[ih] TCP options: Bubba and Skeeter

[Craig Partridge]

In message <200111300021.fAU0LW508101 at boreas.isi.edu>, Christos Papadopoulos wr
ites:

>Hi everyone,
>
>one of my students stumbled accross this one. Look at option
>kind numbers 16 and 17 below.  Anyone knows the history behind it?

Knowles is certainly Stev Knowles -- haven't reached him but found
a co-conspirator.  Attached note sent with permission.

Craig

    From: "Kastenholz, Frank" <FKastenholz at unispherenetworks.com>
    Subject: Re: skeeter & bubba TCP options?

    ah, the sins of ones youth that never seem to be lost...

    it was something that ben levy and stev and i did at ftp many
    many moons ago. bridgham and stev were the instigators of it.
    the idea was simple, put a dh key exchange directly in tcp
    so that all tcp sessions could be encrypted without requiring
    any significant key management system. authentication was not
    a part of the idea, it was to be provided by passwords or
    whatever, which could now be transmitted over the internet
    with impunity since they were encrypted... we implemented
    a simple form of this (doing the math was non trivial on the
    machines of the day). it worked. the only failure that i 
    remember was that it was vulnerable to man-in-the-middle 
    attacks.

    why "skeeter" and "bubba"? well, that's known only to stev...

    f