On Sun, Aug 5, 2012 at 12:56 PM, Ed Murphy <[email protected]> wrote: > scshunt wrote: > >> Based on the arguments to CFJs 3121 & 3122, it appears that the >> consensus among Agorans is that when an infinite rule-defined process >> occurs, it does indeed occur infinitely, but instantaneously, leaving >> the game in a single state afterwards. While the situation giving rise >> to 3122 failed on a technicality similar to that which plagued the >> original CFJs on this matter (see CFJ 3246), this one does not have >> the same issues. >> >> In eir arguments to CFJ 3122, H. Judge Murphy proposes that any such >> action would necessarily introduce ambiguity into the gamestate and >> fail, but I do not agree with this line of reasoning. I see nothing in >> the rules to indicate that an ambiguity causes an action to fail >> inherently; rather, it is the requirement of unamibiguousness written >> out in Rule 478 that prevents actions by announcement from being >> ambiguous. This does not prevent other ambiguity and, indeed, >> ambiguity has been held to exist in situations historically---there >> have been . There is, additionally, a poltiical aspect to it, as >> outlined by H. Judge Pavitra in CFJ 2650, which is a vital read for >> anyone wishing to settle this case. > > > Minor nits: My arguments to CFJ 3121 (not 3122) were not based on > ambiguity, but rather on the rules attempting to deem some finite set > of messages as legally equivalent to an infinite set of messages (which > would be physically impossible to send directly). However, since the > rest of your judgement implicitly refutes the prior precedent anyway, > and is not obviously unreasonable in doing so, I won't bother pushing > for reconsideration. > > I do think that this makes it too easy to set up infinite chains and > get turtles out of them (e.g. by fudging someone's posture, which has > no LFD-style escape clause), but that probably needs to be fixed via > legislation at this point.
Switches do as well, and they are the primary other source of game state. But there are indeed parts of game state without this safety, and I agree, they should be fixed by legislation. >> There is the question, then, of whether it was the same ruble over and >> over again or all of G's rubles. There is no particular indication of >> one over the other. However, Rule 2166 again comes to the rescue with >> "Instances of a currency with the same owner are fungible." This can, >> and in my view, should be interpreted as implying that the distinction >> between rubles of the same owner is irrelevant since they are always >> interchangeable, and that distinction is hence effectively >> non-existent. Common sense and good of the game then lead us to apply >> the simplest interpretation to the situation, which is that only one >> ruble's worth of ambiguity is created. As for CFJ 3256, it the rules >> simply do not allow for drawing this distinction with currencies. >> Moreover, by using the word "currency", it implies that, much like a >> real-life currency, in a situation like this, the actual details of >> the changing of hands are not important, hence I judge CFJ 3256 to be >> IRRELEVANT. > > > Another minor nit: A significantly more reasonable parsing of CFJ 3256 > (precisely because rubles are fungible) is "transferred [a total of] > exactly one", or to make it more explicit: I considered this, but that, too, is irrelevant. If people want me to stick to an interpretation, I'll do so if the judgement is reconsidered. -scshunt
