On Mon, Jul 16, 2012 at 6:09 PM, Kerim Aydin <[email protected]> wrote: > > > On Mon, 16 Jul 2012, omd wrote: >> On Mon, Jul 16, 2012 at 11:54 AM, Kerim Aydin <[email protected]> wrote: >> > then the encoding is interpreted as if its plaintext had been >> > published at the time the original encoding was published. >> >> i think this is too vague - what if the plaintext is "I deregister"? > > Ah, good point. Hmm, the plaintext should be valid if applied to the > actions manipulating the Rules-defined Confidential record, but not > otherwise valid. That might be a little tricky to generalize correctly... > > -G.
The rule specifying the secret should additionally specify the encryption (I hate this word in this context, fwiw) method and the rule allowing secrets should define explicitly what sort of actions or information can be contained within. For instance, on voting, we could say that a player may indicate that their vote is contained in a secret, but they must clearly identify which secret and it cannot contain anything other than a vote. -scshunt
