December 2003 - cypherpunks-legacy

Re: The future will be easy to use
by Bill Stewart 17 Dec '03

17 Dec '03

At 03:30 PM 11/28/95 EST, Carl Ellison <cme(a)TIS.COM> wrote: >BTW -- PGP currently lacks a way for me to note, when I sign a key, how it >is that I trust that key (by personal meeting, by attribution, by message >association, ...). A signed attribute record would let me record that >information for myself as well as for others. That would be a useful feature, even if it's just an unstructured text file. There is a way to do it now, though it's inefficient and hokey - create multiple key-signing keys, with name fields indicating the attribute, sign them with your main key, and use the appropriate one of them to sign keys for people. For instance, I have a key named "Bill Stewart Unauthenticated Pseudonym Signing Key <stewarts(a)ix.netcom.com>" which I use to sign keys of significantly lower trust than my normal key; you could do similar things for higher-quality certification. This does increase the depth of the web-of-trust required, which is less of a problem for low-trust keys than for keys you actually care about :-) #-- # Thanks; Bill # Bill Stewart, Freelance Information Architect, stewarts(a)ix.netcom.com # Phone +1-510-247-0663 Pager/Voicemail 1-408-787-1281

1 0

Code to brute-force PGP passphrase?
by Rich Graves 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- I have an old, forgotten PGP key for which I would like to issue a revocation certificate (or maybe I stole somebody else's with an identical name; who could tell?). I can put the key on a Mac, PC, or UNIX machine (various flavors). I think I have a pretty good idea what the middle of the passphrase contains, so I'd guess there are only about 2,000 possibilities to be explored. Anybody have code for any platform for cracking this? Alternatively, I could just give you the key and wait for the answer. - -rich -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAwUBMLzGmI3DXUbM57SdAQE8NAP9EKVkxVzsTPHQ0NcJJgU4aor8BQbjN2a5 OqNAkpeNoZOcn5NcqGBMjI6mVQbgVjnzDOdoDWngGmWbtruFcFjE1IM8iK/wScah Caqo7SuKQqUMXB8mAC5pVYmdavJrOX4WPcl6nT2MkJTvnJe7YgiCMWIdm+4XCPZ/ YybyYhWf7XA= =W5H8 -----END PGP SIGNATURE-----

1 0

Re: key for Alice as promised (not)
by jimbell＠pacifier.com 17 Dec '03

17 Dec '03

>On Tue, 28 Nov 1995, Adam Hupp wrote: > >> >Can you imagine?? I'm simply not willing to fool myself into thinking >> >that I ahve security by posting a key and using PGP. >> >> Unless you can post some proof that PGP is insecure, stop insisting it is. > >Hold on a minute. Alice is, here, 100% correct. no, he isn't. He isn't even using a minimal effort to prevent spoofing, and he's blaming this on the fact that some particular piece of software (for example, PGP) isn't absolutely foolproof. >If I use PGP to read messages and there's a videocamera trained on the >keyboard, and other people have access to the machine, PGP is not >secure. Similarly, if PGP is on a computer which other people may use >without my supervision, they can monitor keystrokes, etc. and PGP is not >secure. True but irrelevant. This is especially true since the only purpose to signing messages to a public area is to prevent spoofing. If somebody does, indeed, figure out how to break (say) a 1024-bit PGP key and "Alice" gets spoofed, HE WILL KNOW because he will see a message with his signature that HE KNOWS he didn't send. At that point, he will at least be able to reduce the number of spoofed messages to 1 before he alerts us that there is a problem. We won't necessarily know who to believe, of course, but we will know that one of a number of this is probably true: 1 "Alice" is lying to us. (by far the most likely.) 2. Somebody actually has physical access to "Alice's" machine and is actively using it to spoof messages. 3. Somebody found out how to break 1024-bit PGP keys easily. (Very unlikely, of course.) >A chain is only as strong as its weakest link; Alice recognizes this, and >makes no claim that PGP itself is the weak link. But "Alice" resists using ANY chain at all!

2 1

Response to ping re: CJR
by Raph Levien 17 Dec '03

17 Dec '03

I just got a message on my answering machine from Sam Capino. He said they are working on answer to my CJR. They had hoped to get it out by now, but were set back by "the furlough." I had called him about a week ago to check up on the status, but at the time only got his answering machine. Of course, the original 15 days have passed, but I'm perfectly willing to let that slide since I'm not completely sure that the CJR was submitted using exactly the right process. Just thought you'd like to know. Raph

1 0

"Proprietary" internetworking protocols (was RE: The future will be easy to use )
by Pete Loshin 17 Dec '03

17 Dec '03

Perry E. Metzger writes: >attila writes: >>figures. I'll give ipsec and ipsec-dev a look. However, SUN does have >>the power to make something happen on the high-power workstations, and the >>fact they are making a portable package available in source code is >>farther than anyone else has gone. > >Unfortunately, an internetworking protocol used by only one vendor >gets nowhere. much other discussion ommitted... I hate to quibble, but past experience may demonstrate this to be inaccurate. In particular, and most relevantly: NFS (Sun) SSL (Netscape) Another internetworking protocol (IPX) gave Novell a good run. The point is that a working implementation goes a long way for people who want/need _something_ that works, NOW. Consider the relative merits/success of S-HTTP and SSL. ObCrypto: the POTP people are the Boston Internet Expo today and tomorrow, and they have their "technical guy" who is supposed to be answering questions. The nontechnical guy said they generate random numbers using randomness taken from the text of the messages being encrypted. -pl _Electronic Commerce_ (Charles River Media, available now) pete(a)loshin.com

4 3

Question on modular exponents
by Usuario Acceso2 17 Dec '03

17 Dec '03

Hi CP's: I'd be pleased if someone explained me this ASAP: How could I demonstrate that _ (g^a)^b (mod p) = g^{(a*b) mod q} mod p _ when g's order in Zp is q? The same question remains for adition. I have the feeling the demonstration must be similar. Best. Jaime

