December 2003 - cypherpunks-legacy

Idiot's Guide to News via Compuserve
by Duncan Frissell 17 Dec '03

17 Dec '03

Idiots Guide to Reading Banned Newsgroups via CompuServe CompuServe (alone among the Big Three) gives all subscribers a PPP connection to the Net through any CIS node (including those in Germany). This is a real Net connection that makes it possible to use CIS to fully access all the Net's resources including "banned" newsgroups. A Point-to-Point Protocol (PPP) connection makes your computer just another machine on the Net for the time you are connected. Anything that any Net-connected machine can do, you can do. The CompuServe Dialer (formerly CompuServe Internet Dialer) software that has been included with WINCIM at least since version 1.4 gives you a PPP connection but you need additional newsreading software to dodge CompuServe restrictions. So for real beginners, here is how you can use your CompuServe account to access banned newsgroups as well as the rest of the Internet. These instructions are for Windows users. MAC users can do the same things but I don't know the available MAC software. 1) Get the latest copy of WINCIM. GO WINCIM (hit Ctrl-G and type WINCIM in the dialog box) from within CIS. Download is free. Install it according to the instructions and make sure that you can log on to CIS. 2) Next, let's get a copy of Freeagent which is a free usenet newsreader which works with CompuServe's own Internet connection software. Log on to CIS. GO FTP or: Click the Internet icon, click the File Downloads (FTP) icon and proceed to the main FTP screen. Click the Access a Specific Site button. 3) Enter ftp.forteinc.com in the Site Name box and /pub/free_agent in the Directory box. Click OK and you should see some site login information. Click OK again and you should see a check box next to the file name fagent10.zip. Click the check box and then click the Retrieve button. Wincim will show you that it will save the file in the Compuserv\download directory on your hard drive. (Remember where it's going.) 744279 bytes later, you will have a copy of Freeagent. 4) Move the file fagent.zip into a directory by itself (C:\AGENT for example). Unzip it with Pkunzip or one of the many zip utilities available on CIS. 5) What you are going to do is to connect directly to the Internet via CompuServe and read usenet newsgroups using Freeagent. In order to do that, you will need to find a site somewhere on the Net that will let you read Usenet News for free or you will have to obtain an account on another Internet-connected machine. 6) Community ConneXion (c2.org) will give a month of free service to CompuServe members suffering from censorship. For information send email to uncensored(a)c2.org. After your free month, c2.org costs only $7.50 a month for accounts accessed via the Internet. 7) There are some news servers out there that are open to the public. A news server is just a machine connected to the Internet that stores and forwards Usenet news. The IP addresses of two of the open news servers are 198.70.185.5 and 205.139.39.1. Once you become a sophisticated user of the nets, you can use your Web Browser to pick up a longer list of open news servers at http://dana.ucc.nau.edu/~jwa/open-sites.html. 8) Now back to Free Agent to grab those banned newsgroups. In your WINCIM directory (maybe in a subdirectory called \cid), click cid.exe to start CompuServe (Internet) Dialer or find the phone-shaped dialer icon in your CompuServe group window. Inside the Dialer, hit dial to log on to the Internet. 9) Start Free Agent. Click Accept to accept the license agreement. It will prompt you to enter various information including the address of the news server you want to use, your email address, and other info. All that you have to enter is the address of a news server. In the box labeled "News (NNTP) Server:" enter one of the IP addresses of open news servers (198.70.185.5, 205.139.39.1, or another from the list) or the address of the news server of a system you've opened an account on (news.c2.org, for example). 10) Freeagent will ask you if it's OK to retrieve a list of news groups from the server. Click Yes. Free agent will then tell you that it is Retrieving complete List of Groups. Once that's finished you'll be ready to read all the newsgroups you want and neither CompuServe nor the Bavarian prosecutor will have anything to say about it. "The Internet belongs to no one except its users."

