December 2003 - cypherpunks-legacy

Re: RSA's Art gallery is now on the web
by Alan Olsen 17 Dec '03

17 Dec '03

At 09:57 AM 1/5/96 PST, you wrote: > Are you tired of the clip art you downloaded last week? Do you >need a new background screen for you PC? Does your cubicle need something >to give it character? > Look no further. Visit the RSA Data Security Art Gallery and >download to your hearts content. BTW, the URL for this is: http://www.rsa.com/rsa/gallery/gallery.htm Alan Olsen -- alano(a)teleport.com -- Contract Web Design & Instruction `finger -l alano(a)teleport.com` for PGP 2.6.2 key http://www.teleport.com/~alano/ "Governments are potholes on the Information Superhighway." - Not TCMay

1 0

recent garbage posts
by Daniel Leeds 17 Dec '03

17 Dec '03

The user in question has been talked to. DO NOT mailbomb root or any other account any more, this will be seen as hostile activities. We apologize for your inconvenience, but please, the death threats, physical bomb threats, and other nonsense is not called for.

1 0

Re: New Mitnick Book
by Duncan Frissell 17 Dec '03

17 Dec '03

At 02:36 PM 1/5/96 -0500, Brian Davis wrote: >FWIW: When Markoff showed up with Tsutomu, the FBI agents "assumed" that >he was there with Tsutomu as a fellow researcher/grad >student/assistant/whatever. > >They had no idea (initially) that he was with the media. > >... > I take it that you were in the vicinity? Or is this just what you've heard? Any more iside tidbits on the bust? DCF

2 1

Re: Representations of Pi, etc.
by jim bell 17 Dec '03

17 Dec '03

At 11:45 PM 1/4/96 -0800, you wrote: >At 3:19 AM 1/5/96, jim bell 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. And practically everybody else, I'm sure! It was written up in Science News magazine, BTW. >The part about how it works in some bases, but not in decimal. Well, by definition if it works in hex, it'll work in octal, "quadal" (?), binary, etc. That's ASSUMED, since they would be subsets (co-sets?) of each other. >The "hand-waving" (motivational/informal) explanation for why I am >surprised is that "Nature doesn't care about bipeds with 10 digits vs. >bipeds, or whatever, with 2 digits or 16 digits." That is, results >applicable in base 16, hexadecimal, should be easily applicable in base 10. 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. >And there is are interesting properties about the distribution of digits in >"random" numbers. Pi is of course not random by many definitions, but >shares certain important properties with random numbers. (Or sequences, if >you wish.) One of these is properties is that of _regularity_, the >frequency of digits. A regular number is one whose expansion has in the >limit the same frequency for all digits, and this is so in any base. Thus, >a regular number has an equal frequency (in the limit, blah blah) of 0s, >1s, 2s, 3s, etc. And switching to another base will not change this. Not "regular," you used the wrong term. As a mathematician's term of art, this is called "normal." In other words, equal numbers of digits 0-9, equal numbers of digit pairs 00-through 99, equal number of digit triplets 000-999, etc. A series of random numbers, by definition, must be "normal." But "Normal" numbers do not necessarily have to be random. As you might expect, however, testing for "normality" is a good first test of "randomness." (I'm not a mathematician, and I don't play one on TV. Don't be overly impressed with the preceeding knowledge; I'm sure there are dozens if a hundred people reading cypherpunks who know more math than I do. I only got a 780 on the math portion of my SAT, 20 years ago... 790's and 800's weren't all that uncommon.) >I recollect that pi has been proved to me regular, i.e., that pi has an >equal frequency of all digits, in the limit, in all bases. Yes, pi is apparently "normal." (at least four the first 4 billion or so digits...) >(This is the sense in which we can argue that pi is "random." in the sense >that there are no correlations, no dependence of the n+1th digit on the nth >digit, and "no apparent order." Furthermore, there is no effective >compression of pi, except by some tricks, such as _naming_ it (a dictionary >compression, of sorts) or by specifying a program which computes it. Lots >of interesting issues about the real meaning of randomness and >compressability, about the "logical depth" of certain computations, etc. I >recommend "The Universal Turing Machine" (ed. by Haken, as I recall) for a >nice set of articles on these fascinating issues.) Of course, I'm not sure of the ramifications of this new discovery (individual-digit computability of pi) on the facts you list... (for example, inter-digit dependence) However, if the digital representations of the digits of pi were indeed still individually "random" and they could be individually computed rapidly enough, at least hypothetically you could use the digits of pi as a "one-time-pad". Obviously, everyone would know (at least, conceptually) the ENTIRE CONTENT of the pad; the only issue would be the location where you started. If the starting point was the key, and could be defined by a number of at least, say, 256-bits long, it might be a replacement for IDEA whose current length is fixed at 128 bits. I suppose the problem with this technique will probably be that it would take too long to calculate the (a number somewhere around 2**256) bit of pi, and for an n-bit message you'd have to do this n times. >In summary, I would be surprised to find that a method for calculating the >Nth digit of pi works for base N but not for base M (modulo some minor >efficiency factors related to machine architecture, etc.). I'm sure somebody else will have seen it. It was in Science News in the last couple of months, as I recall. Some kind soul will probably type it in. Maybe I can even retrieve my copy from my sister.

1 0

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."

2 1

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