December 2003 - cypherpunks-legacy

Re: Returned mail: User unknown
by aba＠dcs.exeter.ac.uk 17 Dec '03

17 Dec '03

[whoops mispelled cpunks on cc line 1st time, sorry for 2 copies] Perry Metzger <perry(a)piermont.com> wrote: > Gary Jeffers <gjeffers(a)socketis.net> writes: > > Well, I just used MIT's PGP 2.6.2 with 3 different users' public > > keys to encrypt 3 different files. In all 3 files, the first 3 > > characters were the same (an umlauted A, then an i with an up arrow > > over it, and then a heart). This beginning 3 character string is > > apparently the infamous PGP RSA signature. The signature that says > > to spooks' programmed encryption sniffers - "HEY! I'M PGP - GIVE ME > > A LOOK!." > > As if they couldn't figure it out anyway. It isn't an "RSA signature" > by the way. Read format.doc sometime. Yeah, that's what stealth does, removes the boiler plate stuff saying, various things, such as pgp version number, rsa encrypted message, conventional idea block, and as Perry says, it's all in pgformat.doc in the PGP docs directory. > > When are the PGP designers and coders going to get serious and de- > > velope STEALTH PGP inside PGP itself!? > > Never, I hope. It would dramatically lower the utility of the > system. Can you imagine how disgusting it would be to try decrypting > something if you have a dozen keys outstanding? Not to mention how > hard it would be to deal with figuring out that you should even try to > decrypt things in the first place. I reckon it would be a very nice utility, built into pgp as an *option*, some countries it isn't legal to use PGP, and it hasn't been ruled out that the US may not be the next to join the list, all this GAK stuff, son of Clipper, it is a distinct possibility that the law enforcement lot might outlaw other crypto wiht mandatory GAK. Then stealth features for PGP become important for incorporating into stego, of course you can argue that well stealth should be part of the stego app then. It shouldn't be a problem unless it was enforced, as an option I see no problem with it, and lots of possible future advantages even. I'm trying to work on stealth2.1, modifing Henry Hastur's stealth util to added Hal Finney's algorithm for improving the stealthiness (it's not enough to just strip the headers as stealth1.x does, the fact that the RSA encrypted header is always < N, the rsa modulus gives the game away with a few messages to analyse for statistical purposes, ie if you see a lot messages < N, where N is by definition not an even power of 2, and possibly even N is known, or suspected from keyservers). I've been delaying releasing it because I was worrying about ran no generators being good enough, but I think I might have convinced myself that it is not even necessary to use ran no generators, I'd appreciate it if some folks with a bit of crypto expertise could cast their eye over the description below. > > So what, -that "only a few companies" will be discovered to be using PGP > > through the RSA signature!? Those few companies are the seeds for the > > vast numbers of companies that would follow them in using PGP over the > > Internet. The RSA signature is the flag that allows the spooks to easily > > net the bold first companies. The RSA signature is greatly impeding the > > spread of PGP use over the Internet. PGP MUST BE STEALTHED!! > > It isn't an RSA signature. Its a bunch of magic numbers. > > Look, get real already. If someone sees a bunch of random numbers in > mail sent by me, its going to be pretty obvious what the hell is > inside anyway. Yeah, but you wouldn't do it that way. I reckon this is what you'd do: % pgp -es duress % pgp -es msg % stealth < msg.pgp > msg.stl % cat msg.stl >> duress.pgp % pgp +makerandom=1234 noise % cat noise >> duress.pgp % pgp -a duress.pgp % mail someone < duress.asc The pgp +makerandom=<size> <file> is an undocumented feature of pgp > 2.6 (not sure exactly when it got added, Colin Plumb pointed