December 2003 - cypherpunks-legacy

Re: The Terrorists are coming!
by jdd＠aiki.demon.co.uk 17 Dec '03

17 Dec '03

In message <9408031132.ZM695(a)simple.sydney.sgi.com> Ian Farquhar writes: > Out of curiousity, is anyone aware of whether the USSR employed PAL's > (Permissive Activation Links) in their strategic nuclear weaponary? If so, > is anyone aware of how secure the PAL's the Soviets actually used were? > There was a rumor on USENET some time back that the Soviets were using RSA > in their PAL's, but it sounded too much like an urban myth to me. According to recent press reports in the UK, when Soviet bombers were loaded with nuclear weapons, they were actually dummies (because the generals, etc, didn't trust the crews). This suggests that the Soviets did not have reliable safeguards on the bombs. Various arguments between the Ukraine and the USSR about ICBM warheads, also reported in the press, suggest that the safeguards on the warheads were reliable, because both parties seemed to agree that the Ukrainians couldn't set them off, although they are otherwise technically competent in dealing with the missiles. -- Jim Dixon

1 0

Re: EDDB/RN
by jdd＠aiki.demon.co.uk 17 Dec '03

17 Dec '03

In message <9408091536.AA22362(a)ah.com> Eric Hughes writes: > However, there should be a use for persistent store, for a remote > encrypted database accessible anonymously. > > The real questions are "how big is the market?" and "how much revenue > is there in it?". > > Something like this doesn't get made reliable by volunteers. I think that the market is potentially large. > Ideally, the data is stored on a distributed data base, with some > redundancy in case one or more gateways go down > > Look in Schneier for secret sharing. We won't have a copy of Schneier here for three weeks or so. Can you elaborate? However, almost certainly, the crypto aspects would be relatively minor. Most of the work would be in building a reliable widely distributed data base accessible over the Internet. When I last looked at this sort of problem, it was very complex. -- Jim Dixon

1 0

Re: <null>
by jdd＠aiki.demon.co.uk 17 Dec '03

17 Dec '03

In message <9408070032.AA17321(a)ah.com> Eric Hughes writes: > One assumption here is that someone in one country can easily pay > someone in another country, and an automatic currency conversion can > take place. The prerequisites to happen generally for that are the > electronification of retail money in both jurisdictions and a > retail-level currency exchange system. None of this really exists > yet, although the first beginnings are here. Also, for anonymous > payment for such overseas services, anonymous transfer in at least one > of the two currencies is necessary. The last point is certainly not true. If user X communicates with service A (a gateway) in one country to purchase something from service B in another country, X can settle accounts with A anonymously (say in US$) and then A and B can settle accounts with one another (say in sterling) openly. In fact, few Europeans would object to simply trading in digital money denomenated in US dollars. No currency conversion is necessary. Many European companies have US dollar bank accounts because the banks charge so much money for conversion of currencies. -- Jim Dixon

