorithm is different enough from the proprietary code version, with >no clear connection between them, and the author can give full design >rationale as if s/he wrote it from scratch, then what's a company to do? Umm, if the _algorithm_ is different enough, it's a different algorithm and it's not even an issue. I guess you mean if the algorithm is the substantially the same, but the code implementing it is substantially different. But I'm not sure that matters anyway. The way I understand it with trade secrets is: If I'm an employee of PKP (let's pretend they have employees who actually look at code), and they want to keep something a trade secret, they make me sign a non-disclosure agreement. If I break it, and they can prove I broke it, I am in big trouble for breach of contract. I guess if I can make up a convincing enough lie about inventing it from scratch, I can get off. But if I work for PKP, I don't think I'm going to have too much luck convincing a jury that I just coincidentally stumbled on the same algorithm. If I, who has signed a non-disclosure agreement with PKP, takes the trade secret code and sends it to Mutant Rob, and Mutant Rob posts it near and far, Mutant Rob hasn't done anything illegal, and hasn't broken any contractual obligations, and is basically doing fine. As I understand it. So it doens't matter if he pretends he invented it himself or not. Of course, if they take him in the back room and introduce him to "Mr. Thingy" (or make him testify in court, if you prefer), and they find out it was me who sent it to him, I'm still in Big Trouble. Trade secrets don't really have any legal standing or protection, for the most part. They're just things a company is trying to keep secret, for the most part. Generally by using non-disclosure agreements.

1 0

Re: Simpler solutions (was Re: Stealth PGP work)
by Bill Stewart 17 Dec '03

17 Dec '03

At 10:26 AM 2/28/96 -0500, Bruce Zambini wrote: >Or, you can develop a public-key stego system... >ie a stego system that uses bits in specific ways depending on the >private key of the recipient. I assume you mean the public key of the recipient? >Something I've been thinking about, but haven't quite figured out how to >do yet, except that one could use certain bits based on a PRNG and begin >the message with an RSA-encrypted seed (ie the first X bits will be the >seed, encrypted with your public PGP key). That's equivalent to stego(picture, ( RSA(sessionkey,pubkey), Symmetric(message, sessionkey))) where Symmetric is a potentially cheaper (and weaker) algorithm than IDEA that possibly uses transposition rather than strict block-structuring. This approach is no more obscure than stego(stealth(PGP(message, pubkey))) though slightly more obscure than stego(PGP(message, pubkey)), optionally less secure, and probably not much faster except for long messages with wimpy Symmetric. And it still suffers from the non-stealthiness of vanilla RSA that Hal Finney, Adam Back, Harry Hastur, and others have discussed. Might as well keep the stego and encryption parts separate. You gain a certain degree of obscurity by using stego(picture, scramble( PGP(message, pubkey), key )) where scramble is some cheap symmetric encryption algorithm and key is either the recipient's public key or keyid, and PGP is in binary mode. This hides the message from eavesdroppers who don't know the recipient, but not from eavesdroppers who are willing to test against the keys of a list of usual suspects (assuming the recipient is one of them.) #-- # Thanks; Bill # Bill Stewart, stewarts(a)ix.netcom.com / billstewart(a)attmail.com +1-415-442-2215 # http://www.idiom.com/~wcs Pager +1-408-787-1281

2 1

Re: Diffie-Hellman for Matchmaking?
by Bill Stewart 17 Dec '03

17 Dec '03

At 11:37 PM 2/27/96 +0000, Dimitris Tsapakidis <dimitrt(a)dcs.rhbnc.ac.uk> wrote: >This is a followup to my "Dating Problem" question [.....] What's the reason for ZKP below? Is the objective that A and B only know the other one is interested if they have also advertised interest, and that nobody else can tell they're interested in each other? Or just that you can advertise who you're interested in so that only they know you're interested and nobody else does (except possibly a Trusted Third Party.) For the latter case, instead of publishing g^(AB)mod n, publish hash(g^(AB)mod n), where hash is some non-invertible function (maybe strong like MD5, or maybe cheap like a CRC32, depending on how often you're willing to risk collisions.) If you're concerned about replay attacks, use something like (timestamp, MD5(timestamp, g^(AB)mod n)). >Person A is interested to match person B, so he computes >g^(AB)mod n. B is interested in X, where X may or may not >be A, and calculates g^(BX)mod n. Now, they compare these >two "common keys" either using some Zero Knowledge scheme >that ensures fairness (at no point one party has significantly >more information than the other) or through a Trusted Third Party. >If they are the same, then this means X=A, so A and B >have a match (e.g. a date). The common keys must remain >secret (hence the ZK above): if g^(BX)mod n "escaped" >to the public, then the real X would find out that >B is interested in him. #-- # Thanks; Bill # Bill Stewart, stewarts(a)ix.netcom.com / billstewart(a)attmail.com +1-415-442-2215 # http://www.idiom.com/~wcs Pager +1-408-787-1281

