December 2003 - cypherpunks-legacy

The Trouble With Crypto
by blancw＠pylon.com 17 Dec '03

17 Dec '03

Responding to msg by nobody(a)ds1.wu-wien.ac.at: 1 >if [crypto] only >needs to be used once in a while, there will be no >good reason to implement some of the more interesting >protocols, or implement the "maximum strength" >possible. 2 >For the overwhelming majority of >people, the benefits of "digital cash" will not be >worth the time and trouble over "digital cash with >anonimity removed". . . 3 >As for encrypting all email, much like people use >envelopes? Be honest, there isn't sufficient cause to >warrant the time and trouble. 4 >Nor is there sufficient cause to warrant the time and >trouble of signing messages sent to mailing lists or >usenet. . . . 5 >Nor is there sufficient cause to warrant the time and >trouble of communicating via anonymous remailers, >except for say folks like Pr0duct Cypher. 6 >Nor is there sufficient cause to warrant the time and >trouble for banks and stores to offer digital cash. 7 >As for dc-nets, give me a solid example why you ever >need to communicate with one. 8 I see a limited deployment, and almost no >fundamental restructuring of society. 9 >. . . I suppose a discussion about >atomic bombs will likely be of greater impact on our >future than crypto anarchy will. 10 >Cypherpunks write code, but if there is sufficient >cause to warrant the time and trouble! ............................................................... You might be right, having accrued at least 10 reasons why the list discussions do not altogether convince of the importance of using encryption as a matter of course or for the re-structuring of society. The choice to use crypto is a little different from the sense of wanting to use it from desperation; I think it is the difference between determining factors: when it is the individual themselves who decide to employ the tool for whatever reason they have to either use it or not at their discretion, or when the circumstance seems to dictate for the person what they must do - that they must go to desperate means to ensure privacy, from a perceived threat which demands that they hide their communication. One of the important issues regarding the use of encryption is not necessarily whether it is used or not as a matter of course, but rather the controversy over the source of the permission to use it as well as the imposed obligation to participate in self-incriminating applications of it. i.e. do individuals have the sovereign right to use tools which result in a division between public & private existence, or are they obligated to keeping their lives accessible to intervening governing agencies? To me a cumbersome tool would require sufficient cause to use it. However, I would appreciate its existence in case of emergency, if there was no better one available, and I would protest the idea that it was anyone else's prerogative to decide for me when it was an appropriate occasion to do so. Is crypto only a toy with destructive implications for governments & societies, or a tool of subjugation with destructive implications for individuals? If only cypherpunks or only government officionados made the decisions about it the answers would be easier to predict. But they are not the only ones involved, and it is my understanding that not all future developments will be determined on this list. Blanc

1 0

Magic Money / NexusBucks
by nobody＠vox.hacktic.nl 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- To those at the Nexus who are using Magic Money: thank you and good luck. So far you are planning to accept Magic Money as payment, and to pay out Magic Money to a few people who write some code or otherwise help you out. This will not get it into widespread use. Instead, everyone who signs up for one of your services (Internet access, etc.) should get a rebate back in Magic Money. This rebate can be used toward the next month's bill or any other service you provide - but only if the person learns how to use Magic Money and sets up a client. Now there is a real economic motive for quite a few people to use it. Direct exchanges between the users will follow naturally. In your description of Magic Money and the Nexus system, you said that a bug in the Magic Money client would cause loss of coins if you enter a bad coin value while entering coins to be paid out. This is not true, at least not on the systems I've tried it on. You get an error message and a rather ominous-looking blank line, but entering a correct coin value will continue the process. Also, for automation and Perl scripts, you might want to take a look at the v1.1 auto-client. The auto-client is designed especially for control from another program or script. Control from a perl script would be very easy - much easier than controlling the manual client. I also encourage you to standardize on v1.1 now because it adds the bank's keyid to the end of each coins.dat file. The auto-client can retrieve the keyid, and its presence will be very important if you ever start using multiple banks. Pr0duct Cypher Cypherpunks build Bombs? -----BEGIN PGP SIGNATURE----- Version: 2.3a iQCVAgUBLmAfDMGoFIWXVYodAQFmfQQAj79YF9oYtzakrhBgEcLDB8SezWc89o5f TE1jnjKLNnLQGRlzcLuqv0PSRP02ABhAT7VcPkqnOYH28ctvD+SR8MiB65IZI0Ul NHnD6mvSqY7kF0RQVjdVTXPBmGdAq5GZIaLEDl8n1wkh/cf/4XY30J3R13Ac0+Ds ZqrHviE1mQc= =2Dj6 -----END PGP SIGNATURE-----

1 0

software wants to be sold.
by Gary Woodward 17 Dec '03

17 Dec '03

Anyone want to be part of a list to discuss sales and marketing issues of data security software? I will run the list. Regards Gary ............................................................................. Gary A. Woodward internet:p00140@psilink.com SECURITY FEATURES Compuserve:71564,247 P.O. Box 5549 phone: 202.928.1231 McLean, VA 22103 infosec & docsec

1 0

DigiCash ???
by Jim choate 17 Dec '03

17 Dec '03

Hi all, I would like somebody to explain how I would go about using an anonymous digicash system to buy a automobile? Just for arguments sake lets say I want to buy a new Ford Mustang GT from the local dealer and desire that nobody at the dealer has any idea who I am or where I come from.