1 0

Re: GOST for sale
by JMKELSEY＠delphi.com 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- >Date: Fri, 24 Nov 1995 03:43:25 +0100 (MET) >From: Mats Bergstrom <asgaard(a)sos.sll.se> >Subject: GOST for sale >According to a short article in a Swedish newspaper (DN) >with the title 'Spy Code of KGB can make computers safe', >JETICO INC., located in Finland (Tammerfors), introduced >a new crypto system on the world market last week. It's >based on GOST, the Russian federal standard algorithm. >This product, called BestCrypt, is implemented at least >partly in hardware. Alledgedly it uses 'GOST 28147-89', >whatever that stands for. Hmmm. >GOST is probably very secure - a huge keyspace (256 bits) and >KGB would not have left a trapdoor for NSA to take advantage >of, would they? If this is the cipher I've seen, it's not overwhelmingly impressive, though it may be OK. I haven't ever made any serious attempt to attack it, and I don't know anyone else who has. Anyone have a reference? Basically, GOST is a balanced Feistel network (like DES), but its F-function is much simpler. Basically, F(X,K) = Rotate_Left(S(X+K),11), where S(t) denotes parallel application of eight 4:4 S-boxes. Depending on the implementation, these S-boxes' contents could be key-dependent and pseudorandom, or fixed. Unfortunately small random S-boxes are likely to have some weaknesses W.R.T. differential cryptanalysis. This can be true even when the S-boxes are secret and key-derived--see Biham and Birkyov's paper in Auscrypt '94 on a DES variant with variable S-boxes, for a quick discussion of this. And the security of this scheme is very much going to depend upon the S-boxes used. If the S-boxes are generated at (pseudo)random from the key, I'd expect there to be some pretty nasty weak key conditions that could occasionally come up. On the other hand, GOST is defined with 32 rounds, so it may be hard to find any useful differential or linear characteristics, even for relatively bad S-box choices, that have high enough probability to get through 29-31 rounds. And it has a 256-bit key, so even if someone determines some attack which recovers 160 bits of the key, there are still 96 bits of key left to provide security. The GOST key schedule is really simple, though it avoids the most obvious kind of related key attack. I wouldn't be surprised to see some interesting related key attacks be possible. This is interesting because there is also a hash function based on GOST--I'd be pretty reluctant to use this without a lot of analysis. >The newspaper article ends: > >'The Finnish entrepreneurs asseverate that their >Russian colleagues have never had anything to do >with the KGB.' I may be mistaken, but wasn't there some other internal organization in the USSR that did cryptography? >Mats Note: Please respond via e-mail as well as or instead of posting, as I get CP-LITE instead of the whole list. --John Kelsey, jmkelsey(a)delphi.com PGP 2.6 fingerprint = 4FE2 F421 100F BB0A 03D1 FE06 A435 7E36 -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAwUBMLv0kkHx57Ag8goBAQFH3AP+JR1sEQCH0ANmNSDFhzpCZb4jn4uqq23v PjZfFOrYc9SrD1xn+3azqfHFpiHIubfWzZ8u1aNEf4erAErbQLPAUOLPv17zhn// 6UEVKX4GhPqoLfou+wO38rF2DwWdeCjuQm2GgxuFpw5b0D4M3V1mv5ocLOoOMq2N twdfxiMtO6Y= =BOER -----END PGP SIGNATURE-----

1 0

Re: The future will be easy to use
by E. ALLEN SMITH 17 Dec '03

17 Dec '03

From: IN%"perry(a)piermont.com" 28-NOV-1995 13:34:33.55 Jonathan Zamick writes: > I can't agree. The model of a successful enterprise includes feedback from > different levels of participants. This isn't an enterprise. The government is not a participant except by their own desire to interfere. --------- One reason to incorporate a probable opponent into something is to see what methods the opponent will likely use. For instance, their debating points would be nice to know in advance (to the degree they aren't already); while cyberpunks don't have much access to the mass media, that on the Net (where a lot of the initial potential users are likely to be) is another matter. Getting clues to their likely legal strategy may also be valuable. On the other hand, letting them know too much may be a problem... of course, I'm _certain_ the NSA, etcetera get copies of this list. -Allen

1 0

Re: Directed Hamiltonian Path Problem
by E. ALLEN SMITH 17 Dec '03

17 Dec '03

From: IN%"tcmay(a)got.net" 28-NOV-1995 20:15:30.64 The reason we speak in terms of physical volumes of "Adleman computers" is to make concrete the way things scale. If the amount of Adleman computers needed to factor, say, a 2000-digit modulus (or some reasonably equivalent Hamiltonian cycle problem, such as the TSP) is "ten Pacific oceans full of them running for 100 years," then one has a pretty clear feel for just how futile it is to ask about "But what about if we apply MASSIVE PARALLELISM?!?!" ------------ Ah. My objection is probably from being too much of a purist in my own area of science. I do tend to try to be quite careful in whatever I write to use the correct terms (then explain them for those who've heard the incorrect ones). ------------- I don't worry much about factoring breakthroughs. And I don't mean minor improvements, which keep occurring: I mean major breakthroughs which would make factoring a 2000-decimal-digit number "easy." Practically speaking, snarfing private keys is a helluva lot easier, for many reasons. -------------- Umm... it's easier for each one. But the effort in question adds up. Ultimately, for an agency (NSA, CIA, etcetera) wanting to do a lot of such unencryptings, coming up with a factoring method is the most efficient way to go. Fortunately, science doesn't work very well with security classifications (unlike engineering, which is what most military classified "science" is from what I know). -Allen

1 0

Re: Netscape gives in to key escrow
by James A. Donald 17 Dec '03

17 Dec '03

At 12:55 AM 11/30/95 -0800, Jeff Weinstein wrote: > All press accounts I've read so far > have lots of stuff attributed to Jim, but very little of it is actual > quotes. Then we would like to hear Jim "clarify" his position. Absent such clarification, we must assume he considers the press reports of what he said to be reasonably accurate. > I don't want to go down in history as a facilitator of the totalitarian > state, and I don't think anyone else here does either. Most of the major players have taken a position overtly opposed to GAK (whilst leaving themselves some wriggle room, should the heat get too severe). What is netscape's position? --------------------------------------------------------------------- | We have the right to defend ourselves | http://www.jim.com/jamesd/ and our property, because of the kind | of animals that we are. True law | James A. Donald derives from this right, not from the | arbitrary power of the state. | jamesd(a)echeque.com

1 0

Re: cypherpunks at ietf
by James A. Donald 17 Dec '03

17 Dec '03

> OUR CURRENT PROBLEM IS THAT THE AVERAGE CONSUMER CAN NOT COMPLETE AN ACCURATE > SEARCH ON THE WEB. > > [...] > > AT THIS POINT SO MANY CYBERPUNKS HAVE EMBEDDED SO MUCH GARBAGE ON THE WEB, > THAT YOU CAN NOT COMPLETE AN ACCURATE SEARCH. Gee: Someone who has not learnt to use the caps key on his computer cannot do accurate searches, so will everyone else please get off the net. > > [...] > > BUT MORE IMPORTANT, WE NEED TO CLEANUP THE TRASH THAT THE CYBER-PUNKS ARE > PLACING ON THE WEB. Pleasant comic relief. --------------------------------------------------------------------- | We have the right to defend ourselves | http://www.jim.com/jamesd/ and our property, because of the kind | of animals that we are. True law | James A. Donald derives from this right, not from the | arbitrary power of the state. | jamesd(a)echeque.com

1 0

Re: The future will be easy to use
by frantz＠netcom.com 17 Dec '03

17 Dec '03

At 15:30 11/28/95 -0500, Carl Ellison wrote: >BTW -- PGP currently lacks a way for me to note, when I sign a key, how it >is that I trust that key (by personal meeting, by attribution, by message >association, ...). A signed attribute record would let me record that >information for myself as well as for others. There is more to this problem than how it is that I trust the key. There is also what I trust it for. I just added a key to my key ring that I will use for sending confidental data to a client site. I trust that no one can access the secret key who is not also inside their firewall. However, the key is on a multi-user system, so I do not trust that it is accessable to only one person. Since the data I intend to send will be publicly available inside the firewall, I don't have to trust more than the firewall. It is hard to see how to record the information about how much I trust the receipent's systems security. Bill ----------------------------------------------------------------- Bill Frantz Periwinkle -- Computer Consulting (408)356-8506 16345 Englewood Ave. frantz(a)netcom.com Los Gatos, CA 95032, USA