1 0

Re: Anonymous Web Browsing
by Powers Glenn 17 Dec '03

17 Dec '03

- From: tcmay(a)got.net (Timothy C. May) - At 5:24 AM 3/2/96, dwhite(a)pris.bc.ca wrote: - > My question is this: Would it be possible to create a web site that - >would function along the same lines as remailers do? Something that would - >allow a person to browse anonymously. So lets say I surfed into a - Tim May writes: - Your intuitions are correct, that protection steps are needed. The - technology is developing, but the perceived need (anonymous browsing) is - probably less than the perceived need for anonymous e-mail. simple, not real secure solutions: find a unix host, run the cern httpd server, turn on proxying, turn off logs. (about 10-15 minute setup) if this is not run often, it works well. use netcom (or something similiar) that gives you dynamic ip addresses. as a bonus, netcom doesn't require an actual name or address. (yes, they require a CC#). this allows fairly anonymous access. (prehaps one reason the neeed for web proxies isn't that high) glenn

3 2

Re: Anonymous Web Browsing
by Bill Stewart 17 Dec '03

17 Dec '03

At 03:44 PM 3/2/96 +0000, "Mark Grant, M.A. (Oxon)" <mark(a)unicorn.com> wrote: >> These are "web proxies," and searching on this should produce some >> information to answer your questions. >.... hidden mailto: .... even worse with Javascript. > >Note to Netscape people: Is it possible to have an option that will *always* >pop up a mail window and request confirmation before mail is sent out? Or >disable mailto altogether? It would be a lot easier than binary editing >the executable to remove all the mailto strings... At least in the Windows version, you can set the SMTP server you want to use for your outgoing mail. So you could do a proxy email server that listens for SMTP and pops up confirmations for outgoing traffic. For the moment, you'd probably have to write your own, but the new Winsock Remailer probably has the code you need to do most of the work, and you'd only have to add some GUI buttons and whistles. #-- # Thanks; Bill # Bill Stewart, stewarts(a)ix.netcom.com / billstewart(a)attmail.com +1-415-442-2215 # http://www.idiom.com/~wcs Pager +1-408-787-1281

2 1

Re: A brief comparison of email encryption protocols
by Joseph M. Reagle Jr. 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- At 07:20 PM 2/29/96 -5300, Adam Shostack wrote: >key://ftp.clark.net/pub/u/cme/cme-current.asc >key://ftp.clark.net/pub/u/cme/cme-longterm.asc Just as a head up on not duplicating work with regards to incorporating URL's into key certificates and DBs you might want to consider URI's, URC's or URN's as the piece to bind since URL's are considered to be almost as unstable as email.. Check out for more on these naming schemes. Linkname: WRL: Web: URXs URL: http://webreference.com/urx.html -----BEGIN PGP SIGNATURE----- Version: 2.6.3i Charset: cp850 iQCVAwUBMTcfaP0Ael7bLK1/AQG56QQAxr+Tii4UMytdQUXIrKzIp2ckY4UfbGqW BpkduSxha0lL8Mo2kmzOL4Z3s4NiJMExvzUKOo6Y16MuCXhKgJyxH5VN+Nk1Y/ov qNpm4zaFOiAFQTbjMhSWJgFGmm6uaHuOez5jZDpSCpZTSBbXUwR349lxXjEtuvy8 hm+w0ADr/VQ= =SYWa -----END PGP SIGNATURE----- _______________________ Regards, There is no point at which you can say, 'Well, I'm successful now. I might as well take a nap.' -Carrie Fisher Joseph Reagle http://farnsworth.mit.edu/~reagle/home.html reagle(a)mit.edu 0C 69 D4 E8 F2 70 24 33 B4 5E 5E EC 35 E6 FB 88

