RFDC Scenario one

["Alley, Dennis" <dennis.alley@tgslc.org>]

Scenario 1:

Concerning telephone confirmation of signed key-change transmittals,
consider the following scenario:  

One user suspects that their public/private key-set has been compromised
so they create a new key-set and dispatch the new public key via
encrypted and signed cipher to all members on the public key-ring.
Assuming that there are five other users on the key-ring, the original
user will receive one telephone call from each of the five users seeking
to confirm the key change.  

TGSLC is planning for considerable numbers of users within its domain.
Since ALL 
Commonline members should have entries on the aggregate Public Key Ring,
the otal number of links will, most likely, exceed one-thousand.  That
would mean dispatching key-change ciphers to each member on the ring
and, more importantly, subsequently dealing with confirming telephone
calls from each and every one of those members on the ring.  Picture
support staff dealing with more than a thousand phone calls every time
you change your key.  And what about our customers having to change
their keys?  The concept of telephone confirmation is feasible only with
VERY low interface counts (central key servers).