December 2003 - cypherpunks-legacy

Re: Voice/Fax Checks
by Hal 17 Dec '03

17 Dec '03

2 1

Raytheon got a great deal
by rjc＠gnu.ai.mit.edu 17 Dec '03

17 Dec '03

Wow, an 800 billion dollar contract! Someone put me in contact with the government of Brazil! ;-)

1 0

Re: Voice/Fax Checks
by Hal 17 Dec '03

17 Dec '03

1 0

Re: "Key Escrow" --- the very idea
by wcs＠anchor.ho.att.com 17 Dec '03

17 Dec '03

Carl Ellison, who should know better (:-), writes: > if you really want to propose an escrow system we can live with, > I would demand that it include: Arrrgh! I'll try not to flame much here, but this is *wrong*! Please don't buy in to the government's claims of legitimacy for this nonsense merely because they've backed down on the less practical implementations of their abusinve proposal! First of all, _escrow_ systems are something that two contracting parties use when they want a trusted third party to perform a service for them. Holding deposits in real estate transactions is a classic example. Built-in wiretaps are *not* escrow, unless the government is a party to your contract. As somebody on the list once said, just because the Mafia call themselves "businessmen" doesn't make them legitimate; calling extorted wiretaps "escrow" doesn't make them a service. The government has no business making me get their permission to talk to anybody about anything in any language I choose, and they have no business insisting I buy "communication protection service" from some of their friends to do it, any more than the aforenamed "businessmen" have any business insisting I buy "fire insurance" from *them*. If you want to talk about escrow systems, the proper contexts are things like contract fulfillment between anonymous parties... Meanwhile, back to conditions for built-in phone wiretap systems: > 1. unambiguous ID of the person being tapped in the LEAF-equivalent No! I agree that having the government prove your phone was used for a given conversation that you weren't part of is bad, but the only way to have unambiguous ID for wiretappees is to have unambiguous ID for everybody - I certainly don't want to have to insert my National Real American ID Card into a phone to make a call, or into my computer to send email, and in case people start noticing that they can't make phone calls after their wallets are stolen, I don't want to have to wave my arm-tattoo over the scanner either. (Ok, I said I'd *try* not to flame :-) > 2. multiple escrow agencies, at least one of which is the NSA HQ > (for its superior physical security) They're the *last* people I want involved with routine communications between ordinary people. They're an agency that should probably be abolished, but at most they should stick to providing secure communications for the military; I don't want military police agencies or even Federal civilian police agencies getting involved in civil commerce, (especially when they're doing it to find new businesses now that we don't have Commies to kick around any more.) I shouldn't need *anybody's* permission to have a private conversation with anybody, but least of all a secret organization that classifies their activities rather than working out in the open. > 3. watchdogs as escrow agents (e.g., ACLU, Rep & Dem parties, CPSR, > EFF, NYTimes, ...) with authorization to look for abuses of > authority and to refuse to release keys in such cases and to > publicize such cases as well as bringing them to the attention > of law enforcement for prosecution. Realistically, if the government starts allowing non-government agencies as keymasters, it'll probably be banks or phone companies, since they're large cooperative subpoenable organizations that are involved in the communications the government most cares about wiretapping, and they're hard to avoid since they're providing your services. In particular, it'll help set precedents. Bad ones. I'd also worry about the effects on a watchdog group of taking government money for helping the government wiretap people. Wiretap keymastering is likely to be an expensive activity, if done competently, and involves major questions of liabilty. What happens when the government says to your group that they'll cut your funding by $1Million if you don't keep this one quiet? Even if you're honest enough for that not to work, what about the moral effects of being on the government's side in a court case (as keymaster) when you used to be the group that defended the Steve Jacksons and Craig Neidorfs? > 4. user-generated escrow keys, to reduce the chance of anyone having a > backdoor way to get the whole escrow key database. That's a minor technical detail :-) It's also quite possible, and the all-software wiretap version that Dorothy Denning and friends are talking about supports it just fine. A more important detail would be to use genuinely separate master keys instead of one master key split into multiple parts for the keymasters by the trusted NSA, as in the current Clipper system. Bill # Bill Stewart AT&T Global Information Solutions, aka NCR Corp # 6870 Koll Center Parkway, Pleasanton CA, 94566 Phone 1-510-484-6204 fax-6399 # email bill.stewart(a)pleasantonca.ncr.com billstewart(a)attmail.com # ViaCrypt PGP Key IDs 384/C2AFCD 1024/9D6465

