This is a bit of a tangent on this discussion here: http://www.mail-archive.com/agora-discussion@agoranomic.org/msg36920.html , which looks very solid except for the part of "at the moment", which likely makes the flaw not a flaw (because of the use of language).
Anyway, if the following is true, then Ais532's scam might've actually not worked, since IIRC it depended on "withdrawing", which is unregulated. (But it has a few more implications than that, would it be true). So, if R1698 is actually describing a failure case here... "If any other single change to the gamestate would cause Agora to become ossified, or would cause Agora to cease to exist, it is cancelled and does not occur, rules to the contrary notwithstanding" ...that means that "any single change to the gamestate" is *regulated*. Because it satisfies the following in R2125 : "(2) describe the circumstances under which the action would succeed or fail". Note that If we didn't have that rule, an action that would attempt to Ossify Agora *would* work. But since we do have that rule, it doesn't work. Therefore, R1698 (the Ossification rule), describes a case where an action (an action that causes a change to the gamestate) would fail. Does "withdrawing" a vote cause changes in the gamestate? If "Yes", then its a regulated action. Taking into account that "Regulated Actions CAN only be performed as described by the Rules.", then withdrawing can only be performed as described by the rules. There is no description of how to withdraw (just the effects of withdrawing), so there is no way to actually withdraw anything. (Or do any other unregulated action that is relevant to the game).