2 1

Re: key for Alice as promised (not)
by anonymous-remailer＠shell.portal.com 17 Dec '03

17 Dec '03

On Wed, 29 Nov 1995, Adam Shostack wrote: > > PGP is really not the issue. The issue is more my security and the > > environment that I use PGP in. I don't have a trusted machine to run PGP > > on. Anyone who wants to can come up to machine and copy my secret keyring > > or they can even watch me typing my password in. > > Threat, please?? Do people often stand over your shoulder as > you type? Yes. > Enter your office, Yep, especially my night cleaning staff. > point guns at you, Not recently ... I'd co-operate fully in that situation, though. > and take a backup of your entire computer? You mean like the on-site backups that I have in the filing cabinet beside me, or the off-site backups that aren't here (in case of fire, or such) nd are completely outside my control? > Have you considered putting the secret keyring > on a floppy and locking it in your desk/safe when you're not actually > in the office? (Or home..) Yep, I've considerred it. It's still not all that helpful. Cleaning staff has plenty of time when I'm not around to deal with that. > > So, I don't fool myself, and I don't use PGP, except for things like > > exchanging a one-time pad with someone when I've already sent the message > > out across another delivery mechanism, like on a floppy delivered my > > courier. > > I don't follow. You're claiming that PGP is good enough to > transfer OTPads, but not good enough to sign pseudononymous messages? Sure. Two different situations. If I take a message or a data tape and encrypt it with a one time pad. And then I send the message out to someone via Greyhound or DHL. And once they've confirmed that they have the encrypted message safely in hand, then I'll call them and ask them to call me with their public key delivered by voice via telephone. Which I then use to encrypt the one-time-pad, using the PGP key only once. Then, I'm comfortable sending it (not the message, but the pad) over the Internet encrypted with PGP. And I think at that point, I have Pretty Good Privacy. > Adam > > -- > "It is seldom that liberty of any kind is lost all at once." Alice de 'nonymous ... ...just another one of those... P.S. This post is in the public domain. C. S. U. M. O. C. L. U. N. E.

1 0

Re: key for Alice as promised (not) (fwd)
by Adam Shostack 17 Dec '03

17 Dec '03

>On Wed, 29 Nov 1995, Adam Shostack wrote: > >> > PGP is really not the issue. The issue is more my security and the >> > environment that I use PGP in. I don't have a trusted machine to run PGP >> Threat, please?? Do people often stand over your shoulder as >> you type? > >Yes. And you can't ask them to leave, as you send anonymous messages? Or does your whole office know your one of those who post to cypherpunks as Alice d' nonymouys? >> Have you considered putting the secret keyring >> on a floppy and locking it in your desk/safe when you're not actually >> in the office? (Or home..) > >Yep, I've considerred it. It's still not all that helpful. Cleaning >staff has plenty of time when I'm not around to deal with that. Of course, if the cleaning staff cut your safe open, you have a good indication of that in the morning. Not that you've demonstrated that the level of effort to do all of this at all equates to what is gained, namely, the ability to impersonate you. Which everyone in the world has today. >> > So, I don't fool myself, and I don't use PGP, except for things like >> > exchanging a one-time pad with someone when I've already sent the message >> > out across another delivery mechanism, like on a floppy delivered my >> > courier. >> >> I don't follow. You're claiming that PGP is good enough to >> transfer OTPads, but not good enough to sign pseudononymous messages? > >Sure. Two different situations. [...] >And once they've confirmed that they have the encrypted message safely in >hand, then I'll call them and ask them to call me with their public key >delivered by voice via telephone. >Which I then use to encrypt the one-time-pad, using the PGP key only once. > >Then, I'm comfortable sending it (not the message, but the pad) over the >Internet encrypted with PGP. And I think at that point, I have Pretty >Good Privacy. Of course, then your message only has 128 bits of security, because the Greyhound or DHL employee could be easily comprimised, as could the message in their office. So you encrypt the pad with IDEA, getting you IDEA level security, and think you're working with one time pads?? I'm flabbergasted.

1 0

Re: ecash lottery (Was: ecash casino)
by tcmay＠got.net 17 Dec '03

17 Dec '03

At 1:20 AM 11/30/95, Ian Goldberg wrote: >[Disclaimer: The following post is a gedanken experiment. It should not > be interpreted as condoning or encouraging anyone to break any > laws, no matter how stupid the laws are.] I, on the other hand, have no compunctions about "condoning or encouraging" various kinds of actions. When condoning and encouraging becomes criminal, I hope I'll be gone. (Wiseacres will cite hypos in which encouraging a crime, such as a rape, makes the encouragers culpable. A far cry from condoning and encouraging tax evasion, gambling, etc.) On to Ian's proposal: ... >Now: there's likely no Web-of-Trust to the various PGP keys involved, >so the lottery's Reputation will have to be built up. A good way to >do this is to have smaller prizes being won fairly often (e.g. by matching >the last digit/few bits), so that it's obvious the lottery is not just >collecting money without awarding prizes. Have to be _very_ careful here. A variety of scams can be developed which show lots of "small" winners, but which fail to show any large winners. The lottery operators can make a lot of extra bucks by simply not paying off the large winnings, in various ways. More robust protocols may be possible. In particular, I recall that Eric Hughes was working on something he called "encrypted open books." While banking was the obvious domain, it seems plausible that the same methodologies could be adapted for lotteries (some isomorphisms between accounts/balances and tickets/winnings?). I don't follow Ian's proposal in enough detail to comment on it, but thought I should point out the scams over nonpayment of winnings and the "encrypted open books" notions (from sometime in '93, as I recall). Personally, I have long had great _hopes_ for using crypto for non-sanctioned gambling, but I'm pretty skeptical that many people will do it. For one thing, anyone knowledgeable enough to be comfortable with the crytography knows that gambling when a house cut exists is a lose. (And gambling with no house cut is, at best, a wash.) --Tim May Views here are not the views of my Internet Service Provider or Government. ---------:---------:---------:---------:---------:---------:---------:---- 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: y2k as ideological opportunity
by nobody＠REPLAY.COM 17 Dec '03