2 1

Re: "Key Escrow" --- the very idea
by wcs＠anchor.ho.att.com 17 Dec '03

17 Dec '03

Keywords: ranting, crime in cypherspace X-Oops: let's try including the contents file this time:-) Mike_Spreitzer.PARC(a)xerox.com writes > (1) I'm not an anarchist. Does that make me out of place here? No problem, as long as you can handle surprising rhetoric on occasion :-) > I'm willing to live with some amount of government, I don't see that there's much choice, and after we get rid of the first 90% of the government I'll be happy to debate theory for the other 10% :-) > My biggest problem with Capstone is that it > changes the balance of power too much. It's more than a change - it's a declaration that the government already had the *right* to control all your communications, and is just now getting around to implementing it. But aside from the arrogance, I'll agree that it's too much power for them. By the way, you used the term "social contract" in your letter. Somehow, the government has gotten the idea that the "social contract" is between them and us, rather than between us and us. It's not, or at least none of the copies *I* signed included them... > (2) I think crimes can be committed in cyberspace. Substantially, if not > entirely, in cyberspace. Maybe not so many now. I have to agree, and I distinguish between "real crimes" vs. "laws". a) Untraceable payments for physical violent crimes (e.g. kidnap ransom) b) Better communications for conspiring to do violence (murder contracts...) c) Bank Robbery (any respectable digibank can protect itself technically, but we're already seeing Teller Machine card forging in Britain, and other banks will probably have weaknesses as we learn digibanking.) d) Forgery - digital signatures are great, if they're long enough, but protecting your keys is more critical than it used to be. e) Fraud - you'll probably have to do a better job checking reputations for a digital stockbroker living behind anonymous remailers paid with digicash than you currently do for physically traceable brokers like Ivan Boesky. f) Extortion - it's hard to break somebody's legs in cyberspace, but you can send the threat that way, and tell where to send the money; you can also threaten to publish their private key which you stole. Of course, the big "crimes" that the government usually wants to use wiretapping for are things like drugs and money laundering, both of which are none of their business. > I accept the terms of the 4th amendment: search and seizure allowed > when due process followed. The 4th amendment's terms aren't for you - they're for the government to obey. While I suspect the authors of the amendment assumed the government would seize criminals and search for them, they don't claim that power as their right, they only place limits on it. > "Key escrow" is an attempt to implement the cyberspatial analog of search. No, it's not. Wiretapping, electromagnetic eavesdropping, and demands for records you were already keeping are search. Ordering you not to have private conversations without recording them for the government and not to have locks without giving them the keys first are the analogs of so-called "key escrow". > to be worth it. Note that's a comparison of their money and success rate > against our privacy; no wonder they got it so wrong. Well said... > > (4) If you accept points (1) and (2) above, you're left wanting a way to > implement searches in cyberspace when due process is followed. > I hope anarchists won't be the only people opposing changing the > balance of power greatly in the government's favor As a moderate not-quite-pacifist anarchist, I still understand people's desire to protect themselves and their property, though I'm not sure that I agree that revenge after the fact has a real moral justification, but if it does, then you'd probably want to hire some police to get your stolen stuff back or avenge injuries done to you, or at least detectives to find out who injured you so you can publish bad reputations about them. The government aren't always *my* police force of choice, but I certainly have no intention of imposing my spy service on all your conversations. > (by poorly designed key escrow). What are the rest of us left to answer with? > Perhaps a much better key escrow design. "Escrow" is an arrangement between two parties to hire a trusted third party to keep something for them, typically down payments in contracts. If you want to escrow keys in conversations between the two of us, feel free. If the people who work for the government think that *they* are one of the parties to my conversations with other people, when I wasn't talking to them, they're rude and arrogant :-) If they think they *own* my conversations and can limit them, it's time to see how the Bill of Rights limitations on "takings" apply in cyberspace.... > One that integrates the search with the due process in a cryptographically > strong way; one that can't be subverted by a few people in a few organizations. > For example, who says an escrowed key must have only two parts? The Clipper chip only has one master key per chip; the fact that they store it in multiple pieces is a political charade designed to increase its chance of acceptance by focussing on the details. It certainly wouldn't have been hard to design a chip that really *did* have two separate master keys input by separate agencies. Or more. > And again, remember where we're weighing money against freedom. > It may be that we just have to spend more to stay a reasonably free society. > Also, it's worth debating just how strong the protections have to be. Money is part of the issue; the more important part is weighing restrictions on people's freedom against the benefits of order. The government has essentially announced that *they* get to do the weighing and deciding. And the technical issues are all classified, thank you :-) But you can trust the NSA; they're competent professionals. There really *are* benefits to order, and there are real crimes that may be less likely to happen if order is imposed on us. Freedom has risks. I think they're worth it. And unlike the folks who've decided they're in charge of order, I think it's wrong to make that decision for others, at the cost of their freedom, which mandatory escrow does. Bill # Bill Stewart AT&T Global Information Solutions, aka NCR Corp # 6870 Koll Center Parkway, Pleasanton CA, 94566 Phone 1-510-484-6204 fax-6399 # email bill.stewart(a)pleasantonca.ncr.com billstewart(a)attmail.com # ViaCrypt PGP Key IDs 384/C2AFCD 1024/9D6465

