trivia
Owen Lewis
oml at eloka.demon.co.uk
Fri, 13 Jul 2001 10:42:01 +0100
> -----Original Message-----
> From: ukcrypto-admin@chiark.greenend.org.uk
> [mailto:ukcrypto-admin@chiark.greenend.org.uk]On Behalf Of David Howe
> Sent: 12 July 2001 17:32
> To: ukcrypto@chiark.greenend.org.uk
> Subject: Re: trivia
> There is always the "degrees of separation" thing. assume you have a
> provably innocent correspondent. now assume one of THEIR correspondents is
> now a convicted criminal. if they are investigated, then you will
> be a known
> correspondent of a suspect.....
The man who danced with a girl who danced with the Prince of Darkness?
> > How is A to put B into such a position inadvertently? I think
> > that if A is suspected of a serious crime and has been in
> > enciphered correspondence
> > with B, then B might be asked to deliver up correspondence or key.
> > Fair enough.
> > There *has* been correspondence and there *is* (or has been) a
> > secret key
> > that was once in my possession. Were I 'B', I'd have the key and would
> > deliver as demanded (albeit without the best of grace).
> ok, imagine the following.
> A and B want to have a encrypted conversation, and have secure (ssh
> tunnelled) mail.
> A generates a key he does not upload, in your name; his correspondent does
> similarly. both change their keys to their own name, and send a
> copy in that
> form. mail is saved off WITHOUT headers in a subdirectory, which is simple
> enough to do, and both correspondents are careful never to use
> each other's
> name.
> now, one gets arrested; the other immediately uploads the public key (in
> your name) he holds for HIS private key, and the public key (in the others
> name) to the keyservers, then secure-wipes his private key for that
> exchange. what is more likely? that they will pursue all the possible
> correspondents for the suspect, or that they will take the name on the key
> at face value and come looking for you?
Return to the origins of the thread. I've not claimed that using PKC cannot
make trouble for one but only that I would try and act in a way that
minimise that risk. In short, I would not act as you describe.
We are all different and act according to our different needs, education and
experience. In toto, we may know as much and be as experienced as each other
but the vector from the sum of our personal knowledge and experience is
likely to be different.
Much of my education and experience has taught me to be cautious of the
motives of others and, as a result, I would never use a cipher system I
controlled to communicate with persons other than a selected few. With those
selected few, I will conduct my secure communications in a careful and
relatively straightforward way.
Should there be a case that one of them has been conspiring to blow up
Parliament, then as an intimate of his I can fairly expect some questioning
and, with the law as it now is, to have my correspondence from him examined.
However I still believe that it is by no means certain that it would be.
Were it to be so, as I have said I would have the key and would, under
order, provide the plaintexts required of me.
For the few with whom I communicate in cipher, there is a unique key for
every correspondent. Thus, if correspondence with one is compromised it has
no effect on the remainder. It suits me well to conduct my business this
way. There is no reason why others should find such a way best suited to
their needs. That matters not. It suits mine and those of my correspondents.
It seems to me that the main risk that requires some thought to damage
limitation stems from one's selection of intimates. But this has been well
thrashed out here - albeit inconclusively - in the past.
Owen
Owen