1 0

Re: RemailerNet v0.2
by jdd＠aiki.demon.co.uk 17 Dec '03

17 Dec '03

In message <199408090347.UAA24150(a)jobe.shell.portal.com> Hal writes: > What is the goal of the RN as far as defeating traffic analysis? Is it > just to get messages from one "gateway" to another? Or is there also > a desire to prevent traffic analysis from one non-gateway end user to > another? The goal is to completely defeat traffic analysis, while allowing the user the freedom to make use of the system through ordinary email. If email is used, the risk taken by that user goes up, but without reducing the security of other users. > What are the allowed capabilities of the opponent? Can he watch all of > the links? Can he subvert some gateways? In the real world, it would be very difficult to watch all of the links but fairly easy to subvery some gateways to some extent. However, as I have argued elsewhere, I think that all of the central gateways could be compromised and it would make no difference, so long as the number of users was reasonably large and so long as all of the users used gateways. From the opponent's point of view, the problem is that he cannot tell whether there is any traffic at all. Everyone could be whiling away a hot summer afternoon sending noise. The only attack would be to destroy or modify the incoming traffic. If there are any gateways functioning correctly, RN software should detect the damaged packets and route around the gateways that don't work right. This is exactly what the Internet does. > Does every user expose the source and destination information of his > messages to the initial gateway? What other information is sent by the > user to the RN? A user sending encrypted messages via email reveals his source address. He should encrypt his message. The message can be to a 'far' gateway which then remails it; in this case the 'near' gateway does not know the destination address. Messages can be nested to an arbitrary depth. If a user is using a gateway, the other gateways know that the message originated at the gateway, but they cannot tell whether that is the true source of the message. If the destination is another gateway, the other gateways do not know whether that is the true destination. > Are there any limitations on the information which spreads through the > RN? E.g. are gateways allowed to send source/dest information > along with the messages? If the message is to be acknowledged back to the source, the source gateway must be able to receive the acknowledgement. This creates a trail of pointers through the network back to the source. Only the final gateway, which reassembles the message, knows the ultimate destination. > Here are some questions related to Jim's specific points: > > >1.6 the order of dispatch of packets is randomized > For 1.5 you defined what randomized means. What does it mean here? Each gateway must dispatch a certain number of packets. There are a certain number of slots to be filled and a certain number of packets queued for dispatch. Packets are assigned an output slot (that is, they are delayed for a certain amount of "time") according to some sort of probabilistic distribution function. Empty slots are filled with noise packets. Inter-gateway administrative traffic is queued just like any other packet. If a gateway is always connected to the internet, packets can be dispatched at more or less equal intervals (measured in seconds) or they can be batched. > >1.7 on average, all gateways are required to send and receive the same > > number of packets per unit of chronological time > Do you mean that all gateways send the same number of packets per time > all the time? E.g. all gateways send 100 packets per hour all the time Yes, on average, as qualified by 1.8 and 1.9. > >1.8 the dispatch randomization function adjusts the average latency > > and the distribution of latencies so that the preceding commitment > > is met, introducing noise packets as required > This could be accomplished by adding no latency at all during times when > the incoming traffic load happens to equal the desired internal traffic > level. But presumably some latency is actually used to provide reordering. > What rule would determine how much latency would be used in that case? Assume that there are only two links, one in and one out. Packets will be coming in at a more or less fixed rate. Some will be consumed locally, either because they are being used to build messages or because they are noise. So per unit time N come in and C are consumed, on average. The remaining (N-C) packets are available for dispatch. In the same time interval, G packets are generated locally. So a total of N-C+G packets are to be dispatched. The system uses a random number generator to assign a packet a dispatch time slot when it becomes ready. When the clock ticks, the next packet in the queue is dispatched. If there is no next packet, a noise packet is dispatched. The system knows how long the output queue is. If the length of the queue is increasing, the rate at which packets are dispatched will be increased. [I have used the term "latency" here to be provocative.] > >1.10 gateways are required to exchange the same number of packets in > > any session > What is a session? Do you mean, during every session exactly (say) 1000 > packets will be exchanged, or do you mean, during any session the > number of packets exchanged by each gateway will equal the number ex- > changed by every other gateway (but this number may vary from session to > session)? If your gateway connects by dial-up, then the length of time that you are connected to RN is the session time. There must be some handshaking at the beginning of the session and at the end. For machines that are always on line, a session lasts from one breakdown in inter-machine connections to the next. If two machines A and B are connected, then if A sends B 100 packets per unit time, B must send A 100 packets. > >2.4 message delivery is reliable, in the sense that the destination > > gateway will report delivery of incomplete or damaged messages > > to the gateway > To which gateway? The source gateway? To the gateway which packetized the message, the source gateway. Assuming that 'MIRVing' of messages is permitted, the second message in a group could be an acknowledgement back to the originator. > >4.2 where gateways are operated by users, the requirement that gateways > > should exchange the same number of packets per unit time would be > > weakened in some as yet unspecified way > Why do this? I think that you must allow for the possibility that the gateways carry very heavy traffic, say a T1 load (about 1.5Mbit/s). Then if a user's machine was talking down a 14.4Kb/s line, allowing the user to connect would effectively stop the network. There must be some provision for inequality in traffic rates along different links. > >5.1 in either case, users may have accounts with gateways and may be > > charged for usage > What gateways would be in a position to charge users? Only the source > gateway? The destination gateway? Others in between? I assume that in a commercial network, the gateways have accounts with one another that are settled periodically. Essentially they charge each other for non-noise incoming packets at some agreed rate and then pay the accumulated difference every so often. Users should pay the gateway which fragments a message. The charge should be proportional to the size of the message in packets. If messages are nested, you need to include postage. This requires anonymous ecash. > >6.0 RN gateway software should be available only from trusted sites by FTP > What are you trying to prevent by this, and what would happen if someone > wrote his own version of the RN software? I am trying to prevent the inevitable. Weaken this requirement, eliminating the word 'only'. Publish the specs as well, and then say "RN gateway specs should also be available from trusted sites..." > >7.1 established gateways would be encouraged to rate new gateways > What kind of information would be available to them to create the ratings? Gossip, rumors, route announcements and 'hello, here I am' messages from the operators of the new gateways, experience in RN data communications with them, reports from commercial credit agencies, ... whatever information they could lay their hands on. The technical information would be published in some standard format, for example a matrix of claimed lost message rates. -- Jim Dixon