it out when I asked him about ran nos for stealth). So what this means is that you are using PGP it's self to hide a stegoed message. This would be good for the guy from FACTnet (forgotten his name) who just got hit by the CoS, he could hold out for a while, then give up his key, the duress message would appear, and the real message would be explained by having a script to do this on his HD, and having long since burned the disk with a script to do the above on it: % pgp -es msg % pgp +makerandom=4567 noise % cat noise >> msg.pgp % pgp -a msg.pgp % mail someone < msg.asc ie the idea is that you pad your message to a fixed size for the express purpose of hampering traffic analysis (of the type of my, Alice did have a lot to say to Bob that day). It would be even better cover if the thing had gotten sent through a remailer, as this kind of thing is expected of type I remailer traffic (before mixmaster which does the packetizing for you). So the duress message really looks like this: +---------+---------------------------+--------------------+--------+ | pgp hdr | IDEA encrypted duress msg | stealthed real msg | noise | +---------+---------------------------+--------------------+--------+ the IDEA block has a length field, but you can increase the length without damage to include the following stealthed stuff as the underlying stuff which is IDEA encrypted will know it's length on decryption, and the following junk will just be discarded. So, Alice and her secret key ring (encrypted) gets nabbed by the Charlie (CoS?), and coerced into divuling her passphrase. And if and when it is noticed that the message was longer than it ought to be (CoS that smart? substitute the NSA and they'd notice for sure), Alice explains away the junk on the end by pointing them to the fact that all of her messages where exactly (say) 16k long, and that she was using a the noise only script, and that the message really is this: +---------+---------------------------+-----------------------------+ | pgp hdr | IDEA encrypted duress msg | noise | +---------+---------------------------+-----------------------------+ Now we come to arguments about why you might want this built-in to PGP. Well it provides plausible deniablity as you have no extra software which might look incriminating unless you managed to dispose of it first, if it comes as stock. Also the 2nd reason for built-in at least for stealth is if it needs good random number source, but as I said, I'm not sure it needs a random no generator. So comments please, this has been around a few times already, but here's the algorithm for manipulating x which is the RSA encrypted component of the header of an RSA+IDEA encrytped PGP message: (this is a description of my implementation of Hal's algorithm as described on his www page: http://www.portal.com/~hfinney/) consider random no x, RSA modulus N 1 < x < N (that used to say 0 <= x < N, but 0, and 1 being RSA fixed points, 0, 1, and N won't be generated by PGP presumably? I think this shouldn't matter as the keyspace of N is so large that the probability of a 0 or a 1 specifically is nothing to worry about I think.) Hal's algorithm was to convert x to being in the range: 0 <= x' < lim where lim which is the next power of 2 above N * 2^surety, and surety is 64, 64 seems big enough? The recover operation is: x = x' mod N and the create operation is: 1) scale = int( (lim-1) / N ) + 1 2) scale2 = 2^int( log2( scale ) + 1) ie scale2 = next power of two over scale 3) r = MD5( 0, x ) ie MD5 digest of x, as x is an RSA encrypted random session key misc other info, and more random padding to bring up to be close to N in size. It strikes me that we already have a random number, and that provided MD5 can not be inverted (which it is not possible, as it is compressing, and looses info, and the brute force to find which of all the possible y's (0 < y < N) MD5 digests come out to be x. I would have thought it likely that this would be evenly distributed, and that the cost would be enormous? 