[Cryptography] On the 'regulation proof' aspect of Bitcoin

Phillip Hallam-Baker phill at hallambaker.com
Thu Mar 31 10:19:11 EDT 2016


On Thu, Mar 31, 2016 at 8:58 AM, Alfie John <alfie at alfie.wtf> wrote:
> On Wed, Mar 30, 2016 at 01:21:15PM -0400, Phillip Hallam-Baker wrote:
>> On Tue, Mar 29, 2016 at 9:14 PM, Tamzen Cannoy <tamzen at cannoy.org> wrote:
>> Bitcoin is at least two different things, it is a mechanism for
>> assigning value and it is a mechanism for transferring value between
>> parties.
>>
>> I would like to be able to make use of a BitCoin like transfer scheme
>> without the mining bit. I don't think that is actually necessary to
>> maintain the integrity of the chain.
>
> Without mining, what you're talking about is simple data replication. You might
> want to have a look at the the BitTorrent protocol, or even rsync.

Well the Merkle chain is remarkably resilient by itself.

Consider a situation in which we have ten independent notaries
maintaining separate public chains. Every day they take the output
from every other chain and enroll it as an input. Any attempt at
rollback now requires every notary to collude and even then the
defection will be obvious to anyone keeping notes.

This is where I got the idea of calling the system a 'Mesh'. Its like
a huge collection of gears all meshed together. None can turn unless
every other one turns.

It is pretty easy to see that any system that has chained notaries
will quickly end up making interchange agreements and that these will
rapidly converge into one system.


So the upshot of all this is that if you want to fix data D in time,
you simply calculate H(D, tx), where tx is the current output token of
any notary and then enroll that result into a notary. The time at
which this occurs is precisely fixed in time. It cannot be moved
forwards or backwards.

The workfactor for moving the notary timestamp is effectively breaking
the underlying hash. So we can take it as being infinite (I only use
SHA-512).

The notary isn't actually a trusted third party in the long term.
After the data is enrolled in the mesh, it no longer has any more
ability to assist an attacker in any fashion whatsoever.


I had to look at rsync in detail as prior art for a client. Under the
covers it is very similar to git in fact.


More information about the cryptography mailing list