1 0

Pi Stuff
by mpd＠netcom.com 17 Dec '03

17 Dec '03

Various amazed people on the Pi thread wrote: > But BTW, isn't it interesting, that news item from a few > weeks ago, on an algorithm for determining individual bits > in Pi, regardless of whether you've calculated all the > previous ones. Only problem is, it only works in > hexadecimal (and, obviously, binary, etc, not decimal. > I didn't see this result you mention, but it surprises me. > Yeah, well, I understand your frustration, but to me the > really amazing part is that the elephant flies at all, and > not how well he flies (<G>) (In other words, the amazing > thing is that there is a predictable relationship in ANY > base system) Chances are the thing WON'T work in base > 3,5,6,7,9,10,11,12,13,14,15, and any other non 2**n base. > Or, at least, it will take a DIFFERENT equation to work in > those other bases, if they work at all. A few quick comments. The notion that one might be able to compute digits of Pi efficiently at any starting point in the number is not late-breaking news. The Chudnovsky brothers developed a formula which permitted them to do this a number of years ago, and used it to compute Pi to several billion digits. Prior to that time, the Borweins' quartic interation based on Ramanujan's modular identities and AGM techniques was the existing state of the art. Given a base, d, and an finite ordinal, i, the function which computes the ith digit of the Pi in the base d is certainly a computable one. If we can find an algorithm for computing this function whose time as a function of "i" does not include the time required to compute all previous digits, then we are to a certain extent evaluating individual digits of Pi without calculating the previous ones. One should keep in mind, however, that the degree to which such things are true for algorithms lies on a continuum, with a near-constant number of arithmetic operations at one end, and a geometric progression at the other. So it is not the classic either/or situation. Good algorithms whose times are tractable functions of "i" are certainly desirable, and have been discovered. Someone suggested that radix conversion was a time-consuming operation. Modern FFT-based algorithms can do multiplication, division, Nth root, reciprocal, and base conversion in near-linear time. Doubtless good algorithms to compute the Nth digit of Pi in any base do exist, but their form may not be as obvious as those for trivial bases, such as powers of two. -- Mike Duvos $ PGP 2.6 Public Key available $ mpd(a)netcom.com $ via Finger. $

1 0

Re: Massey, CEO of Compuserve, on Internet
by Robert A. Rosenberg 17 Dec '03

17 Dec '03

At 13:55 1/4/96, Tony Iannotti wrote: >On Thu, 4 Jan 1996, Robert A. Rosenberg wrote: > >> Yes it would require that the Node be checked in the Software. What I was >> responding to was a claim that there is no way of telling where I am >> connecting from (which I disproved). As to calling a non-German Node, that >> is always an option. > >Yes, I agree. I think the real difference is that they really cannot tell >where you are calling from, even though they know where you are >connecting. Since the German Government (or the local "DA" who is claiming to represent the National Government <g>) is talking about delivery of banned items in Germany, I think that the relevant location is the node that is being used not where the other end of the call to that node is located. What is being requested is that CIS refrain for delivering the stuff to the Nodes in Germany (they are not being told to monitor someone in Germany who is trying to get it by calling LD to a node in France (or the US for that matter).

1 0

An open letter to Commtouch
by Raph Levien 17 Dec '03

17 Dec '03

Hi Commtouch people, I am intrigued and hopeful about your secure e-mail product, Pronto Secure. However, I am puzzled about its support for POTP encryption. The other encryption protocols (PGP, PEM, MOSS, and S/MIME) have all been reviewed carefully by outside experts, and there is general consensus that these protocols embody state-of-the-art cryptographic technology, and that there are no known major security flaws. POTP stands out on your list because such a review has not been carried out. In fact, grave doubts have been raised regarding its security, and (to my taste, anyway) not satisfactorily answered. I do not wish to raise those points here, nor do I wish to claim here that POTP is insecure. However, I believe the reputation of your product is drawn into question by association. Should POTP be definitively demonstrated to be weak, then it would not be the case that using your product according to the instructions would provide "security." Further, I would consider it slightly misleading to describe it as "mission-critical." I feel the situation is analogous to that of a hypothetical networking company claiming that their product delivers high bandwidth by offering the choice of ATM, Myrinet, 100Mbps Ethernet, or string and tin cans. That said, I applaud your multiprotocol approach in general. In fact, I feel it is the future of Internet security tools. I hope your product gains widespread acceptance, and helps to further the cause of deployment of strong crypto. Raph Levien

1 0

No Subject
by Marc J. Wohler 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- Anyone know of the staus of the Phil Zimmermann Grand Jury investagation? Is there a Statuate of Limitations or what? -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAwUBMO2XK2eikzgqLB7pAQE6wgQAg7bs/b9qkayCQSUz8Ql8vcVNF74cSAl+ TRs/HQWA8g99x5j0Boircf7kpO8LR8orFkHdsSApYWCOCTDTHpiQNmcLYcSbTiFn td3cJsponuKLa0qTjRpo3e+yIo8ebDC1tCrJaFXTZD8hHGLchr+rz8CLJR8R7pLS LxYr3TFMM3s= =x1Al -----END PGP SIGNATURE----- ***Preserve, Protect and Defend the private use of Strong Crypto*** * * * PGP for the masses * * * Finger mjwohler(a)netcom.com for Marc Wohler's public key fingerprint= F1 70 23 13 91 B5 10 63 0F CF 33 AD BE E6 7B B6