17 Dec '03

Blanc wrot: >On the PBS daily economics news report the other day, and an economist was >warning that this y2k problem could lead to a recession (don't remember if >he said it would be mild or what). This is a possibility that the rest of >us ought prepare for (guess I'll have to see what Harry Brown has to say >about it). A recession? Say that 20 million people figure out that they don't have to pay their taxes. Those 20 million will stay wealthy. The 25 million people employed by the federal govt in some parasitic capacity or another will find themselves without enough money to go around. So, that means the FBI, DEA, BATF, Secret Service, Treasury and all the other heavily armed, officially sponsored gangs can't feed their habits. They look at their badges and weapons and they get hard-ons and decide to extort more money from the drug traders. Only now they want 50% instead of the 25% they take today. Net result: the street price of drugs will rise, causing more crime. It'll be anarchy in the streets. Look for Tim May carrying this sign: +--------------------+ | I Told You | | The End Was Near | | Seven Years Ago, | | Check The Archives | +---------++---------+ || || The only hope for America? Buy more drugs to keep the traders cash flowing. An artificial drug-induced happiness will descend upon the land, and we all reelect Gore in 2004, because we're STILL STUPID SHEEPLE. THE DRUGS WON'T HELP! Some things will never change. GangMongrel

1 0

RE: ecash lottery (Was: ecash casino)
by Pete Loshin 17 Dec '03

17 Dec '03

Ian Goldberg wrote: > >Here's an off-the-cuff idea: > >[Disclaimer: The following post is a gedanken experiment. It should not > be interpreted as condoning or encouraging anyone to break any > laws, no matter how stupid the laws are.] > >Anonymous email lotteries: [details deleted] >A lottery consists simply of displaying the winning number and random >noise, whose hash was previously posted (so the participants know >you're not cheating). Traditional illegal numbers rackets usually use some publicly available number so everyone knows it's all on the up and up (I used to know what the source for the winning numbers actually _is_, though I think it has something to do with the stock market, e.g., the last three digits of the number of oddlot trades, or something else that is usually essentially random. In any case, using such a "public" and daily number eliminates the problem of proving the game isn't fixed. -Pete Loshin pete(a)loshin.com

1 0

Because they want to be "players"
by Tim May 17 Dec '03

17 Dec '03

At 5:53 PM -0700 9/18/97, Declan McCullagh wrote: >But wait... The version of SAFE "as is" contains the ***FIRST EVER*** >domestic restrictions on encryption! Why should it be passed intact? For >the sake of Beltway politicking and deal-cutting? > >It includes very, very troubling severe criminal penalties for the use of >encryption in a crime. When encryption is in everything from light switches >to door knobs, any crime will include crypto, no? It would be like >criminalizing "breathing air in the commission of a crime..." Yes, SAFE is an evil bill. And anyone who supports the "use a cipher, go to prison" language (*) is guilty of a most serious crime. (* Some object to this slogan, saying that crypto use is not actually outlawed. Well, neither were guns outlawed when the billboards went up in California and other states: "Use a gun, go to prison." The meaning is clear.) >Why not just say "stop SAFE altogether?" No new laws are better than bad >new laws. And even if the crypto-in-a-crime provisions are yanked, SAFE may >be a bad bill. I wrote about this in June: Indeed, I also condemned it earlier this year. .... >Why are these organizations -- CDT, VTW, Wired, EFF, ATR -- urging it be >passed intact, as is? Why should Americans give up their rights so business >can make more money on encryption exports? > Because they want to be "players." And to be "players," they feel they have to get in their at the pork barrel and bargain away our rights. Fuck them all. They are all scum. (The ACLU is far from perfect, but I am gaining new respect for their absolutist stance on most civil liberties issues. I am assuming that the ACLU will not support the "speak in Spanish while committing a crime and get an extra 10 years" language? This, by the way, is what the "use a cipher" language is exactly parallel to.) --Tim May The Feds have shown their hand: they want a ban on domestic cryptography ---------:---------:---------:---------:---------:---------:---------:---- Timothy C. May | Crypto Anarchy: encryption, digital money, ComSec 3DES: 408-728-0152 | anonymous networks, digital pseudonyms, zero W.A.S.T.E.: Corralitos, CA | knowledge, reputations, information markets, Higher Power: 2^1398269 | black markets, collapse of governments. "National borders aren't even speed bumps on the information superhighway."

3 2

Re: ecash lottery (Was: ecash casino)
by tcmay＠got.net 17 Dec '03

17 Dec '03

At 2:00 AM 11/30/95, Dr. Dimitri Vulis wrote: >iagoldbe(a)csclub.uwaterloo.ca (Ian Goldberg) writes: >> So; is there a problem with the implementation outlined above? > >What about the income tax on winnings? That's a good one! A bit subtle as humor, though. (For many of us, even most of us, the whole _attraction_ of such gambling schemes--to the extent any gambling scheme can ever be attractive to a rational person--is the tax avoidance. And any scheme that allows taxation of winnings, but not losses on individual bets, would kill this proposal dead.) --Tim May Views here are not the views of my Internet Service Provider or Government. ---------:---------:---------:---------:---------:---------:---------:---- 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

Fertile ground for papers....
by tcmay＠got.net 17 Dec '03

17 Dec '03

At 3:22 AM 11/30/95, Pete Loshin wrote: >In any case, using such a "public" and daily number >eliminates the problem of proving the game isn't fixed. Maybe. In passing, the whole field of cryptography is filled with interesting papers with titles like "How to toss a fair coin over the telephone" and "Avoiding double spenders in digital money systems." Meaning, how robust cryptographic protocols can be used to avoid scams familiar to the most junior grifter. The theme of "proving anonymous betting schemes are not fixed" is in the same vein, and could establish the reputation of any budding Amos Fiats or David Chaums out there. --Tim May Views here are not the views of my Internet Service Provider or Government. ---------:---------:---------:---------:---------:---------:---------:---- 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

Re: politics aren't all or nothing
by John Young 17 Dec '03

17 Dec '03

Mark Hedges wrote: >Does anyone know who is behind this crap that the Empire's pushing on the >committees? Louis Freeh is the only name I've heard. I've not heard >Tenet's name mentioned in any of this. It would be interesting to research >the other people involved, to try to find some dirt. There's considerable push coming from domestic law enforcement agencies of many types and jurisdictions. See letters some of them have made in support of encryption controls in a House statement by Representative Weldon (co-sponsor with Dellums of the National Security Committee amendment to gut SAFE): http://jya.com/weldon.txt These letters are regularly cited Freeh to show that the FBI and other federal agencies are not alone in pushing for GAK, and to remind legislators that the law and order folks at home are a threat to re-election if they don't "give something to law enforcement." And, Freeh often reminds that half of wiretaps are state rather than federal, as shown in the 1997 annual report on electronic surveillance: http://jya.com/wire97.htm This is not to diminish the signficance of the burgeoning national security agencies, especially the 13 members of the intelligence community and those new-born aspirants for funds to protect the commonweal -- nearly every governmental jurisdiction now aims to grab a piece of the Nat Sec booty being converted to dual- and domestic-use, usually through various "justice" and "law enforcement" programs: http://jya.com/lec-links.htm An avid drumbeater for more funding for protection against the nation's high-tech enemies (thus bringing the foreign wars home) is the President's Commission on Critical Infrastructure Protection (PCCIP). See a recent news report on PCCIP's call for $1 billion to fight the homefront foe: http://jya.com/pccip.htm In short, there is a growing list of self-interested domestic parties who have a stake in GAK and you can bet they are leaning on their legislators just as hard as the federal briefers who whisper secrets and lies celebrated this week for 50 years of domestic deception in the national interest. Domestic controls on crypto and other munitions are just as certain to come as the export kind, and for the same reasons -- security of the state. If there are insufficient external enemies to justify stupendous public expenditures, why, just cook up the internal kind -- as done in the 1900s, the 1920s, the 1930s, the 40s, the 50s, the 60s, and to this day. The Bible Code foretells all.