3 3

someHost@nsa.gov
by Pat Farrell 17 Dec '03

17 Dec '03

In message Sun, 28 Aug 94 03:11:00 EDT, Carl Ellison <cme(a)tis.com> writes: > I ran into an NSA policy office person at CRYPTO'94 and suggested to him > that the NSA should set up a WWW page from the crypto museum -- and also > post occasionally to sci.crypt from some nsa.gov host. He was very open > to the first idea but not to the second. (problem apparently with the > standard disclaimer not being good enough for them. :-) This is strange. I see no difference between somehost.nsa.gov and docmaster.ncsc.mil ncsc is nothing but a public front for NSA. It is located at Ft Meade, etc. This is not news. Why wouldn't they be willing to post from dockmaster? Pat Pat Farrell Grad Student pfarrell(a)cs.gmu.edu Department of Computer Science George Mason University, Fairfax, VA Public key availble via finger #include <standard.disclaimer>

2 1

Re: DigiCash ???
by jdd＠aiki.demon.co.uk 17 Dec '03

17 Dec '03

In message <9408281521.AA13945(a)ua.MIT.EDU> Jason W Solinsky writes: > > I would like somebody to explain how I would go about using an anonymous > > digicash system to buy a automobile? > > > > Just for arguments sake lets say I want to buy a new Ford Mustang GT from > > the local dealer and desire that nobody at the dealer has any idea who I > > am or where I come from. > > You hire an anonymous reshiper... no different from an anonymous remailer > only you can't use an automated program (yet :) and its damn difficult to > avoid traffic analysis. Do we care that you can't get it on the road without insurance in most states, and that the registration must be in someone's name? -- Jim Dixon

2 1

FCC and Crypto
by John Young 17 Dec '03

17 Dec '03

Keep it up. Literate writing is ur-code. Crypto-tech entrepreneurs will smell the coffee. John

1 0

Re: DSPs
by norm＠netcom.com 17 Dec '03

17 Dec '03

At 13:09 1994/08/26 -0700, Phil Karn wrote: .... >But then I hear people say that it's not the multiplication that slows >down modular exponentiation, it's the modular reduction. .... Modular reduction is scarcely worse than the multiplication. If I have a 60 word multi precision number N to be reduced by a 30 word number M, I compute a guess by dividing the 32 bit most significant bits N by the most significant 32 bits of M. I then multiply this quotient by M and subtract that from N. That reduces N by some multiple of M leaving N mod M unchanged. The error in the guess might mean that N is less than 32 bits shorter than it was before the operation but this method gets nearly 32 bits per pass. The inner loop of the is the same as in multiplication. For all of this using the floating point unit wins on most modern CPUs.

1 0

Golbal Econ.
by Rick H. Wesson 17 Dec '03

17 Dec '03

With topics like buying a car with digital cash make me think some may be missing my point. I don't want to replace all currencies world wide with some implementation of digital cash, be it an on or off-line system. I would like to help get an internet economny going! For the internet to become a source of greater revenue I'd like to see a system that can securely make business transactions that we are all happy with. The amounts should be small transactions compaired with selling a car, I'd like to know how I can pay for some online information without having to have an account based with the information supplier. For starters how can I pay for some HTML page or purchase a physical product via CURRENT technologies on the internet. Lets first start an internet cash system before you go off and let our ego's make all other forms of hard currency obsolette. -Rick

1 0

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

17 Dec '03

In message <199408281656.JAA14318(a)netcom.netcom.com> Norman Hardy writes: > Modular reduction is scarcely worse than the multiplication. If I have a 60 word > multi precision number N to be reduced by a 30 word number M, I compute a guess > by dividing the 32 bit most significant bits N by the most significant 32 > bits of M. On a DSP chip like the Texas C40, 32-bit multiplication takes one clock cycle. Modular reduction will take something on the order of one hundred clocks. Modular reduction is much more expensive than multiplication. > I then multiply this quotient by M and subtract that from N. That reduces N by > some multiple of M leaving N mod M unchanged. The error in the guess might > mean that N is less than 32 bits shorter than it was before the operation but > this method gets nearly 32 bits per pass. The inner loop of the is the same as > in multiplication. > > For all of this using the floating point unit wins on most modern CPUs. Not on DSP chips. On the C40, reals are only 32 bits long, so there is no benefit to using them. They are less precise than integers. -- Jim Dixon

1 0