- cypherpunks-legacy - lists.cpunks.org

Re: A problem with anonymity
by tcmay＠got.net 17 Dec '03

17 Dec '03

At 11:32 PM 8/31/95, Scott Brickner wrote: >I was thinking about some issues related to electronic commerce, and it >occurred to me that there is a significant problem in conducting >business with untraceable pseudonyms (anonyms?). The problem occurred >to me while considering inheritance. > >If one operates a business under an anonym (as opposed to the sort of >conditionally traceable pseudonym proposed by AT&T in "Anonymous Credit >Cards" <URL:ftp://ftp.research.att.com/dist/anoncc/anoncc.ps.Z>), >there's a strategy for transferring unlimited funds to one's >posterity. > >Consider a business which typically has a lot of assets, but which are >offset by a lot of liabilities --- almost any sort of VAR will do, for >instance. In your will, you leave the key to unlock a private message >to your heir, in which you hand over the information necessary to >assume your anonym. Since the heir presumably has his own identity >(whether anonymous or not is immaterial, except to *his* heirs), and >the anonym can't be linked to you, he has no reason to care about >maintaining the reputation of the anonym. In dismantling the anonym, >he sells its assets to his own identity at a fraction of their worth, >and defaults on the liabilities. ^^^^^^^^^^^^^^^^^^^^^^^^^^^ You don't have to look to death and inheritance for this problem to crop up. Similar situations arise when: - a pseudonym simply decides to dissolve the current pseudonym and shift focus to another pseudonym (perhaps transferring a bunch of assets, then simply vanishing and leaving "no forwarding address') (This is of course the basis of any number of scams and "boiler room operations." Crypto does not completey eliminate scams like this, and, in fact, generates some new kinds of scams.) - this is also a well-known problem with any services that handle money, valuables, etc. For example, the money courier who vanishes to Rio de Janeiro. This is one thing that _bonding_ is designed to partially ameliorate. One posts a bond which is greater than the amount being carried, or at least is some large amount. (Calculations are complex, and various agencies may have various policies, depending on other reputation factors.) >Since the anonym behaved reputably during its life, it developed what >would have been a credit-worthy reputation, had it been a (traceable) >pseudonym. But, since there's nothing to link the anonym to its heirs >(or ancestors), the creditors of the anonym must eat the loss. The concept of "reputation capital" is a critical one. ideally, one never "trusts" an agent with a transaction greater than the value of the reputation capital he will lose if he defaults. There are still scams and manouvers to thwart this reputation capital scheme. The agent planning to "defect" (default, split, abscond, renege, etc.) can try to pile up as many pending transactions as possible, anticipating that the various transactees will be unaware of each other. (This of course happens in real life.) Whether cryptographic protocols (cf. the "encrypted open books" proposal by eric Hughes for one approach which may be useful) solve this problem is not known at this time. But the non-crypto world has of course not solved this problem, either. ... >A market which permits anonyms to have credit based on reputation will >probably have a constant stream of defaults caused by such behavior, >representing a significant risk factor in extending credit to anonyms >which can't be predicted by reputation. > >Comments? Lots of issues need to be thought about. My hunch is that economists, game theorists, and scam artists will all discover digital money and pseudonyms and will explore various aspects of this situation. I devoted a pretty big chunk of my Cyphernomicon to these "darker sides" of anonymity, of reputation capital, and suchlike. By no means did I cover all the issues of "crypto anarchy," but I suggest interested folks take a look at the chapter on crypto anarchy for more discussion. --Tim May ---------:---------:---------:---------:---------:---------:---------:---- Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay(a)got.net 408-728-0152 | anonymous networks, digital pseudonyms, zero Corralitos, CA | knowledge, reputations, information markets, Higher Power: 2^756839 | black markets, collapse of governments. "National borders are just speed bumps on the information superhighway."