1 0

Re: FreeBSD user <fb@sponsor.octet.com>
by Steve14571＠aol.com 17 Dec '03

17 Dec '03

In a message dated 96-01-04 18:44:46 EST, you write: > Am I the only one who has received three unreadable >messages from this address on cypherpunks? No, I have them too... What is going on? Pvt Stephen Herbert United States Marine Corps

1 0

Re: New Mitnick Book
by John Young 17 Dec '03

17 Dec '03

Responding to msg by bdavis(a)thepoint.net (Brian Davis) on Fri, 5 Jan 3:4 PM Here's Littman on what the Feds knew about Markoff at Raleigh (Orsak and Murphy are with Sprint Cellular; Kent Walker was US Attorney in San Francisco): Monday afternoon, Special Agent LeVord Burns sits by the coffee pot and vending machine at the Sprint switch and debates the legal issues with Shimomura. "Tsutomu wanted us to kick his door down," recalls Orsak, who along with Murphy, listened in. "Burns was talking about what warrants had been issued, what the FBI was going to do." Burns impresses Orsak. A well-built, bespectacled black man in a suit and tie, Burns looks like the kind of FBI agent that doesn't miss details. As Burns recounts Mitnick's background, Orsak is surprised by what the agent says about Mitnick. "Burns said there were a lot of guys that as far as national security went were a lot more dangerous than Mitnick -- that a lot of professional hackers are a lot more dangerous." To Orsak, cyberspace's Most Wanted Hacker doesn't sound all that threatening. "One of the more interesting things, I thought, was the FBI goes, 'As far as hackers go,' Mitnick was 'benign.' They didn't have evidence he was in it for the money." A little later, John Markoff and Shimomura's girlfriend, Julia Menapace, who just flew in, arrive at the switch. Orsak and Murphy invite Shimomura's team, Burns, and two other FBI agents from Quantico, Virginia, out to Ragazzi's, a casual Italian restaurant nearby. Orsak spreads out a Raleigh street plan on the checkered tablecloth and pinpoints Mitnick's location. "LeVord was telling us what his involvement was for the FBI," recalls Murphy. "It was light banter. LeVord assumed like we all did, that Markoff was just another guy out of California. Just another egghead. One of Tsutomu's." Markoff gets everyone's ear when he mentions Mitnick inspired the hit movie WarGames. "Markoff was filling us in on Mitnick's typical behavior, the different people Mitnick had run-ins with," recalls Murphy. "A guy in England, a guy in Princeton, one at Digital." Then, Markoff runs through some of Mitnick's aliases. One of the phony names rings a bell with Murphy. After dinner, the whole crew heads back to the switch, and just as Murphy suspected, he finds a memo describing a recent attempt by someone using the alias to social engineer a new bunch of MINs. Meanwhile, the FBI is bumping up against a technical problem. The agents had planned to install the FBI's own bulky scanning equipment in a rental van, but they can't find one. Murphy suggests using his co-worker Fred's minivan. Burns gives the idea the green light, and Orsak helps the agents set up and calibrate their equipment in Fred's van. Around midnight, Fred chauffeurs the two agents to circle the cell site to calibrate their scanning equipment. Fred and the FBI agents get to talking. "He [Fred] let the cat out of the bag," confides Murphy. "We didn't tell him not to say anything. We weren't trying to hide it, but we were also not trying to convey it. He told them Markoff wrote a book on this guy." The boys from Quantico aren't happy. "They freaked," recalls Murphy. "They thought Markoff would tip the guy [Mitnick] so he could write another book." One of the Quantico agents phones the Sprint switch to confirm Markoff's identity. "Me, Markoff, Tsutomu, and Julia were at the switch," remembers Murphy. "One of the Quantico guys was on the phone. He wanted to talk to Tsutomu." Murphy passes the phone to Shimomura. "He [Shimomura] wasn't about to lie," says Murphy of the tense moment. "He [Shimomura] was trying to evade a little bit. He said that Kent Walker knew about Markoff being there, which of course Walker did." Murphy, Markoff, and Menapace listen to Shimomura. "Kent knows about it," insists Shimomura to the agent from Quantico. "He's cleared through Kent." But Kent Walker later denied ever giving Shimomura such approval or knowing John Markoff was in Raleigh. Shimomura later disputed Murphy's account and said he "never told anyone from law enforcement that anyone had authorized Markoff's presence in Raleigh." (pp. 357-58)

