- From: Pete Wenzel <pete@seebeyond.com>
- Date: Thu, 11 Jul 2002 23:28:20 -0700
- To: Joseph Hui <Joseph.Hui@exodus.net>
- Cc: www-ws-arch@w3.org
Thus spoke Joseph Hui (Joseph.Hui@exodus.net) on Thu, Jul 11, 2002 at 07:06:14PM -0700: > > From: Pete Wenzel [mailto:pete@seebeyond.com] > > Sent: Thursday, July 11, 2002 5:59 PM > > To: Joseph Hui > > Cc: Christopher B Ferris; Hal Lockhart; www-ws-arch@w3.org > > Subject: Re: "Onion model" explained > > > > > > Thus spoke Joseph Hui (Joseph.Hui@exodus.net) on Thu, Jul 11, > > 2002 at 04:27:32PM -0700: > > > [snip] > > > >> >Presumably she requires some proof of possesion, a certificate > > > >> >alone proves nothing, but this is a side issue. > > > >> > > > >> A certificate proves nothing? By certificate I meant a cert > > > >> that Alice > > > >> deemed trustworthy, say a CA signed cert. Now, if a CA > > signed cert, > > > >> which millions of Internet shoppers trust their money to, > > > >> means nothing > > > >> to you, then I don't think we'll go anywhere on the issue. > > > > > > > > A certificate alone means nothing, even one that is CA > > > > signed, absent a CPS. > > > > > > Well, it means whatever Alice deems trustworthy or otherwise, > > > assuming Alice already practices "safe certs," such as > > > checking the CRL (Cert Revocation List) with due diligence. > > > [snip] > > > > These are all important considerations for establishing a > > certificate's trustworthiness, but I think you are both still missing > > Hal's "proof of possession" point. > > If there was point there, that wasn't a salient one wrt > the gist in my original (first) response to Hal: authN > alone can suffice in some cases. I have paid careful attention to your authentication-only examples, but still believe that they involve a binary authorization policy, whether it is implicitly or explicitly stated. For Alice & Bob, it is "All authenticated parties are authorized to submit messages, which I will in turn accept for further processing." If Bob can authenticate his identity, Alice will accept. If he cannot, she will reject. Or, "If I succeed in authenticating the server at the other end of my connection that claims to be www.amazon.com, it is then authorized to receive my credit card info, so I will type it in. Otherwise, it is not authorized, and I will not proceed with the transaction." Yes, authentication happens first, then an authorization decision is made, so they can be treated separately; but having authentication alone is not useful. It is not a very interesting application that performs authentication, then continues in the same manner regardless of the outcome of the authentication. > > The fact that I can present some > > random certificate to you, even if it passes all validity checks, only > > serves as identification. My identity is not authenticated (bound to > > the identifiers in the cert) unless I can actively prove to you that I > > am the subject of said cert. (This by demonstrating that I can sign a > > challenge nonce, verifiable using the cert's public key; or that I can > > decrypt something that you have encrypted with it.) > > Again, let me reiterate what I expressed in previous messages of > this thread, it all comes down to Alice's discretion: in what > she trust? In trust and sec, you get you pay for, but don't > pay for what you don't need. There're many happy PGP folks > out there. Not everyone, in fact most e-commerce practitioners > don't, find it necessary to add challenge/response on top of > trusted certs. (BTW, this has been a well treaded area by > non-repudiation enthusiasts.) True, the authentication I spoke of need not be of a challenge/response form. The PGP or e-commerce user's goals may be satisfied as long as *something* is signed or decrypted, proving possession of the correct private key. > Say, if you buy stuff from an > https website, do you chllenge the sellers? I bet you don't, > even though it's your money that's at stake. I trust that my SSL/TLS-enabled browser challenges the seller's web server. It encrypts the pre-master secret using the (supposed) server's certificate. If the server is unable to decrypt it, that proves it likely to be unauthentic, and the protocol terminates. --Pete > Joe Hui > Exodus, a Cable & Wireless service > ======================================== > > > > --Pete > > Pete Wenzel <pete@seebeyond.com> > > SeeBeyond > > Standards & Product Strategy > > +1-626-471-6311 (US-Pacific)
Received on Friday, 12 July 2002 02:28:52 UTC