4) r = MD5( r, rand() ) it will fail some of the time, so in this case repeat with another random number, Istirred it in first but perhaps you would only do it if it failed. Or perhaps it is enough to MD5( r, r ) ie stir r into itself to generate another ran no? So long as there are no rare cycles? Presumably either impossible or infinitessimally small probability. 5) r = r mod scale2 6) if (r > scale) goto 4) 4),5) & 6) are designed to generate an evenly distributed random number in the range: 0 <= r <= scale 7) x' = N * r + x 8) if (x' > lim) goto 4) an example, with small numbers, and surety set to 8, x = 7, and manufactoring the ran nos manually: lim = 2^( int( log2(N) + 1 ) + surety ) = 2^( 4 + 8 ) = 4096 1) scale = int( 4096 / 13 ) + 1 = 316 2) scale2 = 2^( int( log2( scale ) ) +1 ) = 2^(8 + 1) = 512 3) say MD5( x ) = 0x00000000000000000000000000001234 4) say MD5( x, rand ) = 0x00000000000000000000000000001235 = 4661 (base 10) 5) r = 4661 mod scale2 = 4661 % 512 = 53 6) if ( 53 > 315 ) - it's not 7) x' = N * r + x = 13 * 53 + 7 = 696 8) if (696 > 4096) - it's not) so x' = 696 reversing: x = x' mod N = 696 mod 13 = 7 when 6) is true, ie r > scale which will happen half the time on avg with the first ran no, and so on like a coin flip, the chances of many coins flipping the wrong way diminishes rapidly with number of flips. when 8) is true, ie x' > lim, x' can only be > than lim when x = N - 1, and given that x = N - 1, the probability is less of x being in N * scale <= x < lim but this will only occur with probability: 1 1 -------- < ----- 2^surety scale I think. So you'd have to collect one mess of messages to even catch one sample, let alone enough messages to have a statistical proof of steganography being present. Pls check the maths a bit, the above describes the software implementation of Hal's algorithm which he has on-line, on his page: http://www.portal.com/~hfinney/ But the main question I'd like to get verification on is if it is safe to use the MD5 of the RSA encrypted message to perform the operation. I'm essentially doing: x' = N * f(MD5(x)) + x where f(y) is a function which converts from range 0 <= y < 2^128 to a range 0 <= f(y) < scale. Is that safe? x is random, and will be different even for a repeat encryption of the same file, as PGP is using a random IDEA session key. So are there any brute force attacks on that which would be cheaper than attacking 128 bit IDEA? PGP's random number generator also makes extensive use of MD5, so I'm taking the use of MD5 as secure as a given. If it is thought to be dangerous for some reason (it is after all some kind of signature on it's self, presume that you know N, and x' but not x, the question is can the equation be brute force reversed in a less than 128 bit brute force attack. I'm neglecting to consider the rand() calls, which I'm not expecting to add security, but are just a mechanism to stir the value with to get more random nos, as occasionally the alogrithm needs more than one, if the first fails, etc. If people reckon it's insecure, then it would be ideal to include the stealth functionality into PGP, so that integral use of PGPs ran no routines can be made. I was previously using an MD5 digest of PGPs randseed.bin for this, but you can't stir it (well you could but that would considered a security risk diddling with PGPs files), and not stirring means you have no improvement over the above, in the event that that your system is captured with the known plaintext. If you stirred it there would be no proof of that being the message in the file. Even unstirred, if the N * f(MD5(x)) + x is no good, the inclusion of a digest of the randseed.bin would be a big improvement. It seems rather messy to have all of the keyboard sampling stuff, for PGP keys duplicated for this. Hope the x' = N * f(MD5(x)) + x construct is secure as this will avoid the issue. Adam

