December 2003 - cypherpunks-legacy

Re: Clarification of my remarks about Netscape
by kipp＠warp.mcom.com 17 Dec '03

17 Dec '03

In article <199412130729.XAA01473(a)jobe.shell.portal.com>, you write: > -----BEGIN PGP SIGNED MESSAGE----- > > It is nice to have a lot of people on the list from Netscape. > Here is a question about SSL relating to the use of certificates: > > + The issuer name must resolve to a name that is deemed > acceptable by the application using SSL. How the application > using SSL does this is outside the scope of this memo. > > What does Netscape actually do about this? If I want to make a server > which will interoperate with existing Netscape clients what kind of > certificate do I need, and what kind of name should be in there? > Thanks - > > Hal Finney > hfinney(a)shell.portal.com > > -----BEGIN PGP SIGNATURE----- > Version: 2.6 > > iQBVAwUBLu1NOxnMLJtOy9MBAQGItwIAr4eerI+FSmPpOIcwITepnXzcUUFkPwsK > +Rz2FC4Y6hV0HoDEt1JnpvCPVV5N74Jtc9xMmF8CcRlBybk25PkxVQ== > =LOql > -----END PGP SIGNATURE----- Because online directory services are not one of the extant solved problems on the Internet, Netscape uses a simple approach - a small set of "important issuer" certificates are compiled into the browser. A future release will support "key rings" ala PGP. This is all we had time for in this release... All you need to do is get your server certificate from one of several places, including: RSA (commercial CA or server CA) Netscape (not likely; we can't afford the liability) MCI (I don't know if they are selling this). So the short answer is: it's hard to do right now. In six months it should be a very different scenario. --------------------------------------------------------------------- Kipp E.B. Hickman Netscape Communications Corp. kipp(a)mcom.com http://www.mcom.com/people/kipp/index.html

1 0

Re: the netscape/ssl controversy
by kipp＠warp.mcom.com 17 Dec '03

17 Dec '03

In article <199412131448.JAA07613(a)cutter.clas.ufl.edu>, you write: > have to get my two cents in: > > netscape has a proposal that secures the *transport* of files. pgp and > the like secure the actual files themselves. the question that will > have to be answered is "which method of security is more valuable to > the internet as a whole?" > > of course i have made up my mind, but i'll waste some space and share > those thoughts here. > > ...much elided... > > so while ssl may well be a wonderful protocol, it does not address the > problems that many cypherpunks see as being real. i think it would do > netscape good to listen to and consider the views of many on this > list, as they have many genuinely good ideas, even if they choose a > confrontational manner. My personal opinion is that using MIME encoded documents with embedded security information including digital signatures combined with SSL or other trasport layer security, will be a reasonable solution. Please check out: ftp://ds.internic.net/internet-drafts/draft-ietf-pem-sigenc-02.txt and ftp://ds.internic.net/internet-drafts/draft-ietf-pem-mime-07.txt --------------------------------------------------------------------- Kipp E.B. Hickman Netscape Communications Corp. kipp(a)mcom.com http://www.mcom.com/people/kipp/index.html

1 0

Re: BofA+Netscape
by jgrubs＠voxbox.norden1.com 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- Rick Busdiecker <rfb(a)lehman.com> writes: > Even Apple & Microsoft agree that Netscape is brain dead... Is that the same Microsoft that brought us Edlin??? -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAwUBLu3Lud74r4kaz3mVAQEWvwQAp/fQ5DmE28saQziqd+AJU9RIaqc9lQ8k /oDfOGLmA9RFL1R1hHbTBmQqd+rdjZgtW+o1J9WTMZ3ttoR03G8Nhr7gtrZncMTD /OmL7BmKGQTjp24MsIXk+oBdpyN/X1VqNrGZvuK9/GIDpkEA5d7Q8qlAHLYDRkv8 KcGJBMRRBFU= =0gC5 -----END PGP SIGNATURE----- ... "The greatest dangers to liberty lurk in the insidious encroachment of men of zeal, well meaning but without understanding." - Justice Louis Brandeis -- jgrubs(a)voxbox.norden1.com (James C. Grubs, W8GRT) Voxbox Enterprises, 6817 Maplewood Ave., Sylvania, Ohio 43560-1956 Tel.: 419/882-2697