2 1

Re: wanted SueDNym messages
by Peter Murphy 17 Dec '03

17 Dec '03

But what's this MEDUSA's TENTACLES program meant to be? An artificial intelligence modelled on the personality of LD? <yeegawdz!> Peter.

2 1

Re: Double DES calculations
by norm＠netcom.com 17 Dec '03

17 Dec '03

At 09:05 1994/07/22 -0700, Hal wrote: >I missed the start of this double-des thread due to system problems and >being gone, and I've never been able to pick up the main point since. It >sounds like some kind of meet-in-the-middle attack is being discussed. >It is true that with current technology MITM generally seems more costly >in terms of space than time. However, I have seen references to techniques >which shift this tradeoff some, costing more time and less space. Un- >fortunately, I can't remember where I saw them! ... There may be more than one way that MITM (meet in the middle) may be used to attack Double block cyphers. I assume the following attack. You know some block of plain-text P and corresponding cypher text C. You believe that C = E(k, E(j, P)) where E(k, p) is the encypherment of p with key k. D(k, E(k, p)) = p. You need to find keys k and j. Classic MITM is to produce a file A with records: <k, E(k, P)> for each k, and file B with records <j, D(j, C)> for each j. Sort both A and B on the second field. Pass over the sorted files looking for a record from file A whose second field is the same as a record in file B. To substantially shorten the ammount of tape used by a factor 2^n at the expense of evaluating C and D 2^n more often do the following: For m from 0 to 2^n-1 Do Produce file A with records: <k, E(k, P)> for each k where (the right n bits of E(k, P)) = m. (discarding other records) Produce file B with records <j, D(j, C)> for each j where (the right n bits of D(j, C)) = m Sort files A and B on second field. Pass over files looking for records from A that match records from b in the second field. Enddo. This is still a daunting job and evaluating its magnitide requires several assumptions. The most obvious is the cost of evaluating C and D. Next is the cost of reading and writing tape.