1 0

Re: "Got a subpoena?"
by Jeff Hupp 17 Dec '03

17 Dec '03

On 30 Nov 95 at 12:56, Jay Campbell wrote: : >> "finessing" laws will become more important. In short, they WILL have : >> a subpoena--then what? : > : > Then they'll find out that I don't have any information that : >could help them, anyway. : : What about a court order to (a) start comprehensive logging, and (b) not : tell anyone under penalty of ______ . And at that point, you shut down the remailer and log each and every bounce. Now, they could get an order forcing you to run a remailer ~ but that would open up a whole new can of worms i.e. can you be forced, under court order, to commit a crime? -- JHupp(a)gensys.com |For PGP Public Key: http://gensys.com |finger jhupp(a)gensys.com Autocracy is based on the assumption that one man is wiser than a million men. How's that again? I missed something.

2 1

Re: Digital Fingerprinting
by Peter Trei 17 Dec '03

17 Dec '03

> Finding the mail mixed in there is evidence that supports the "charge" > but surely can be rebutted. When I was practicing in a small town c. > 1984, people would be summoned (sent a paper telling them to show up) to > court for illegal dumping if their mail was found with other trash, old > refrigerators, etc., in a creek somewhere. Those charged could always > put on proof that Billy Bob's garbage service picked up their trash ... > The sentence was almost always to go and pick up some multiple of the > garbage found and take it to the landfill. This was Stockbridge, in MA, right :-? That's what we did, and drove back to the church, had a thanksgiving dinner that couldn't be beat, went to sleep and didn't get up until the next morning, when we got a phone call from officer Obie. Said "Kid, we found your name on an envelope at the bottom of a half a ton of garbage, and just wanted to know if you had any information bout it". I said "Yes Sir Officer Obie, I cannot tell a lie, I put that envelope under that garbage." (c) Arlo Guthrie Peter Trei Senior Software Engineer Purveyor Development Team Process Software Corporation http://www.process.com trei(a)process.com

3 3

Re: Questions/Comments on ecash protocol
by shamrock＠netcom.com 17 Dec '03

17 Dec '03

At 22:40 12/2/95, Michael Froomkin wrote: >thank you for the sterling analysis. I for one am following this with >enormous interest, even though some of the details are lost on me right now. [Well done, Ian!] [...] >1) What information about Charlie/customer is encoded onto the coin? None. >(There must be some, right, since the serial number is blinded?) Since the >bank doesn't know what serial number it is signing, it needs to put info >about Charlie onto the coin so that it can track him down if he double >spends. Lacking such info, the bank can refuse to honor a double-spent >coin, but has no way to know who the double-spender is. Since an online clearing protocol is being used, the bank has no need to identify double spenders. The bank will simply refuse to honor a double spent coin. In fact, cancelling a payment in this protocol is done by just depositing the coin yourself. >3) Is there a way [how hard is it] for charlie to extract a coin and >either > (i) copy it and/or > > (ii) send it to David [3rd party] in such a way >that David could insert it into David's MTB software and then spend it to >Sam without Sam or the Bank noticing that anything was wrong. If Charlie >and David do this, David now has a coin that is from his point of view >both payee and payor anonymous, although Charlie has a risk that David >will double-spend and expose Charlie to the bank's wrath. I can't help the feeling that I am missing something whenever you bring up this question. Assuming it could be done. What would David gain? He as the payor is anonymous to Sam either way. Sam still would have to be worried about being identified, since if Charlie gives David access to Charlie's wallet, it is safe to assume that Charlie will give David (and the mint) access to his blinding factor. Which in turn would reveal Sam as the payee. The protocol you suggest gives the parties exactly what they would have if they just used Ecash "out of the box": full payor anonymity, no payee anonymity. So why bother? <insert standard disclaimer here> -- Lucky Green <mailto:shamrock@netcom.com> PGP encrypted mail preferred.

