On Tue, 2009-07-21 at 16:01 -0700, Kerim Aydin wrote: > I still don't understand this scam. > > coppro argued that: > > it is IMPOSSIBLE to send a public message except where allowed > > by the rules. > because of the equivalence of "acting on behalf of oneself" and just > general "acting by sending a public message." > > But R478 authorizes just that sort of action, thus allowing it by > the rules: > Where the rules define an action that CAN be performed "by > announcement", a person performs that action by unambiguously > and clearly specifying the action and announcing that e performs > it. > > so R478, if it authorizes acting by sending a public message, also > (by the very equivalence that coppro uses) authorizes acting on behalf > of oneself at the right level of Security. And an act by announcement > of "I do x" fits all the (a-c) of what needs to be specified to act- > on-behalf. > > Now, either the authorization is either by R478, or by whatever rule > defines the act. In the case of CoEs, R2201 defines CoEs, and it > is power-3, so the ability is there. > > (Note, in passing, that a CoE is not a tightly defined "document" that > is created, but just a result of noting an error, so also, as long as > R101 protects the act of sending a message, the fact that it > contains a claim of error is a fact of the message. "Claim" is given > the same general definition of "says" or "writes" so all I have to > do to have made a claim is to send the message - R101 protected.) > > This just seems straightforward that R478 works, at least for actions > defined at power 3, if the equivalence is both ways. What am I > missing here? Something deep in the missing IRC logs?
I'm actually starting to agree with this argument; actions defined in power-3 rules probably are possible due to the definition of CAN. I still don't see anything that allows lesser-powered rules to define actions, though; so instead of a dictatorship scam, we just have an interesting brokenness. -- ais523