1 0

Re: e$
by frissell＠panix.com 17 Dec '03

17 Dec '03

At 01:51 PM 8/9/94 -0400, Perry E. Metzger wrote: > >> > Yes it does. Bearer bonds are illegal in the US. >If the check is written to "cash" and is used to avoid federal >transfer reporting requirements, it is CURRENTLY illegal. I suspect >that checks made out to "cash" would be illegal in all cases if they >were widely used to avoid reporting requirements. > >Perry > Bearer bonds are not illegal in the US. Under the Tax Equity and Fiscal Responsibility Act of 1982 (TEFRA), any interest payments made on *new* issues of domestic bearer bonds are not deductible as an ordinary and necessary business expense so none have been issued since then. At the same time, the Feds administratively stopped issuing treasury securities in bearer form. Old issues of government and corporate debt in bearer form still exist and will exist and trade for 30 or more years after 1982. Additionally, US residents can legally buy foreign bearer securities. DCF "Who promises the Black Caucus that if they will help kill the Crime Bill, he will support affirmative action in executions -- starting with the White Guys responsible for the Waco Massacre."

2 1

Re: e$
by jdd＠aiki.demon.co.uk 17 Dec '03

17 Dec '03

In message <9408101407.AA24943(a)ah.com> Eric Hughes writes: > When you fly into the US, you must fill out a customs declaration. > You are required to declare money in various forms (cash, checks, > etc) > > What they ask for and what is required by law are two different > things. It's not generally illegal to allow people to volunteer > information that increases the power of the state. As I recall, you are not asked to volunteer information, you are required to provide it, and the form specifies penalties if you do not. But I will have someone who is going to the States in the next week or two get me a copy of the form. -- Jim Dixon

1 0

Re: e$
by jdd＠aiki.demon.co.uk 17 Dec '03

17 Dec '03

In message <9408101428.AA26732(a)snark.imsi.com> perry(a)imsi.com writes: > > My point is that checks made out to cash are not regarded as an > > alternative currency. > > My point is that the government doesn't give a flying fuck. They are > simply trying to stop you from playing games. The law isn't like > geometry -- there aren't axioms and rules for deriving one thing from > another. The general principle is that they want to track all your > transactions, and if you make it difficult they will either use > existing law to jail you, or will produce a new law to try to do the > same. On what experience or observation do you base these rather extreme remarks? > Your hair spliting is really completely irrelevant. If you are saying that any form of legal argument is irrelevant to the issues being discussed, then you are simply wrong. The government is staffed by a population which has more or less the same distribution of attributes as the rest of the population of the US, except that, probably they are on average somewhat better educated, somewhat more intelligent, somewhat, generally, more middle class. There are thugs working for the government who will speak and reason much as you do. There are also many reasonable and intelligent people. And there are real, genuine lawyers who understand precisely what they are doing and are moved by legal arguments. Demonization of the government is simply silly, as is dismissing all logical argument. -- Jim Dixon

1 0

future existence of free remailers?
by Jonathan Rochkind 17 Dec '03

17 Dec '03

People often like to postulate on the list that eventually there won't be any more of these philantropic free remailers, and people will be charging small amounts for every remailed message, to make some money off it. I've thought of a pretty good reason why this might not ever happen. Hal Finney (or maybe it was Jim Dixon. Probably both) recently realizd and revealed to us that if one operates a remailer oneself, you effectively hide your identity from even the _first_ remailer on your chain, because it doesn't know if the mesage is in fact from _you_, or from someone using your remailer. In practice, simply having some remailer software running that no one other then you uses wont' work. You've got to have a busy remailer running with lots of traffic coming in as well as going out. This means that if someone wants to send secure anonymous mail using remailers, it's in his best interest to operate a remailer himself, and to try and get as many people to use it as possible. So philanthropy or profit aren' the only reasons to run a remailer; one's primary reason might be to ensure oneself anonymity. You could try to get some profit out of it too, but the more people who use your remailer, the better for you, so it's in your interest to make it free so many people will use it. This motivation could provide us with lots of free remailers for years to come. Maybe. It's an interesting thing to think about, anyhow.

2 1

soda.csua.berkeley.edu
by pstemari＠bismark.cbis.com 17 Dec '03

17 Dec '03

Does anyone know what the status of soda.csua.berkeley.edu is? Neither soda.csua.berkeley.edu or soda.berkeley.edu seems to be accepting anon ftp connections today. --Paul

2 1

Re: e$
by Perry E. Metzger 17 Dec '03

17 Dec '03

Jim Dixon says: > In message <9408101428.AA26732(a)snark.imsi.com> perry(a)imsi.com writes: > > They are simply trying to stop you from playing games. The law > > isn't like geometry -- there aren't axioms and rules for deriving > > one thing from another. The general principle is that they want to > > track all your transactions, and if you make it difficult they > > will either use existing law to jail you, or will produce a new > > law to try to do the same. > > On what experience or observation do you base these rather extreme > remarks? Plonk. .pm

1 0