2 1

Zensoren ueber Alles
by tallpaul＠pipeline.com 17 Dec '03

17 Dec '03

>From initial reports, the challenge to CompuServe did not come from the German government. It did not come form the German Parliament. It did not come form the German judiciary. It came from one (repeat one) prosecutor in one (repeat one) city. I also suspect the U.S. journalists who reported the news may not know the difference between a "prosecutor" and a "procurator." It is a significant differnce in many legal systems. (Not to slag U.S. journalists too hard because I myself don't know which of the two Germany has.) The news reports on New York City television focused exclusively on the "sexual" character of the banning, without directly mentioning that other groups were also banned. "But," wrote Jonathan Oatis of Reuter in an article available on the Clarinet news feed, "many more span a host of topics, including Barney the dinosaur of children's television, Estonian politics and the New York Yankees baseball team." Also included in the reports of news groups banned is the clari.news.sex feed containing articles routinely filed by the Associated Press and Reuter. It has also been reported that self-help recovery groups were banned because the groups also had the word "sex" in the name. (In order to save the victims it was necessary to destroy them!) How did the 11 o'clock news here in New York handle the story? Two of the three channels stated that CompuServe banned "chat rooms," not news groups. The third got it right and reported on internet groups. The two channels who spoke of the "chat room" ban also stated that the rooms were banned because they contained "explicit" sexual images or graphics "depicting" sexual topics. What can we infer? No single thing directly save for the inaccuracy of the reports. But certainly even national-level tv editors are not yet sufficiently informed about the internet to know the difference between chat rooms and news groups. I'd say that this level of ignorance rather impacts on their ability to do their job in a professional fashion. (It also points to our collective failure to adequately inform them.) Second, the editors -- whether under pressure of deadline, personal psychological bias, or more sinister things -- can't pick up the idea that the verbal ("chat") and the visual ("explicit" sexual images) are two different things. Third, the sexual hysteria of the editors themselves significantly erodes their ability to perform their jobs in an objective fashion. CompuServe also got off easy under initial press inquiry. One spokesman for the company announced they were required to do it. He also stated that there was no way to cut the German customers off from the groups will making the groups available to other CompuServe customers. Based on other news reports, I conclude that CompuServe lied in both areas. (BTW, this is the first time I recall using the word "lie" on any post to the cypherpunks list.) I also infer that CompuServe did not "roll over" on this issue. The evidence shows, I think, that CompuServe is merely using one German prosecutor (or procurator) as an excuse to implement their own desired and previously prepared policy. CompuServe had, I think, several actions open. First, if the news reports that it was not "forced" to do anything by the single German, it could simply not have done anything. Second, it could have appealed the decision by the prosecutor to the courts (or submitted accurate information to the procurator and demanded that he consider it.) Third, it could have narrowly targeted the banned groups to alt.binary groups dealing with sexual issues. Fourth, it could have easily used software to cut off the feed to Germany. It did none of these things. It cut off all customers to an enormous number of groups. It inferrentially violated property rights (i.e. contracts) to customers promised internet access and now provided only a crippled version thereof. And it lied about the whole thing. Interestingly, none of the classic cypher-nasties were behind CompuServe's decision. The "big statists" in Washington didn't tell CompuServe to do it. The "hell with private property rights" bureaucrats didn't force CompuServe to do it. Nor did the taxman. The taxmen historically rarely do; they do not seek to ban "sin;" they tax it. The Treasury Department's Bureau of Booze, Butts & Bazookas (aka Bureau of Alcohol, Tobacco, and Firearms) is not behind this country's anti-booze, anti-butt, or anti-bazooka movement. They just tax all three. (They may kill you if you buy, transfer, or manufacture your bazooka without paying the US$ 200 (?) excise tax, but they're not out there in the forefront of those pushing gun control.) Additional facts that will be forthcoming in the future will point, I believe, to two things behind CompuServe's decision. The two leading causes will, I predict, be: First, the growing abstract systemic fear in this society produced by a society in crisis. This is a fear unnaturally re-directed at things like PGP and anonymity by various political poo-bahs to both deflect the citizens' fears from real causes and to rechannel that fear into areas where the same poo-bahs can claim credit for doing the "something" in "something has to be done." Second, the growing sexual hysteria within large sections of the population that does not exist in an abstract form and is not being artifically rechanneled but rather appealled to. CompuServe, in a rather brilliant move, managed to handle both groups, and blame a foreign force to boot. But while brilliant tactically, I do not believe they will succeed in continuing their policy. --tallpaul

2 1

Re: Guerilla Internet Service Providers (fwd)
by Jim Choate 17 Dec '03

17 Dec '03

Forwarded message: > From owner-cypherpunks(a)toad.com Mon Jan 1 17:53:37 1996 > Message-Id: <199601012332.PAA10533(a)desiree.teleport.com> > Subject: Re: Guerilla Internet Service Providers > To: cypherpunks(a)toad.com > In-Reply-To: Your message of Mon, 01 Jan 1996 13:41:13 PST. > <v02120d09ad0e047046b5(a)[192.0.2.1]> > Date: Mon, 01 Jan 1996 15:32:37 -0800 > From: Felix Lee <flee(a)teleport.com> > Sender: owner-cypherpunks(a)toad.com > Precedence: bulk > > Lucky Green: > > Some site in physical space has to host the nntpd, the ftpd, and the httpd. > > That site will be subject to search, seizure, and arrest and conviction of > > owner. > > but if it turns out that 30% of home PCs have to be seized to prevent > dissemination of dangerous-information-X? > > though we're not quite there yet... eg, it's a little expensive to run > your own httpd 24 hours/day. > -- > If you live in the Austin, TX area I will setup a dedicated slip for you at only $100/mo.