1 0

Your faith is PGP is charming and quaint, but wrong
by Hroller Anonymous Remailer 17 Dec '03

17 Dec '03

All of these articles from people claiming to be me! I repeat, for the slow learners, that I consider PGP to be untrustworthy, at least on my machines. I don't want those Mounties to do to me what they did to Mulroney en passant. You in the southern colony are free to use PGP, but I prefer other means of establishing my uniqueness. Style, for example, cannot be easily copied. I am, as always, Alice de 'nonymous ... ...just another one of those... P.S. This post is in the public domain. I. A. M. A. M. O. R. O. N.

2 1

Re: GOST for sale
by mjsus＠atlanta.com 17 Dec '03

17 Dec '03

>> > I have non-commercial demo version of the product (370 Kb arjed > distributive) and I can send it to people who will be interested in it. > Please, send it to me. Thanks. Mark

1 0

Crypto-ban will pass Commerce cmte next week, from Reuters
by Declan McCullagh 17 Dec '03

17 Dec '03

Attached below is a Reuters article on how the House Commerce committee is about to approve an amendment banning the distribution of encryption products without a government backdoor. The Commerce committee would then join the Intelligence committee, which approved a similar measure last week. I generally agree with Aaron's (who wrote the article) take on the situation. Late this evening some lobbyists called me to promise some good news tomorrow, but momentum is clearly against them. After the full committee vote next Thursday, some sort of compromise legislation will likely go to the House floor -- at least if the leadership lets it. It's unclear whether that would happen this year or not. So much for the millions of dollars industry wasted on crypto-lobbying. The lobbyists (and high-tech executives) should have realized long ago that legislatures often are enemies of civil liberties. Civil liberties are anti-majoritarian, and usually unpopular, by their very nature. Look at what happened with the CDA -- it took the Supreme Court to undo Congress' bad work. Also, I recently spoke with Rep. Michael Oxley (R-Ohio), a former FBI agent, who's on the House Commerce committee that will consider SAFE next week. A version of his amendment that will be introduced is at: http://www.cdt.org/crypto/legis_105/SAFE/Oxley_Manton.html I wrote about crypto-compromises recently at: http://cgi.pathfinder.com/netly/editorial/0,1012,1412,00.html Attached is an excerpt from my conversation with Oxley... -Declan ***************** IS THERE A COMPROMISE IN THE WORKS? I can't give you any definitive answer on that. They have to recognize that our committee is prepared to accept my amendment. The Intelligence Committee has already passed an amendment unanimously. The National Security Committee passed an amendment 45-1. It's going to be the amendment in some form... The briefing that law enforcement and the FBI and DEA and the NSA and the CIA -- that was a pretty sobering briefing that the members got. At least it was in our committee. It brought up a lot of issues that hadn't been considered before... If it's a fight and it's up or down on my amendment, my amendment passes. Goodlatte told me that. He knew I had the votes. [Rep Goodlatte is the chief sponsor of the original SAFE bill] WOULD YOU TAKE SEN. KYL'S LEAD AND MOVE TO BAN NOT JUST THE DISTRIBUTION, BUT THE //USE// OF ENCRYPTION WITHOUT A KEY RECOVERY FEATURE? I've had some discussions with Kyl and we're working with him to find out what the best approach is. WOULD YOU INTRODUCE THE INTELLIGENCE DRAFT? That's still an option but we haven't made that determination yet. [The Intelligence committee bill is a slightly more extreme proposal.] I think the overwhelming sentiment in the House is towards protecting national security and providing law enforcement with the tools they need to combat this activity. This is a very complicated technological issue. My sense is that Bob Goodlatte and the folks who support that bill want other members to say we have to give up: we can't protect people from terrorists and organized crime since the technology has run. HAS YOUR EXPERIENCE AS AN FBI AGENT SENSITIZED YOU TO LOUIS FREEH'S ARGUMENTS? Oh clearly, sure. I used to tap phones for a living. That was old technology, but the concept is clearly the same. ----- Date: Thu, 18 Sep 1997 18:39:35 -0400 From: Aaron Pressman <aaron.pressman(a)reuters.com> To: declan(a)well.com Subject: reuters encryption story U.S. encryption bill inspires little give and take By Aaron Pressman WASHINGTON, Sept 18 (Reuter) - Like many who have tried to broach a compromise on the knotty issue of regulating computer encoding technology, members of the House Commerce Committee are finding their negotiation efforts futile, people involved said on Thursday. With talks deadlocked, the committee is likely to vote next week on virtually the same amendments that would have been considered previously, and most participants said a restrictive proposal favored by law-enforcement agencies would carry the day. An essential means of safeguarding communications and electronic commerce over the Internet, encryption products scramble information and render it unreadable without a password or software "key." Barring a surprise change in momentum, the committee will likely approve an amendment from Rep. Mike Oxley, Republican of Ohio, to impose restrictions for the first time on domestic sales and distribution of encryption products, staffers and lobbyists said. "He has the votes," said an industry lobbyist opposed to Oxley's amendment. "To call this negotiation is kind of a joke." Oxley's staff also conceded that talks were not proving fruitful. "We're negotiating with the other side, but I don't think there's a lot of middle ground here," Oxley spokeswoman Peggy Peterson said. One week ago, the committee postponed a vote on a bill by Virginia Republican Bob Goodlatte that was originally intended to relax strict U.S. export limits on encryption and prohibit mandatory backdoor access for the government. The Republican leadership gave the committee two weeks to find an approach that would be acceptable to law enforcement and national security agencies on one side, and software companies, civil libertarians and Internet users on the other. Rep. Rick White, Republican of Washington, and Rep. Ed Markey, Democrat of Massachusetts, had prepared an amendment requiring a study on backdoor access technologies but their proposal has not attracted much support from the FBI or the leadership of the committee. "Language can be changed but at the end of the day, it doesn't get law enforcement on board," one staffer said. FBI Director Louis Freeh has repeatedly told lawmakers to adopt restrictions forcing encryption manufacturers to include features allowing law enforcers to decode any message covertly. Freeh said without such a law, criminals and terrorists would increasingly use encryption to thwart FBI surveillance. Software companies counter that adding such features may be impossible and would reduce the level of security provided to legitimate citizens and businesses seeking to safeguard their communications. And civil libertarians said the FBI proposal would permit an Orwellian intrusion of government snooping into everyone's private affairs. Domestic restrictions like those in the Oxley amendment were added to the original bill by the Select Committee on Intelligence in a classified session last week. A few days earlier, the National Security Committee gutted the export relaxation provisions drafted by Goodlatte and replaced them with export limits tighter than current rules. Thursday, 18 September 1997 18:06:50 RTRS [nN18296706]

