home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.crypt
- Path: sparky!uunet!charon.amdahl.com!pacbell.com!mips!darwin.sura.net!jvnc.net!princeton!iso!dla
- From: dla@iso (Don Alvarez)
- Subject: Re: Secure netnews
- Message-ID: <1992Aug19.172555.14545@Princeton.EDU>
- Originator: news@nimaster
- Sender: news@Princeton.EDU (USENET News System)
- Nntp-Posting-Host: iso.princeton.edu
- Organization: Princeton University
- References: <9208191619.AA25204@news.cis.ohio-state.edu>
- Date: Wed, 19 Aug 1992 17:25:55 GMT
- Lines: 51
-
- In article <9208191619.AA25204@news.cis.ohio-state.edu> Marc.Ringuette@daisy.learning.cs.cmu.edu writes:
- >dla@raven.princeton.edu (Don Alvarez) writes,
- >> There is no way to defend against this because there is no way to prove
- >> who sent the message stating that machine foo has received an article.
- >
- >Sorry, Don, but you're just totally 100% wrong. Digital signatures exist!
- >They are practical, and they work.
- >
- >You might want to have a look at any tutorial on RSA, which will tell you
- >how to do secure authentication (digital signatures) assuming RSA is secure.
-
- I am well aware that RSA exists, and that you can use it for digital
- signatures. RSA is an encryption technique. It happens to be a
- particularly good one, but it is still just an encryption technique.
-
- The problem has NOTHING to do with encryption. No matter how you
- encrypt/sign/etc your message, we are still faced with the fact that
- you and I have never met and know nothing about each other. I could
- call you on the phone or walk up to you on the street and you still
- wouldn't be able to trust whether I sent a message, because you just
- plain don't know who I am. That's why it's hard to be a spy, and
- that's why it's hard to authenticate mail on the net. Multiply this
- problem by the tens of thousands of people using the net who have
- never met each other, and you realize why authenticating messages on a
- net-wide basis is a highly non-trivial task.
-
- If you don't believe me, tell me how you and I can use RSA to let us
- authenticate messages we wish to send to one another. Remember that
- (to the best of my knowledge) we have never met, and assume that
- George Bush's evil twin Skippy wishes to prevent us from communicating
- (which I personally doubt, but without an adversary we have no need to
- authenticate anything). We each have the suite of RSA code, as well
- as any other software you think would be helpful. You will quickly
- find that we still can't communicate securely.
-
- If you think you have a way, post it, and I will post a way for Skippy
- to spoof the method. Skippy _doesn't_have_to_crack_RSA_. Skippy only
- has to figure out a way to prevent us from getting a bona-fide copy of
- the other's digital signature validator. Cracking RSA is hard.
- Preventing us from getting bona-fide copies of each other's digital
- signature validators is easy.
-
- >You might want to have a look at any tutorial on RSA, which will tell you
- >how to do secure authentication (digital signatures) assuming RSA is secure.
-
- You might want to look at the literature on Key-distribution protocols.
- THAT'S where the problem is. How to START the communications link.
- Once we each have copies of each other's public keys, then it's easy.
-
- -don
-
-