2 1

Re: A problem with anonymity
by tcmay＠got.net 17 Dec '03

17 Dec '03

At 2:06 AM 9/1/95, Mac Norton wrote: >On Thu, 31 Aug 1995, Timothy C. May wrote: > >> You don't have to look to death and inheritance for this problem to crop >> up. Similar situations arise when: > >But these are problems of fraud that the criminal and civil >justice systems already comprehend. I think what is posited >is something a bit more unique. The guy who takes in money and vanishes, or skips to Rio, out of the reach of extradition treaties, is not all that different from the posited situation of a nym dematerializing. The "law" can't really touch either situation, efficiently, and so other mechanisms are generally used (or recommended, though many con victims have fail to do so). I've not claimed the scenarios are identical, only that the issue of agents taking in money or promising services and then vanishing is as old as history. That such things will happen with digital pseudonyms is assured. Fortunately, countermeasures appear practical. ... >> - this is also a well-known problem with any services that handle money, >> valuables, etc. For example, the money courier who vanishes to Rio de >> Janeiro. > >I don't think bonding is applicable in this posit, given the >assumed established creditworthiness of the original anonym. >Perhaps he originally posted a bond, but the market would >probably have dispensed with that condition as an unnecessary >formality at some point in his glowing credit history. > >MacN If the "market" (actually, some players in the market) chose to dispense with bonding and then got burned, so be it. Later iterations of the market, and the players, will thus likely _not_ dispense with bonding and other such measures. Think of it as evolution in action. --Tim May (P.S. Could people try not to include all of the post they are responding to in their replies?) ---------:---------:---------:---------:---------:---------:---------:---- Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay(a)got.net 408-728-0152 | anonymous networks, digital pseudonyms, zero Corralitos, CA | knowledge, reputations, information markets, Higher Power: 2^756839 | black markets, collapse of governments. "National borders are just speed bumps on the information superhighway."

1 0

LuRaSHA vaporware...
by Deranged Mutant 17 Dec '03

17 Dec '03

The encrypted filesystem I promised never came through, partly because I've been busy but also because I'm toying with writing it for OS/2 instead... alas I know little about OS/2. Either way, the encryption/decryption routines are nicely written, and I am willing to make them available for perusal either way (their written in 386 Assembler code). Drop me a note if you're interested.

1 0

sums with BIG numbers
by Sherry Mayo 17 Dec '03

17 Dec '03

Can anyone point me to any books, documentation or whatever that will explain the methods used in routines like bignum for doing sums with 'too-big' numbers. I'm having a tough time trying to figure it out from the code ;-( Sherry