1 0

More Noise Sphere Noise (simple source code)
by wlkngowl＠unix.asb.com 17 Dec '03

17 Dec '03

Ok, no language holy wars. This was quickie to test out the Noise Sphere plotting. It's in Pascal, but it's understandable. { Simple demo of a Noise Sphere in Turbo Pascal } { (If only I had a really awful RNG to test it with...) } program NoiseSphere; uses Graph, Crt; const BGIPath = ''; { where those silly Borland *.BGI drivers are } var GraphMode, GraphDriver: Integer; type Polar = record r, theta, phi: Real; end; Cartesian = record x,y,z: Real; end; procedure PolarToCartesian(var P: Polar; var C: Cartesian); begin C.x := P.r * Sin(P.phi) * Cos(P.theta); C.y := P.r * Sin(P.phi) * Sin(P.theta); C.z := P.r * Cos(P.phi); end; procedure Plot(var C: Cartesian); begin with C do begin PutPixel(100+(Round(100*y)), 200-(Round(120*z)), Yellow); PutPixel(320+(Round(100*x)), 200-(Round(120*y)), Red); PutPixel(540+(Round(100*x)), 200-(Round(120*z)), Blue); end; Delay(1); end; function ByteToReal(b: Byte): Real; begin ByteToReal := b / 256; end; function InitScreen: Integer; begin GraphMode := VGAHi; GraphDriver := EGA; InitGraph(GraphDriver,GraphMode,BGIPath); InitScreen := GraphResult; end; var n: LongInt; X: Array [ 0..2 ] of Real; P: Polar; C: Cartesian; begin InitScreen; Randomize; for n := 0 to 2 do X[n] := {$ifdef USEDEV} {$else} ByteToReal(Random(256)); {$endif} n := 0; repeat with P do begin r := Sqrt(X[(n+2) mod 3]); theta := pi * X[(n+1) mod 3]; phi := 2 * pi * X[n]; end; PolarToCartesian(P,C); Plot(C); X[n] := {$ifdef USEDEV} {$else} ByteToReal(Random(256)); {$endif} n := (n + 1) mod 3; until KeyPressed; ReadKey; RestoreCrtMode; end.

1 0

Mixmaster In a Box
by Adam Shostack 17 Dec '03

17 Dec '03

Towords the goal of making Mixmasters in a box, I've written an installer script for mixmaster. If you're running on one of the supported platforms (alpha, bsdi, hpux, linux, sunos, solaris), the script will walk you through everything from the make to setting up cron jobs & /etc/aliases. If you've been putting off setting up a remailer because its a pain, give this a shot. Lance will probably be including it in the next release of mixmaster, but you can get it now by sending me a message with the Subject: get mix-installer. Comments, bugs, bug fixes are welcome. Thanks to Rich $alz for a extensive comments & suggestions for portability. Adam -- "It is seldom that liberty of any kind is lost all at once." -Hume

1 0

Re: More Noise Sphere Noise (simple source code)
by wlkngowl＠unix.asb.com 17 Dec '03

17 Dec '03

On Fri, 05 Jan 96 07:24:03, I wrote: >Ok, no language holy wars. This was quickie to test out >the Noise Sphere plotting. It's in Pascal, but it's >understandable. >{ Simple demo of a Noise Sphere in Turbo Pascal } >{ (If only I had a really awful RNG to test it with...) } Odd... That came through on the mail/news gateway but the preceeding article that explains what that was about didn't.... did anyone receive it? In case no, it's based on Clifford A. Pickover's paper "Random Number Generators: Pretty Good Ones are Easy to Find" which doesn't deal with useful RNGs for crypto sense, but the paper does explain ways to visually represent RNGs so that seemingly good RNGs show their awful correlations, etc. My previous post that should have showed up mentioned that these methods might be useful for checking out crypto functions (hashes, pRNGs, ciphers). The full reference is The Visual Computer (1995) 11:369-377, Springer Verlag, 1995.

1 0