2 1

Re: More 40-bit RC4 nonsense
by kipp＠warp.mcom.com 17 Dec '03

17 Dec '03

In article <199412131742.JAA27330(a)netcom5.netcom.com>, you write: > -----BEGIN PGP SIGNED MESSAGE----- > > If I recall correctly, the first byte out of the RC4 stream has > about a 40% chance of being the first byte of the key. Thus, if the > 40-bit "secret" part of the key is the _beginning_ of the full 128-bit > key, then the keyspace is effectively reduced by about seven bits, > meaning that I would be able to crack a key on my PC in a couple of > days or so. > Of course, if the "clear" 88 bits went first, there would be no > advantage whatsoever. The SSL document very carefully does not say > how they combine the two key parts to form the 128-bit key. Does > anyone know? > > Raph > > -----BEGIN PGP SIGNATURE----- > Version: 2.6 > > iQCVAwUBLu3cI/4BfQiT0bDNAQEToQQAtcy2v0sBd+g5GBrm+Pa1AykqS4tTctfu > EYga7kPry4wvGmI7/HpD+SVVDQRcJe+O9CxH9cpvRgBRIBhyvsFXVBSTW0OTJgXb > 1bYh5qerD5J/gXAs0XWIp0+Hj8GqeTIRkFTseU4MDcDfQ7tOSEFvul97iSNYIytX > AMkmAEmMXxU= > =S80T > -----END PGP SIGNATURE----- OOPS. This is a spec ommission. The clear key data (aka "salt") is combined with the secret portion as follows: The bytes of the salt are concatenated with the secret portion with the secret portion making up the least significant bytes of the concatenation. I will spec'ize the english... By the way, where did this 40% number come from? For some reason RSA never told me this... :^( --------------------------------------------------------------------- Kipp E.B. Hickman Netscape Communications Corp. kipp(a)mcom.com http://www.mcom.com/people/kipp/index.html

1 0

Re: Clarification of my remarks about Netscape
by Rich Salz 17 Dec '03

17 Dec '03

> Please define a solution that is: > distributed > reliable > supports an unforgeable name to public-key mapping > standard > not-bulky > not-expensive Build a new API out of sockets and GSSAPI? ssl_t s = ssl_bind(char *hostname, int port, char *servername); ssl_status_t ssl_secure(ssl_t s, oid_t security_type, unsigned32 flags); etc... /r$

1 0

Re: News->Mail & CMU's Andrew Message System
by Rich Salz 17 Dec '03

17 Dec '03

> AMS is pretty strongly tied to AFS (now known as part of DCE) At some levels DCE's distributed file system (DFS) is "AFS version 5" (sic), but not really. The original intent was that DFS was AFS with the IPC replaed with secure DCE RPC, but there's a whole mess of protocol changes since then. They're different beasts. /r$