3 2

Payment Systems
by rah＠shipwright.com 17 Dec '03

17 Dec '03

--- begin forwarded text From: "John Hemming CEO MarketNet" <JohnHemming(a)mkn.co.uk> Date: Tue, 12 Sep 1995 15:38:00 PM PDT To: www-buyinfo(a)allegra.att.com Mime-Version: 1.0 Subject: Payment Systems For the edification of those concerned we have issued our first Electronic Cheque today. The browser now has been tested linking signed instructions directly to the web. The browser, however, is still being developed to resolve the problems with shortage of space in Windoze DGROUP. There is a copy of the browser at ftp://193.119.26.70/mktnet/pub/echeque.zip (or something similar) we would recommend waiting a while before you take it however and to take the horse.zip as well because of the DGROUP issue). The only form that will accept Echeques at the moment is http://193.118.187.107/load ordflow The trick in this form (if you are using our browser) is that it allows payment either by ECheque or by credit card and only generates the signed instruction for echeques. The signed instruction is then sent to our servers encrypted with 128 bit RC4. ttfn John --- end forwarded text ----------------- Robert Hettinga (rah(a)shipwright.com) Shipwright Development Corporation, 44 Farquhar Street, Boston, MA 02131 USA (617) 323-7923 "Reality is not optional." --Thomas Sowell >>>>Phree Phil: Email: zldf(a)clark.net http://www.netresponse.com/zldf <<<<<

1 0

NOISE.SYS /dev/random driver for DOS, v0.3.3-Beta
by wlkngowl＠unix.asb.com 17 Dec '03

17 Dec '03

The leatest version is ready to be examined. Much rewriting. Send a message with the subject "send noise033" and an automated responder should eventually send you a uu-encoded Info-Zip file with a compiled version of the driver, assembler source-code, and pgp-sig. Again, comments & criticism would be appreciated. Help is especially needed trying to find IRQs to sample for sources of randomness. Finding a good way to sample mouse or video scan lines would also be a help.... etc. --Rob

2 1

PGP duress code, and stego (was Re: PGP in UK - snooped unSTEALHed?)
by aba＠dcs.exeter.ac.uk 17 Dec '03

17 Dec '03

erm darn, I posted a huge post which should have the above subject, but instead came out with the subject line: Re: Returned mail: User unknown due to typo, could someone replying please correct that, if following up, I won't bother posting it again in respect of those who actually pay for their mail feed per k. It had a lot of questions which I'd really like people to look at, skip back a few, and take a read, it's for the implementation of stealth2, Henry Hasturs stealth util. Adam

1 0

NSA says strong crypto to china??
by Jeff Weinstein 17 Dec '03

17 Dec '03

This is a quote from an article by Paul Vallely in The Independent, London, which can be found here: http://nytsyn.com/live/News3/006_010696_101827_2723.html > What one government regards as harmful material is an instrument of freedom > and democracy to another. Officials at the US National Security Agency have > suggested that Internet encryption technology - a sophisticated method of > encoding information - be deliberately exported to Chinese dissidents to > help them in their fight against their government - even though its export > is otherwise banned under US arms control regulations. Does anyone know of real documentation of this "suggestion" from the NSA? It quite telling, though no surprise to any of us I'm sure, that they would think that strong crypto should be a tool of freedom in china, but not in this country. Maybe they should get Microsoft to insert subliminal directions for downloading PGP into a future episode of "My Computer Family". :-) --Jeff Jeff Weinstein - Electronic Munitions Specialist Netscape Communication Corporation jsw(a)netscape.com - http://home.netscape.com/people/jsw/ Any opinions expressed above are mine.

7 6