1 0

Re: Voice/Fax Checks
by Pierre Uszynski 17 Dec '03

17 Dec '03

A couple of pointers on current outfits trying to undercut the "transaction cost", none of them the ultimate we all root for, but nonetheless. hfinney(a)shell.portal.com writes > Robert Hettinga writes: > > >I've gotten stuck on > >exactly how to "*undercut*" the transaction costs of existing methods. Got > >any ideas? > > It's pretty clear that credit cards don't work for some of the transactions > people want to do: > [...] > It's not that easy for a mail-order shoestring startup to get the ability to > accept VISA cards. Because of the danger of fraud, the credit card companies > like to see a storefront and/or some previous history. Someone who writes a > nifty PGP shell and wants to sell it for $10 per will have this problem. It's actually getting easier. Small card service outfits have noticed the problem. They have looked at the danger of fraud, and when accepting a new company, still conduct a sufficient investigation (so they claim :-). They have also looked at the way small outfits operate. The result is that it seems it is now possible for "home businesses" like BBS's or software sellers to get cheaper service than store fronts. It comes out to something like $25 a month + 2.5-3% per charge ($0.20 minimum per charge, $25 minimum per month) + equipment, roughly, for VISA and Mastercard. As an example, in the San Francisco area, try ... errr... <grumble>... (Darn it, I buried that guy's card... try the phone book :-) > 3) People who don't like giving out their credit card numbers to an unknown > email address. Or to a small random unknown business... Which is a problem even if everybody could accept VISA, as you mentioned. Others already mentioned on the list the 1-900 phone based "netcash" <shrug> service described in July 1994 Boardwatch Magazine. You call the 900 number with your modem, you get a random string for it, and your phone company bills you $10 (info at netbank-info(a)agents.com) You then can get change (down to $0.25 strings), do on-line validation of transactions, and eventually redeem the strings, minus a 20% redemption fee. In addition to that one service, I already heard of two others that cater in particular to BBS's, similar system where you get billed $10 for a 900 call, and get a random string to use as a voucher. These two don't let you "get change" or any similar facility, but they let small businesses get rid of the "collection" process, for a fee. In all these cases, the fee is around 20-25%, so there may be space for competition (I don't have the coordinates for these, just know where to ask). That was/is actually the strong point of the French Minitel system (as much as I can't stand that piece of plastic ****): Minitel services are accessed (mostly) through 900 style numbers, and France Telecom handles all the billing and collection for the Minitel service providers, allowing charges as low as 7 cents per call to a service (most are much higher, WAY higher). We already discussed the "phone check" idea, even Western Union advertised it, and maybe even provides it :-) Clearly all these systems have disadvantages, and are some way from what some of us want, but they show that the current "billing" services are not standing still, and are working on this access barrier to credit card payment. Clearly too, they don't seem to be working in the direction we want. Pierre. pierre(a)shell.portal.com

2 1

Re: Voice/Fax Checks
by Hal 17 Dec '03

17 Dec '03

Eric Hughes writes: >You can still use an account mechanism, but with an intermediary whose >business it is to aggregate small amounts as these proposed and clear >the total periodically. That's now one account setup for the >customer. How, though, would the ftp site which wants to know whether I'm "good for" the one cent charge to download PGP do so? Does it have to check with an agent on the net somewhere which will vouch for me? Aren't the communica- tion costs then the same as an online system? Or does it extend me the one cent as credit and hope that I really do have an account with that agent (or bank)? Then that seems like a basic off-line system. So I don't understand the role of agents in solving this problem. I find it confusing to imagine a situation where large numbers of goods are sold for very low prices. Will people tend to cheat, since it's easy to get away with it (all those systems offering you one cent credits), or will they tend to be honest, since the per-use cost is so low (but perhaps adds up over a month)? I suspect that nobody will pay if there is a way they can use the servers without paying, even though they are only saving a fraction of a cent each time. Maybe that's just my jaundiced view of human nature. Hal

2 1

Re: CYPHERPUNKS TO THE RESCUE
by norm＠netcom.com 17 Dec '03

17 Dec '03

At 09:51 1994/07/26 -0400, Russell Nelson wrote: >Why not generate a random number, checksum it, and sign it using a >public key? Or is that overkill? ... Seems good. But to thwart replay of the signed message the garage unit must never accept the same signed number twice. How about the car unit signing successive numbers. The garage unit would remember the last number that it accepted and only accept signed numbers larger than that. Garbled transmissions would then cause no problems. They would be fixed by yet new transmissions, just as with current units.

1 0