5 4

Re: A problem with anonymity
by tcmay＠got.net 17 Dec '03

17 Dec '03

At 10:45 AM 9/2/95, David Murray wrote: >I've often thought that in a system of digital pseudonyms, where no-one need >trade with a negative reputation (a reputation deficit?), something like >Akerlof's Market for Lemons will arise, and _all_ pseudonyms will be treated >as (reputationally) worthless. Doubtful, as we already have evidence that people are treating some digital pseudonyms as reputationally valuable. Examples abound. >[Akerlof, if I remember my economics right (and I am confident that I will >be corrected if I don't) analysed a market for used cars. There were two >types of cars: good ones, and lemons. A purchaser couldn't tell the difference >until she had bought the car. Since the expected value of a used car was less >than the value of a good car, purchasers wouldn't pay the good car price. >But that would mean owners of good cars wouldn't offer them for sale (in >this market). So the only cars for sale would be lemons :-)] I haven't encountered this example, but it clearly misses some important real-world issues. First, people buying used cars are strongly advised to take the car to an independent mechanic to be checked out (a kind of variant of the "cut-and-choose" protocol at work). This often reveals lemons. Second, people take test drives, look under the hood, kick the tires, etc. This also often reveals lemons. Third, the reputation of the used car dealer is, despite nearly a century of jokes to the contrary, often very important. The last two cars I've bought I bought used--albeit with low mileage on each--from car dealers. I took test drives and got a limited warranty in each case. The first car I drove for 12 years with essentially no problems, the second I've been driving for almost 3 years. Like a lot of simple game-theoretic models, the application to the real world is quite different. But I certainly agree that crypto will reignite interest in analyses of such game theory questions. Another way of viewing anonymity vs. non-anonymity is that knowing the True Name of a party with whom one trades is just _one element_ of a transaction. By no means is it essential. (Think of various trading situations where one has no idea of the True Name of the other parties: cash-and-carry transactions, flea markets, many international trade arrangements, etc. As we have discussed in past discussions of anarchy, the international trading regime is essentially an anarchy, in that no Higher Authority exists to resolve disputes in a top-down way...the so-called "Law Merchant" evolved to resolve disputes in such trading situations.) >As Tim points out, this is a non-crypto problem as well, and devices such as >bonds or (which are game-theoretically similar) expensive advertising or >plush premises [if they spent an unrefundable $20million on the Rolling >Stones, they're not likely to throw it all away by ripping you off for >$100 ;-)] are used to convince potential customers of one's bona fides. > >How these transfer to the world of cyber-finance, I'm not sure, but I suspect >it leaves a role for True Names in the management of credit risk: as escrow >agents, middlemen, clearing houses etc. [Although, having said that, if the >Akerlof analysis applies, you just *can't* grant (unsecured) credit to >pseudonyms - the percentage of defaulters will be 100...] But these Names are >True only in the sense that they are juridically persistent (that is, if >they transact today, they can be sued tomorrow), and need not be traceable >to any True People (Warm Bodies?) - anonymously held corporations, for >example. > >If you can't rely on the unsecured promise of a digital pseudonym, and you >can't accept reputation as 'security', how do you extend credit? I am willing to extend some amound of credit to PrOduct Cypher, Black Unicorn, etc., based on their past reputation and on the fact that I can show to others the transactions into which their pseudonyms entered and thus expose them if they default. Now _how much_ I'm willing to extend is of course a more complicated issue, but the principal is still there: a purely digital pseudonym, with no possibility of being tied to a True Name, can still be extended credit....I just said I would do so. --Tim May ---------:---------:---------:---------:---------:---------:---------:---- Timothy C. May | Crypto Anarchy: encryption, digital money, tcmay(a)got.net 408-728-0152 | anonymous networks, digital pseudonyms, zero Corralitos, CA | knowledge, reputations, information markets, Higher Power: 2^756839 | black markets, collapse of governments. "National borders are just speed bumps on the information superhighway."

1 0

Flame Wars
by zoetrope＠infinet.com 17 Dec '03

17 Dec '03

I am sorry if I have annoyed or offended anyone here, It was not my intention of course......I dont want to hold any grudges especially with Sameer....I havent done much cracking and have alot of questions I bet alot of you had when you were 13 anyway zoetrope as in coppolla's studio....again sorry If I offended or annoyed you.... Sincerely, CyberPunc X

3 2

Re: A problem with anonymity
by Bill Stewart 17 Dec '03

17 Dec '03

Scott Brickner's example has fairly close analogies today. Corporations, cash, and assets with variable value are already useful. Alice is the dying parent who runs a business or has a good credit-rating. Bob is the irresponsible kid, Carol is the about-to-be unlucky creditor. Alice borrows lots of money from Carol, tells Bob that it's in the mattress / numbered Swiss Bank Account / collection of $$$$ artwork at home. Alice dies, and Bob absconds with the money (probate isn't needed for Swiss bank accounts or unrecorded stuff in mattresses) or sells the paintings to the highest bidder of Dave, Eve, and Freddie (with a bribe to all of them, and a kickback from Dave when he finally sells the painting for its real value.) (Or maybe it's not bribery and kickbacks, he's just a major shareholder in the Glorkspitz corporation, which is a major shareholder in their businesses.) Carol is stuck trying to get money out of The Alice Estate, which turns out to be surprisingly broke. If she lent the money to The Alice Company (which looks more creditworthy than Chronologically Challenged Widow Alice), Alice would have had to leave Bob the yellow-sticky with the store safe combination instead of leaving the cash in her mattress. Now, one reason we have probate courts is to try to stop this sort of fraud, making sure creditors get paid and heirs get their shares. Another reason, of course, is to make sure "Uncle" Sammy gets to inherit part of it, even though he's not a relative. (That's of course one of the reasons for having Swiss bank accounts you didn't remember to mention on your tax forms, fireproof mattresses, foreign corporations, and friends in the art dealer business :-) Does good anonymity make this easier? Sure. And if all the heirs are cooperative, they can conspire to rip off their parents' creditors, if they're dishonest, or at least to avoid taxes, if they're not tax-believers. Of course, if they're dishonest, they'll have plenty of opportunities without having anybody die, and that'll probably affect bankruptcy law and increase Federal self-justification for watching the money supply long before anonymous estate ripoffs become common. #--- # Thanks; Bill # Bill Stewart, Freelance Information Architect, stewarts(a)ix.netcom.com # Phone +1-510-247-0664 Pager/Voicemail 1-408-787-1281 #---

