From: "rudy (r.) rawlins" <rudy@bnr.ca>
Date: Tue, 15 Feb 94 10:40:52 PST
To: cypherpunks@toad.com
Subject: Re: Clipper and Traffic Analysis
Message-ID: <"2152 Tue Feb 15 13:25:01 1994"@bnr.ca>
MIME-Version: 1.0
Content-Type: text/plain


In message "Clipper and Traffic Analysis", pmetzger@lehman.com writes:


> 
> Normally, one can only determine the endpoints of a conversation. With
> clipper, however, one can deduce a lot more, since when people move
> around, go to hotels, phone booths, etc, you can still track their
> clipper serial numbers.
> 
> Perry
>             

Maybe we are missing something here, when people move around if they
simply use their hotels phones, phone booths, et al, how can one track
their clipper serial number? Unless there is a personal Clipper serial
number? I do not see any extra info being garnered from Clipper phones,
just less.

Consider the case where there is a wiretap in progress: In the past one
end was tapped and both sides of the converstion were heard. You could
always find out who was calling, but you could not go and bug the
calling party's phone without a court order. With Clipper, you tap the
outbound voice/data, but every inbound voice has to be decoded with its
own key. Now will the wiretap allows blanket decryption for all Clipper
phones? and if so where does the tap begin and end. If no blanket
decryption is allowed then  must they record outgoing voice/data and
based on the content of  one side of the conversation  convince a judge
to let them decode the other side? Both scenario is problematic. 

If I'm cop and the suspect does not have a Clipper phone at home? can I
choose  which end of the switch to tap?  I would choose the analog side
that still has voice -- both voices, so even if the network is
Clippered, why go through the hassle of managing keys when I can always
hook onto an old fashioned analog voice line.

Am I seeing correctly or did I miss something?