1 0

Winsock & PGP Integration
by Johnathan Corgan 17 Dec '03

17 Dec '03

-----BEGIN PGP SIGNED MESSAGE----- It has been a long discussed and desired feature for MS-Windows users to be able to integrate PGP functionality into the various Windows Sockets based mail and news readers. After a discussion with Raph and Pierre at the cypherpunks meeting this weekend, I took a look at the paradigm that premail uses and gave some thought to how this same concept could be applied in the Windows environment. On a Un*x machine, premail works by impersonating the mail transfer agent and intercepting the flow of mail in each direction, adding encryption, signing, and anonymous remail services in a rather elegant fashion. As the mail system under Un*x (indeed, the entire OS) is designed to be a "piped and glued together" batch of smaller utilities, premail's method works well and is very 'unix-like'. How to achieve the same under Windows with winsock based SLIP or PPP access? The various mail agents such as Eudora and Chameleon are integrated packages that do everything from using SMTP and POP for mail transfer to providing the user agent that reads and writes mail. There really is no simple way to wedge into the package and replace or supplement functionality. Except one. What all of these agents have in common is that they interface with the Windows Sockets API to establish TCP streams that are used in the POP and SMTP protocols. Since these are well known and standardized protocols, this gives us our toehold. Picture this. Using a replacement WINSOCK.DLL, we can intercept those specific API calls that an application would use to open the appropriate sockets and establish a TCP stream to a particular remote socket. For all other API calls, we hand these off to the "real" WINSOCK.DLL (that we have renamed and loaded after our DLL has loaded). Our replacement DLL now has the ability to make the mail application think it is talking to a remote SMTP or POP process, when in fact, it is talking to our agent on the local machine. Our local agent spoofs the SMTP protocol, accepts the mail, does all the neat crypto stuff, and stores the mail in an outbound queue. At whatever preprogrammed intervals, _our_ agent makes the _real_ SMTP call to the remote and delivers the mail. The reverse process would work as well. Our local agent periodically makes a POP call to the remote end, retrieves mail, and stores it in an inbound queue. Again, we do all the neat crypto stuff, and finally the next time Eudora or whatever mail agent is in use makes its POP run, we spoof POP and deliver the mail to the mail program. A little thought on this and I realized that this could be a generalized process. Want to implement an anonymous remailer under Windows? Use this mechanism and replace "neat crypto stuff" with "even neater anonymous remailing stuff." Majordomo for Windows? No problem, same method. By intercepting the flow of mail in each direction, there is no limit to the stuff you could come up with. Why bother fiddling with Windows anyway, when one has all the beauty and power of Un*x at one's disposal (such as Linux)? Well, think about where the personal computing sector is going to grow in about six months or so with the advent of Windows95. Okay, it's a great OS for what it is designed for, but it doesn't have pipes or perl or twenty or thirty command line parameters for every executable. So what. I garauntee that there will be more laptops and desktop PC's with Windows95 than with any flavor of Un*x. If we are to promote and encourage, as well as make the tools for, the use of strong crypto by the average Joe, then Windows is where it is at. I've just enough Windows programming experience to know that the scheme I worked out above is feasible and would work well. On the other hand, I have no experience with Windows Sockets at the api level, nor with any implementation of SMTP or POP (though I know the protocols themselves well.) Yep. It would be cool to write a message in Chameleon (such as this one), press the send button, and have a dialog box pop up asking me if I wanted to sign, encrypt, or remail this message. Or to have it done automatically for me. And this would work with existing mailers today. Sheesh, somebody burst my bubble. Comments, criticisms, suggestions, and especially financial donations, are quite welcome :) ======================================================================= Johnathan Corgan "Violence is the last refuge of the incompetent" jcorgan(a)netcom.com -Isaac Asimov PGP Public Key: http://www-swiss.ai.mit.edu/~bal/pks-toplev.html Or send email to: pgp-public-keys(a)pgp.ai.mit.edu Subj: GET jcorgan ======================================================================= -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAwUBLu0w2k1Diok8GKihAQGdnQP8DiqQt3820dhKHY3dbXAZnl/11eEce4z+ /oFZMKVBHlCJAxCSucnK31dcyRbvXiOkAt9x9EMinXC26VWh3sETd+YbaJNznkx6 VmM1UNID2bQ+Xpcc4ANJQx6CgPrRxspr3reVeuFv7QLtKGMy5Mucl5mXGp829f6i Gj7NMlhcu+s= =rgyq -----END PGP SIGNATURE-----

3 2

James Donald's comments
by perry＠imsi.com 17 Dec '03

17 Dec '03

James Donald has claimed that I've been talking about vaporware and that Netscape has produced "real cool crypto for the masses". As I've noted, real code to implement proposals quite similar to IPSP is out there right now, and has been for half a year. The reason that it isn't very widely deployed is largely because we haven't been pushing it because it isn't fully clean yet, not because it couldn't be used. (It is indeed being used by real people with real applications). .pm

2 1

Re: James vs. Eric
by Blanc Weber 17 Dec '03

17 Dec '03

From: "L. McCarthy" Eric & James, how about taking the "You're flaming me !" "No, I'm just insulting you !" portion of this debate to private email ? I sincerely doubt that very many people on the list give a damn about any of the above jousting. .................................................................. Actually, I find it mildly amusing, myself, in between the Netscape thing, and I'm also picking up on logic, game theory, and white spaces. Except I don't know what "De mortuis mentis, nil nisi Latinum" means ("the brain dead can't understand Latin"?) :>) Blanc

1 0

What, exactly is elliptic encryption?
by dfloyd＠io.com 17 Dec '03

17 Dec '03

What, exactly is elliptic curve encryption? (Only thing I knew that the NeXT nearly had it in its OS, but the heavy hammer of ITAR squashed that...)

5 6