1 0

Re: Phil Zimmermann/Amnesty International?
by rudy (r.) rawlins 17 Dec '03

17 Dec '03

In message "Phil Zimmermann/Amnesty International?", rrothenb(a)ic.sunysb.edu writes: > > > > I was wondering if the Zimmermann case would be a proper concern of > > Amnesty International. [snip] > > > I've heard that A.I. does not discuss political persecution as much in > the countries that are doing it, since they do not want to offend the > powers-that-be in that country, though they will note a case in another > country. (In other words, if A.I. did note the PRZ case they'd only > discuss it outside the United States.) > > This is just speculation though... > You're right on the speculation; A.I. has no such policy. As a matter of fact, they will meet face to face with government officials to seek the release of 'prisoners of conscious' - which is a possible category for Phil Z. Though Phil, despite what we may think, is not a prisoner, yet..

1 0

No Subject
by Pasquale Piombino 17 Dec '03

17 Dec '03

Hi all, I am searching PGP software. Does anyone know where I can download it? Thanks for answers. -+-+------------------->->--------------------------------:-Q------------ ° Pasquale Piombino | | Via Colombo, 35 Email: mr.xxx(a)ce.flashnet.it ° ° I-81100 CASERTA Phone - Fax: 0039 823 329152 | | ITALY ° +-+-------------8-)---------------------:-))-----------------:-o---------

1 0

Re: Some details on RSA Secure
by Bill Stewart 17 Dec '03

17 Dec '03

At 08:48 AM 8/31/95 PST, "baldwin" <baldwin(a)RSA.COM (Robert W. Baldwin)> wrote: >Here are my biased comments on RSA Secure. They are biased by >the fact that I an a techie who works for RSA Data Security. >Neatest Features: >- It is integrated with the File Manager on Windows and the > Finder on Mac System 7. It's an interesting approach - simpler than installing as a file system (I'm not using SecDrv because it needs a separate partition), but so far it seems relatively useable. On the other hand, since there are file types it refuses to encrypt, like DLLs, there are some things it can't protect (like DLLs with passwords embedded in them.) >- A settable list of files can be automatically decrypted (or > encrypted) on system startup (or shutdown). I've got mixed feelings about this - since my password has normally timed out by the time I shut down, it demands a password before shutting down, and if I had any large amount of data there, I assume it would take a while to reencrypt on my 386-box. (Laptops are generally slower than similar-age desktops, and you often want to shut them down quickly, even when they aren't shutting themselves down to save power.) I haven't tried powering it off during this process, but I assume that at best the files would be unencrypted and at worst there'd be one half-decrypted and sitting in limbo to be trashed the next time the system starts up? >Technical Features: >- The user's passphrase unlocks a master key that is used to unlock > the file encryption key for each file. Since the documentation mentions 512-bit and 1024-bit RSA keys, I'd guess that each file has a separate random RC4 key which is stored in the file header, encrypted with the user and escrow RSA keys plus the MD5 hash, plus the encrypted real filename? (The alternative would be that the userpref.!!! file contains an RC4 key encrypted with RSA, which is used to encrypt the file keys, but that would lead to much shorter headers, and cracking that master RC4 key would then allow cracking of all documents on the system, so I'm guessing that's not the approach used.) >Request for Improvements: 1) The "Emergency" menu item in the File Manager is annoying. Could it be combined with the RSA menu item, eliminated, or at least have the option of abbreviating the name? 2) Can keys and files be shared between multiple machines, i.e. the same user keys on a desktop and laptop, so that individual encrypted files can easily be moved back and forth? #--- # Thanks; Bill # Bill Stewart, Freelance Information Architect, stewarts(a)ix.netcom.com # Phone +1-510-247-0664 Pager/Voicemail 1-408-787-1281 #---

1 0