-----BEGIN PGP SIGNED MESSAGE----- Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit To: cypherpunks@toad.com Date: Wed Jan 08 08:39:12 1997 Having gotten the subject in the spotlight (!), I'd like to clarify a couple of points about my draft proposal: 1: it is aimed at supporting _automatic_ realtime - as in supporting crypto-aware email packages, so that if an unknown key turns up, the encryption program can go out and collect key data _without_ user intervention. 2: port "assignment" was arbitrary; if the protocol would work better with a single port for send/receive, then "make it so, Number One" :-) 3: I expected that data transfer would work best using ascii text (like ascii-fied keys); if something else would work better, so be it. 4: if checksums or other error-detection/correction isn't necessary, then by all means, leave it/them out. I offered the draft as a _proposal_, fully expecting that it would require modification, editing, revision, et al. I'm not sufficiently pompous/egotistical to believe that I am He Who Knows All. Too, I have a reasonable idea of my own limitations; I'm simply trying to establish *some* kind of common reference point toward a solution of something I see as a possible problem. Happily, I haven't received any _personal_ flames, the comments I've received have been professional/technical. If I get a brownie point or atta-boy out of this, cool. If not, that's fine, too, _as long as the problem is addressed_. I didn't dare propose anything so grand as a distributed system - it seemed that a variation of DNS would handle that. We return you now to FlameFest '97 :-) Dave Merriman -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAwUBMtMI1sVrTvyYOzAZAQH5fgP+M5kJntTNM+LWd67jt+WGoiGLq8SdRwCa 0MljMYuO0oMzgiTrMTchNaj+cxtVkyUrzz+cgj3XQdJF6cdRlDnSW9xGV9rJJC/B lPf6RD9Vp5Pih9KkqWc+DqKMynGzEm1WXvNZKHDtn3/1etkv0RsEeQejqdNG+5dI rfZhi4xp9mk= =1Pcn -----END PGP SIGNATURE-----