[Cryptography] Blockchain to Secure Nuclear Weapons?

[Ben Laurie]

On 13 October 2016 at 15:14, Natanael <natanael.l at gmail.com> wrote:
>
> Den 13 okt 2016 11:50 skrev "Ben Laurie" <ben at links.org>:
>>
>> On 13 October 2016 at 01:54, Peter Todd <pete at petertodd.org> wrote:
>> > This is similar to how Certificate Transparency must be more than just
>> > timestamping to give the auditability guarantees it attempts to provide.
>>
>> Curious what gap you think there is? Or do you mean CT does provide
>> the "more" that is needed?
>
> The gap is the ability vs lack of ability to say that there's no data being
> hidden from you, the ability to know when you have access to one entry that
> there is no entries that conflict with yours.
>
> A pure timestamping service has no concept of conflicts or versioning.
> There's no concept of exclusivity or "rivalry". But in Bitcoin's blockchain
> you can't spend an UTXO twice,

Which requires clients to have a complete copy of the blockchain,
right? Its not efficient, though. That's why we're doing Trillian
(https://github.com/google/trillian).

> in Certificate Transparency you can't issue
> two separate certificates for one domain undetected (revocations are also
> tracked).
>
> Factom (works on top of Bitcoin) goes even further, when you look up entries
> you should be able to confirm that what you hold is the latest version, or
> to otherwise find the one correct latest version. It ensures there's only
> one active version at a time, no forking. Each new one replaces the previous
> one. They call it proof of process.

Sounds like Trillian.