1 0

Re: Your faith is PGP is charming and quaint, but wrong
by Rich Salz 17 Dec '03

17 Dec '03

>Style, for example, cannot be easily copied. Au contrair. Cypherpunks share code... #! /bin/sh # This is a shell archive. Remove anything before this line, then feed it # into a shell via "sh file" or similar. To overwrite existing files, # type "sh file -c". # The tool that generated this appeared in the comp.sources.unix newsgroup; # send mail to comp-sources-unix(a)uunet.uu.net if you want that tool. # Contents: Makefile PATCHLEVEL README markov3.6 markov3.l # Wrapped by rsalz@sulphur on Thu Nov 30 00:04:29 1995 PATH=/bin:/usr/bin:/usr/ucb ; export PATH echo If this archive is complete, you will see the following message: echo ' "shar: End of archive."' if test -f 'Makefile' -a "${1}" != "-c" ; then echo shar: Will not clobber existing file \"'Makefile'\" else echo shar: Extracting \"'Makefile'\" $53 characters$ sed "s/^X//" >'Makefile' <<'END_OF_FILE' X Xmarkov3: markov3.o X $(CC) -o $@ $(CFLAGS) markov3.o END_OF_FILE if test 53 -ne `wc -c <'Makefile'`; then echo shar: \"'Makefile'\" unpacked with wrong size! fi # end of 'Makefile' fi if test -f 'PATCHLEVEL' -a "${1}" != "-c" ; then echo shar: Will not clobber existing file \"'PATCHLEVEL'\" else echo shar: Extracting \"'PATCHLEVEL'\" $2 characters$ sed "s/^X//" >'PATCHLEVEL' <<'END_OF_FILE' X1 END_OF_FILE if test 2 -ne `wc -c <'PATCHLEVEL'`; then echo shar: \"'PATCHLEVEL'\" unpacked with wrong size! fi # end of 'PATCHLEVEL' fi if test -f 'README' -a "${1}" != "-c" ; then echo shar: Will not clobber existing file \"'README'\" else echo shar: Extracting \"'README'\" $2400 characters$ sed "s/^X//" >'README' <<'END_OF_FILE' XThis is a cleaned-up reposting of the markov3 program. The following Xchanges have been made: X XThe null pointer dereferencing bugs have been fixed (I hope). X XThe code that uses "rand" should now be portable (the patches posted Xto the net to fix this problem were wrong, they break the code on Xsome machines in order to fix it on others. I stole some code from X"hack" to do things right. If hack works for you, this should). X Xmarkov3 now understands "notes" cruft (thanks to Rich Salz). X XBecause of the 50% rule in news 2.11, people often use some other Xcharacter than ">" for inclusions. markov3 assumes that lines Xbeginning with any of X X > < ) | # } ] X Xare inclusions (without this rule, funny-looking output results if Xanyone uses non-standard "quoting"). X XThe random number generator is initialized using the time, if neither Xthe -s flag nor the new -x flag is given. X XThis will be the last complete posting; a "patchlevel" file is included Xand I will send out patches if there are further bugs or improvements. X XHere's the original README. X--------------------------------------------------------------------------- XI created a bit of a stir with this program in December 1986 when I Xused an earlier version of it to simulate a certain well-known net Xpersonality (Hi Laura!). It digests Usenet articles and spits out Xother articles with similar characteristics. You need lex to run it, Xbut otherwise it should run on any Unix I know of. X XI had several requests for the program but didn't consider it X"ready". It's as ready as it will ever be now. X XThe program uses getopt(3). There are several public-domain versions Xavailable for Berkeley systems from the mod.sources archives. Since Xit's small, I've included Henry Spencer's version, but you'll have Xto change the Makefile to use it. X XFor best results, feed it at least ten articles by the same person Xor on the same subject. If there are fewer articles the output Xresembles the original too much; if there is too much variety in Xthe articles the output is more incoherent than it otherwise is. X XThe program requires lots of memory if it is given lots of input; Xthe small-model people will have problems. X XPlease don't post the output to the net (though I'd be happy to Xsee some of the more interesting results). X XSend comments, suggestions for improvement, fan mail, and flames Xto me: {sun,hplabs,ames,ihnp4}!oliveb!epimass!jbuck. END_OF_FILE if test 2400 -ne `wc -c <'README'`; then echo shar: \"'README'\" unpacked with wrong size! fi # end of 'README' fi if test -f 'markov3.6' -a "${1}" != "-c" ; then echo shar: Will not clobber existing file \"'markov3.6'\" else echo shar: Extracting \"'markov3.6'\" $2503 characters$ sed "s/^X//" >'markov3.6' <<'END_OF_FILE' X.\" markov3 X.\" @(#)markov3.6 1.1 3/6/87 epimass!jbuck X.TH MARKOV3 6 "3/6/87" X.UC 4 X.SH NAME Xmarkov3 \- Digest and spit out quasi-random Usenet articles X.SH SYNOPSIS X.B markov3 X[ X.B \-pv X] [ X.B \-n X.I n_articles X] [ X.B \-d X.I dumpfile X] [ X.B \-s X.I seed X] [ X.B \-x X] Xfiles X.SH DESCRIPTION X.PP X.I Markov3 Xdigests Usenet articles and builds an internal data structure that Xmodels the articles as if they came from a random process, where Xeach word is determined by the previous two. It then emits a series Xof articles on the standard output that have the same distribution Xof words, word pairs, and word triplets as do the input files. XThe name X.I markov3 Xcomes from the fact that this structure is called a Markov chain, Xand that the statistics for word triplets are modeled. XHere, a "word" is a sequence of printable characters surrounded by Xwhitespace. Paragraph breaks (blank lines) are also treated as a X"word". Paragraphs of included text are treated as single "words" Xand printed as "> ...". X.PP XBy default, the program expects to be fed Usenet articles; it strips Xoff headers, included text, and signatures (or at least it tries). XThe X.B \-p X(plain) option disables the header-stripping feature (otherwise Xeverything is skipped until a blank line is encountered). X.PP XBy default, 10 articles, separated by form feeds, are written on the Xstandard output. The X.B \-n Xoption lets you specify a different number. X.PP XThe X.B \-x Xoption does not seed the random number generator; this is useful Xfor simulating people who repeat themselves. X.PP XThe X.B \-d X(dump) option dumps a representation of the internal data structure Xbuilt by X.I markov3 Xon the named file. X.PP XFinally, the X.B \-v X(verbose) Xoption prints some statistics on the standard error. X.SH "CAVEATS" XThis program allocates lots of memory if given large amounts of input. XOn virtual memory systems, the paging behavior