Responding to feedback.
juan via agora-discussion [2023-02-28 16:30]:
> Here's an idea for a rule. […]
>
> {
> A Fingerprint for a document (the Plaintext) is a document that could
> not have been reasonably created without knowledge of the Plaintext, and
> which is, clearly and unambiguously, uniquely related to that Plaintext
> in some specified way.
>
> Where the rules define an action that a person CAN perform "by
> commitment" to a particular kind of document, that person performs that
> action by performing it by announcement while, in the same message, also
> publishing a Fingerprint for a document of that kind.
> }
Janet Cobb via agora-discussion [2023-02-28 14:38]:
> This results in every purported "by commitment" attempt diverging the
> gamestate, since it's impossible to verify the hash at the time of
> publication.
True. I fixed that by adding that a player needs only to provide a
document which purportedly is a Fingerprint.
nix via agora-discussion [2023-02-28 13:46]:
> Doesn't "unambiguously uniquely" exclude hashes? As I understand it, hash
> collisions are theoretically possible, even for all the uncracked encryption
> methods. I think you need a weaker standard like "The fingerprint must be
> unique enough to not be reasonably used to identify two separate documents"
> or such.
Good point. Changed it to a reasonability standard.
Here's the new version:
{
A Fingerprint for a document (the Plaintext) is a document that could
not have been reasonably created without knowledge of the Plaintext, and
which is related to the Plaintext in such a way that one could not
reasonably produce another document related to that Fingerprint in the
same way.
Where the rules define an action that a person CAN perform "by
commitment" to a particular kind of document, that person performs
that action by performing it by announcement while, in the same
message, also publishing a what is purportedly a Fingerprint for a
document of that kind.
}
--
juan
