On Mon, 31 Jan 2011, omd wrote: > On Mon, Jan 31, 2011 at 6:51 PM, Kerim Aydin <ke...@u.washington.edu> wrote: > > I agree that legislative clarification would help here, strongly. There's > > an interesting history to "gamestate" in reset proposals along with "deem" > > and other terms of legal fiction that have been used until strongly > > questioned and then discredited due to a weakness in the boundaries of > > the term. > > CFJs are judged based on the state of the game at a certain point in > time, which is no different than, say, the registration rule, which > examines whether the person was ever registered in the last thirty > days. But it seems to me that if a rule says it's looking at what the > game was like in the past, it should examine what the game *actually > was like then*; without an explicit rule to the contrary, saying "when > a rule purports to examine the past state of the game, it actually > refers to what's stored in this gamestate variable", the rules should > mean what they say.
Without definition, gamestate is the whole game as of now. Anything needed to interpret its current and future behavior is part of that state. So if a CFJ called now finds that history is needed to interpret the present or the future, that history (the state that "X happened on day Y") is part of the state. Mathematical description (definition?) on wikipedia puts it well: "Intuitively, the state of a system describes enough about the system to determine its future behaviour." Limiting the state to a subset of document-recorded quantities (some but not others! where to draw the line!) is a legislative matter that hasn't been performed. -G.