is atrocious because Xpointers tend to point every which way, and many pointers are dereferenced Xfor every word processed. This could be improved, I'm sure. X.PP XPosting articles generated by X.I markov3 Xto the net may be hazardous to your health. X.PP XNot as smart as Mark V. Shaney. X.SH "PORTABILITY" XAn effort has been made to make this program as portable as possible; Xan earlier version was much less portable because of problems with Xnull pointers and rand(3). Please let me know if you have further problems. X.PP XIf you don't have lex, you'll need to rewrite the lexical analyzer Xbut most of the program is in C. END_OF_FILE if test 2503 -ne `wc -c <'markov3.6'`; then echo shar: \"'markov3.6'\" unpacked with wrong size! fi # end of 'markov3.6' fi if test -f 'markov3.l' -a "${1}" != "-c" ; then echo shar: Will not clobber existing file \"'markov3.l'\" else echo shar: Extracting \"'markov3.l'\" $11822 characters$ sed "s/^X//" >'markov3.l' <<'END_OF_FILE' X%{ X/* X * Copyright (c) 1986, 1987 by Joe Buck X * X * Permission is granted to use, redistribute, or modify this program, X * as long as you don't pretend you wrote it. Send improvements or X * bug reports to {ihnp4,hplabs,ames,sun}!oliveb!epimass!jbuck. X * X * The program generates simulated Usenet articles, given Usenet articles X * as input. X * X * This program constructs a table of frequencies for a token appearing, X * given the two preceding tokens. A "token" is a sequence of non-blank X * characters. An entirely blank line is also treated as a token, as is X * the beginning and end of an article. X * X * The program is designed to process news articles, rejecting text from X * the header, signature, and included text, together with cruft inserted X * by rn and notes. A paragraph of included text is treated like a token. X * X * After the table is built (and it can be big), articles are generated X * on the standard output. X */ X#ifndef lint Xstatic char *sccs_id = "@(#)markov3.l 1.1 3/6/87 epimass!jbuck"; X#endif X#include <sys/types.h> /* for time_t */ Xint in_included_text = 0; X#ifdef yywrap X#undef yywrap X#endif X%} X%Start HDR BODY SIG X%% X<HDR>^[^ \t]+:.*\n ; /* Header line, e.g. "From: foo(a)bar.UUCP" */ X<HDR>^[ \t]+[^ \t].*\n ; /* Continuation of header line */ X<HDR>^[ \t]*$ BEGIN BODY; X<BODY>^"-- "$ BEGIN SIG; X<BODY>^[><)|#}].*\n { /* 50% rule gets people to change ">" X to other characters; this gets most of them */ X if (!in_included_text) { X in_included_text = 1; X process_token ("\n> ...\n\n"); X } X } X<BODY>"]".*\n { /* should have been included in the above. My X lex generates bad C code if I say [[><...] X even though ed(1) says that's a valid regular X expression. */ X if (!in_included_text) { X in_included_text = 1; X process_token ("\n> ...\n\n"); X } X } X<BODY>^"In article".*\n ; /* Reject rn crud */ X<BODY>^"/* Written".*"*/"\n ; /* Also NOTES crud */ X<BODY>^"/* End of text from".*"*/"\n ; /* NOTES */ X<BODY>^"/* ----------".*"----------*/"\n ; /* NOTES */ X<BODY>[ \t]+ ; /* Skip white space */ X<BODY>\n[ \t\n]*\n { process_token ("\n"); /* Paragraph break */} X<BODY>^\..* ; /* Ignore format directives. */ X<BODY>[^ \t\n]+ { in_included_text = 0; process_token (yytext); } X<HDR>. ; /* Eat anything that escaped */ X<HDR>\n ; X<BODY>\n ; X<SIG>. ; X<SIG>\n ; X%% Xextern int optind; Xextern char *optarg; X X/* X * hashtab is a hash table storing all the tokens we encounter. X */ Xstruct htentry { X char *htext; X struct htentry *hlink; X}; X X#define HSIZE 3557 /* Should be prime */ X#define Fprintf (void)fprintf X#define Printf (void)printf X Xstruct htentry hashtab[HSIZE]; X X/* X * node and succnode are portions of the big structure we're going to build. X * node represents something like ("was", "a") in a binary tree. X * a linked list of succnodes contain tokens that may follow ("was", "a") X */ Xstruct node { X char *text; X char *text2; X int ocount; X struct node *lc, *rc; X struct succnode *succ; X}; X Xstruct succnode { X struct node *scnod; X int count; X struct succnode *link; X}; X X Xstruct node *prev_code = NULL; Xchar *prev_token = NULL, **Argv; Xint init_state = HDR; Xint verbose = 0; Xstruct node *root = NULL, *tknptr; Xstruct succnode *start = NULL; Xint n_pairs = 0, n_tokens = 0, n_files = 0, n_total = 0; X Xstruct node *insert_token(); Xchar *savetoken(); X Xprocess_token (txt) Xchar *txt; X{ X struct node *code; X char *token = savetoken (txt); X/* We have a new token. Say the previous two tokens were "one" "way" X * and the current token is "to". Then prev_code points to a node X * for ("one", "way") and token is "to". This function adds ("way", "to") as a X * successor to ("one","way") and makes prev_code point to ("way","to"). X */ X code = insert_token (prev_token, token); X insert_pair (prev_code, code); X prev_code = code; X prev_token = token; X return; X} X X/* X * here it is, the main function. X */ Xmain (argc, argv) Xint argc; Xchar **argv; X{ X int i, c, n_articles = 10, sflag = 0; X char *dumpfile = NULL; X extern int optind; X extern char *optarg; X X while ((c = getopt (argc, argv, "pxvn:d:s:")) != EOF) { X switch (c) { X case 'v': X verbose = 1; X break; X case 'p': /* Input is plain text, not Usenet stuff */ X init_state = BODY; X break; X case 'n': /* # articles to generate */ X n_articles = atoi (optarg); X break; X case 'd': /* where to dump the data structure */ X dumpfile = optarg; X break; X case 's': /* Set the seed for rand; fall through */ X srand (atoi (optarg)); X case 'x': /* set flag to prevent srand */ X sflag++; X break; X default: X Fprintf (stderr, X "Usage: markov3 [-pvx] [-s seed] [-n n_art] [-d dump] files\n"); X exit (1); X } X } X BEGIN init_state; /* initial state of lexical analyzer */ X if (!sflag) /* set random number generator */ X srand ((int)time ((time_t *)0)); X/* Note: if optind == argc, there are no file arguments. yyin is left X * initialized to stdin. X */ X if (optind < argc) { X/* yyin is lex input stream. Point to first file. */ X if ((yyin = fopen (argv[optind], "r")) == NULL) { X perror (argv[optind]); X exit (1); X } X optind++; /* skip to next file */ X } X Argv = argv; /* make it global so yywrap can access it */ X n_files = 1; X/* yylex puts all the input files through the lexical analyzer and builds X * the database. X */ X (void) yylex (); X if (dumpfile) X dump_database (dumpfile); X if (verbose) X Fprintf (stderr, X "Total of %d tokens (%d different), %d different pairs, %d files\n", X n_total, n_tokens, n_pairs, n_files); X/* Generate the articles, separated by form feeds */ X for (i = 0; i < n_articles; i++) { X if (i > 0) output_word ("\n\f\n"); X generate_article (); X } X return 0; X} X X/* X * Lex calls this when EOF is reached. It opens the next file if there X * is one. Lex interprets a return value of 1 to mean "all done" and 0 X * to mean "keep going". X */ Xyywrap () { X (void) fclose (yyin); X insert_pair (prev_code, (struct node *)0); X prev_code = NULL; X if (Argv[optind] == NULL) return 1; X else if ((yyin = fopen (Argv[optind], "r")) == NULL) { X perror (Argv[optind]); X exit (1); X } X optind++; X in_included_text = 0; X if (verbose && n_files % 10 == 0) X Fprintf (stderr, "%d files\n", n_files); X n_files++; X BEGIN init_state; X return 0; X} X X/* X * This function saves a token in the hash table (if it isn't there X * already) and returns a pointer to the stored copy. X */ Xchar * Xsavetoken (txt) Xchar *txt; X{ X int h; X char *p; X struct htentry *hp; X X n_total++; X for (p = txt, h = 0; *p; h += *p++); X hp = hashtab + (h % HSIZE); X while (hp->hlink) { X if (strcmp (hp->htext, txt) == 0) { X return hp->htext; X } X hp = hp->hlink; X } X/* OK, it's a new token. Make hp->hlink point to a new, X * null block and make hp->htext point to the text. X */ X hp->hlink = (struct htentry *) malloc (sizeof *hp); X hp->htext = malloc ((unsigned)(strlen (txt) + 1)); X (void) strcpy (hp->htext, txt); X hp->hlink->hlink = NULL; X hp->hlink->htext = NULL; X n_tokens++; X return hp->htext; X} X X/* X * This recursive function inserts a token pair into the tree. X */ Xstruct node * Xinsert_in_tree (p, txt, txt2) Xstruct node *p; Xchar *txt, *txt2; X{ X int cmp; X if (p == NULL) { X/* Create a new node. */ X p = (struct node *) malloc (sizeof *p); X p->text = txt; X p->text2 = txt2; X p->lc = p->rc = NULL; X p->succ = NULL; X p->ocount = 1; X tknptr = p; X n_pairs++; X if (verbose && n_pairs % 1000 == 0) X Fprintf (stderr, "%d pairs\n", n_pairs); X return p; X } X cmp = my_strcmp (p->text, txt); X if (cmp == 0) cmp = my_strcmp (p->text2, txt2); X if (cmp == 0) { X/* It's a match. Increment the count. */ X tknptr = p; X p->ocount += 1; X } X/* Look in the subtrees. */ X else if (cmp < 0) p->lc = insert_in_tree (p->lc, txt, txt2); X else p->rc = insert_in_tree (p->rc, txt, txt2); X return p; X} X X/* X * This just calls insert_in_tree starting at the root X */ Xstruct node * Xinsert_token (txt, txt2) Xchar *txt,*txt2; X{ X root = insert_in_tree (root, txt, txt2); X return tknptr; X} X X/* X * This function adds a successor. X */ Xstruct succnode * Xinsert_in_succ_chain (sp, np) Xstruct succnode *sp; Xstruct node *np; X{ X if (sp == NULL) { X sp = (struct succnode *) malloc (sizeof *sp); X sp->scnod = np; X sp->count = 1; X sp->link = NULL; X } X else if (sp->scnod == np) X sp->count += 1; X else sp->link = insert_in_succ_chain (sp->link, np); X return sp; X} X X/* X * This calls insert_in_succ_chain starting at the right place. X */ Xinsert_pair (p1, p2) Xstruct node *p1, *p2; X{ X if (p1) p1->succ = insert_in_succ_chain (p1->succ, p2); X else start = insert_in_succ_chain (start, p2); X} X X/* X * This function dumps the stored data structure onto a file. X * Now if only I had a function to read it back in. X */ Xchar * Xpr_token (txt) Xchar *txt; X{ X if (txt[0] != '\n') X return txt; X return txt[1] ? "<INCL>" : "<LF>"; X} X Xtreedump (tree, fp) Xstruct node *tree; XFILE *fp; X{ X if (tree) { X treedump (tree->rc, fp); X Fprintf (fp, "( %s %s ) %d", pr_token (tree->text), X pr_token (tree->text2), tree->ocount); X chaindump (tree->succ, fp); X treedump (tree->lc, fp); X } X} X X/* X * Subroutine of treedump; it does one row. X */ Xchaindump (p, fp) Xstruct succnode *p; XFILE *fp; X{ X char *text; X while (p) { X if (p->scnod == NULL) X text = "<EOF>"; X else text = pr_token (p->scnod->text2); X Fprintf (fp, " %s %d", text, p->count); X p = p->link; X } X putc ('\n', fp); X} X X/* X * This routine generates the dump file (-d option) X */ Xdump_database (file) Xchar *file; X{ X FILE *fp = fopen (file, "w"); X if (fp == NULL) { X Fprintf (stderr, "markov: can't open "); X perror (file); X exit (1); X } X Fprintf (fp, "START:"); X chaindump (start, fp); X treedump (root, fp); X} X X/* roll (n) generates a uniformly distributed rv between 0 and n-1. X * This code is stolen from "hack" and should be portable. If you X * change this, remember that different systems have rand functions X * with different ranges, and the bottom bits are often no good. X */ X#define roll(n) ((rand() >> 3) % n) X X/* X * This function generates an article by traversing the X * structure we've built. X */ Xgenerate_article () { X struct succnode *p = start; X int ncounts = n_files; X int n, accum; X char *tp; X X while (1) { X/* Roll the dice to find out the next token. The code below selects the X * next token, and the new state, with a probability corresponding to the X * frequency in the input. X */ X n = roll (ncounts); X accum = p->count; X while (accum <= n && p->link) { X p = p->link; X accum += p->count; X } X if (p->scnod == NULL) X break; X tp = p->scnod->text2; X/* Check for "end of story" */ X if (tp == NULL) X break; X output_word (tp); X ncounts = p->scnod->ocount; X p = p->scnod->succ; X } X output_word ("\n"); /* This will flush the buffer as well. */ X return; X} X X/* X * This version handles null strings * X */ Xmy_strcmp (a, b) Xregister char *a, *b; X{ X if (a == NULL) return b ? -1 : 0; X if (b == NULL) return 1; X return strcmp (a, b); X} X X#define LEN 75 Xoutput_word (word) Xchar *word; X{ X static char line[LEN+1]; X static int room = LEN; X int l; X X if (word == NULL) return; X l = strlen (word); X/* If word won't fit, or starts with \n, dump the current line */ X if ((l >= room || word[0] == '\n') && line[0]) { X Printf ("%s\n", line); X line[0] = 0; X room = LEN; X } X/* If word won't fit in the buffer or starts with \n, print it now */ X if (l >= LEN) X Printf ("%s\n", word); X else if (word[0] == '\n') X Printf ("%s", word); X/* Otherwise fill it in */ X else { X (void)strcat (line, word); X (void)strcat (line, " "); X room -= (l + 1); X } X return; X} END_OF_FILE if test 11822 -ne `wc -c <'markov3.l'`; then echo shar: \"'markov3.l'\" unpacked with wrong size! fi # end of 'markov3.l' fi echo shar: End of archive. exit 0

1 0