1 0

RE: Nortel "Entrust"
by droelke＠rdxsunhost.aud.alcatel.com 17 Dec '03

17 Dec '03

Applied Crypto 2nd Edition - page 334-335. Brief stuff: Block Cypher (symetrical), 64 bit key, 64 bit block size "The S-boxes are implemenation dependent, but not key dependent." resistant to differential and linear cryptanalysis "Northern Telecom is using CAST in their Entrust security software package for Macintoshes, PCs, and UNIX workstations. The particular S-boxes they chose are not public. The Candadian government is evaluating CAST as a new encryption standard. CAST is patent pending." If you don't have the book - GET IT! ;-) Dan > From: "Stephen A. Gutknecht" <sgutknec(a)computek.net> > To: "'cypherpunks(a)toad.com'" <cypherpunks(a)toad.com> > Subject: RE: Nortel "Entrust" > > I too would like to see an open discussion on this. Microsoft uses this for > their new Exchange Server client/server email product.... > > ---------- > From: Jonathon Fletcher[SMTP:jonathon@japan.sbi.com] > Sent: Friday, March 01, 1996 10:17 AM > To: cypherpunks(a)toad.com > Subject: Nortel "Entrust" > > > Can anyone tell me anything about a product called "Entrust", by Nortel > (Northern Telecom). The notes talk about the software using DES ("which > employs a 56-bit key") so I guess it's single DES, not triple DES. It also > mentions a proprietary algorithm called CAST. > > Is this worthy of further investigation, or is it suspect ? What is > CAST, and would it be classed as snake oil ? > ------------------------------------------------------------------ Dan Oelke Alcatel Network Systems droelke(a)aud.alcatel.com Richardson, TX

1 0

Re: "Louis Freeh is a Cocksucker"
by jimb＠alpha.c2.org 17 Dec '03

17 Dec '03

Alan Horowitz says: > OVer on alt.something.somesomething-or-other.whitewater, there > are people keeping track of dates and events. It has been > noticed that Freeh was sworn in _just before_ Vincent Foster's > body was found. That's certainly very suspicious! I bet that Freeh *is* Vince Foster with plastic surgery! -- JimB "I feel good!"

1 0

Re: going back to stone axes
by Bill Stewart 17 Dec '03

17 Dec '03

At 02:32 PM 3/1/96 +1100, Steven Legg <s.legg(a)trl.telstra.com.au> wrote: >DistinguishedName wasn't defined the way it was because of ASN.1. I'd always assumed it was just the ugly X.400 email naming convention. >No one at Telstra Research would want to replace DistinguishedName with >a byte string because of all the constant reparsing, cutting and splicing >that would entail. We would end up writing more code, not less. Given the >sort of heavy duty processing we do to DNs, trying to pack a DN into a >string is considered brain dead. Those rich semantics matter to us. I don't see how a DN is any better than an Internet-style Domain Naming Service name, except that it puts a bunch of pre-defined labels on subdomain parts which can generally be guessed from context and aren't very relevant to computer programs. Compare joeuser(a)purchasing.fnord.edu.au vs. /FN=Joseph/Ln=User/NN=Joe/OU=Purchasing/O=First.National.Organization.- [for].Research.[and].Development/S=Academic/C=AU/ Is there really any additional information in the latter, except possibly some detail you could have looked up with a "whois" request? #-- # Thanks; Bill # Bill Stewart, stewarts(a)ix.netcom.com / billstewart(a)attmail.com +1-415-442-2215 # http://www.idiom.com/~wcs Pager +1-408-787-1281

1 0

No Subject
by owner-cypherpunks＠toad.com 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- At 07:20 PM 2/29/96 -5300, Adam Shostack wrote: >key://ftp.clark.net/pub/u/cme/cme-current.asc >key://ftp.clark.net/pub/u/cme/cme-longterm.asc Just as a head up on not duplicating work with regards to incorporating URL's into key certificates and DBs you might want to consider URI's, URC's or URN's as the piece to bind since URL's are considered to be almost as unstable as email.. Check out for more on these naming schemes. Linkname: WRL: Web: URXs URL: http://webreference.com/urx.html -----BEGIN PGP SIGNATURE----- Version: 2.6.3i Charset: cp850 iQCVAwUBMTcfaP0Ael7bLK1/AQG56QQAxr+Tii4UMytdQUXIrKzIp2ckY4UfbGqW BpkduSxha0lL8Mo2kmzOL4Z3s4NiJMExvzUKOo6Y16MuCXhKgJyxH5VN+Nk1Y/ov qNpm4zaFOiAFQTbjMhSWJgFGmm6uaHuOez5jZDpSCpZTSBbXUwR349lxXjEtuvy8 hm+w0ADr/VQ= =SYWa -----END PGP SIGNATURE----- _______________________ Regards, There is no point at which you can say, 'Well, I'm successful now. I might as well take a nap.' -Carrie Fisher Joseph Reagle http://farnsworth.mit.edu/~reagle/home.html reagle(a)mit.edu 0C 69 D4 E8 F2 70 24 33 B4 5E 5E EC 35 E6 FB 88

1 0