[Cryptography] Plan to End the Crypto War

[Phillip Hallam-Baker]

On Mon, Jan 18, 2016 at 11:35 AM, John Gilmore <gnu at toad.com> wrote:
>> Not singular "split golden key"  but plural this involves key pairs for
>> both ends of the conversations.
>
> The discussion is slightly interesting, but please notice that it
> isn't relevant to Chaum's proposed design.  Beating the stuffing out
> of a propped-up strawman isn't much of a feat of strength.  cMix
> doesn't have a golden key, it doesn't split that key, etc.
>
> It's worth reading Chaum's cMix design to at least *understand* what
> you may want to criticize.  Try:
>
>   https://eprint.iacr.org/2016/008.pdf

Quite.

And moreover, Chaum is first and foremost showing a new technique
here, applying homomorphic encryption to a particular problem. Folk
who are spending their time complaining about the requirements he has
chosen should note that as with most cryptography, the same technique
can be applied to many requirements.

The impossibility of government backdoor access doesn't come from any
lack of cryptographic capability, it comes from the utter incoherence
of the requirements. I was reading a piece by Kaminski on BitCoin
where he complained that it claims the advantages of both its current
and its future architecture and leverages one off against the other. I
get a similar feeling when I am reading the government backdoor
proposals, they are fully accountable but do not want to have to
bother with warrants, they are only demanding exceptional access but
every transaction must be visible, etc. etc.

Chaum's proposal is a useful contribution to that debate insofar as
the FBI will clearly reject it and so would anyone who might choose to
use a mix network. Identifying a mid point between the positions of
the two camps that is obviously infeasible is a good step towards
showing that the intersection between the two positions is the null
set.