Life In 19x19

I also think that is a valid example of moonshine life, and should ideally be treated as such (hence J89 flaw in not practically doing so). With Chinese rules, I wouldn't bet on whether a given referee would recognize this or rule a triple ko. The problem is those rules never even attempted a logical definition of which repetition is moonshine and which is real triple ko, so it is probably up to the individual judging it.

santo wrote:If I remember correctly, the 1979 WAGC rules say that "in a mannenko like this, the player that can fill the ko and create seki MUST do so at the end of the game", so in that case none of this applies. But the 1989 Japanese rules do not have any such special ruling for this as far as I know (the 1979, like the 1949 rules, have lots of special-cases and no "general principle").

I correct myself: the WAGC 1979 rules just say (emphasis in upper case is mine):

"(4) In a thousand-year ko, if neither side is willing to start the ko, the player WHO IS ABLE by capturing and connecting to make the position a seki shall do so."

So it is not at all clear to me that they would apply to the situation at hand (because the player is not able, as his opponent will use infinite ko threats to deny such opportunity). Thus it seems that versions prior to the 1989 rules remained ambiguous about this.

I think the WAGC explicit mannenko mention was a J49 remnant, from the Segoe-Takahashi dispute, but probably not really necessary anymore. WAGC rules were before the pass-for-ko rule, but after passes become full moves, and the latter seems enough for normal mannenko resolution (OC moonshine-in-seki is different).

Upon thinking about this position again... I now think that it has quite an important difference with the standard moonshine life. So, since there is no "official" definition of what is moonshine life, it is open to interpretation whether this difference is enough to make it not moonshine life, and thus enough that the precedent in the Chinese rules no longer apply (Chinese rules show examples of the "standard" moonshine life explicitly, saying that it should be considered dead and the ko unfightable due to the repetition).

The difference is that standard moonshine life has the following fairly strong property (which we might choose to take as the definition of moonshine life, not being there an official definition): The defender can avoid their group being captured, if they have an infinite supply of ko threats, which can be used indefinitely to establish an infinite sequence. If however, the defender is given an infinite supply of ko threats, but is instructed to "resolve" the position, so that after playing for as long as they want under these conditions (being granted the power of infinite ko threats, a power which their opponent does not share) they have the obligation to eventually say "ok, it is resolved now", and give up their superpower, then the group cannot live.

This basically boils down to the fact that in a (standard) moonshine life position, there is absolutely no way for the defender to create a new position achieving life independent of ko threats, EVEN if given INFINITE, ARBITRARILY LARGE ko threats as a help in doing so. There is simply nothing to do in the standard moonshine life examples: the one and only possible thing to do in the position is to keep the ko open, which can only be done while there are ko threats.

In a mannen ko however, this is not the case: if white is given large enough ko threats, they can simply turn the mannen ko into a direct ko, and win the ko fight. After that the position is resolved.

The argument would then be that in the diagram in the original post, the lone white stone is "more alive" than in standard moonshine life: having large ko threats only temporarily would be enough to avoid its capture, while in standard moonshine life that would never be enough, and only a PERMANENT supply of ko threats can avoid capture.

Yet another perhaps simpler formulation of the same idea:

Imagine that you play the local situation with the added rule that:

Scenario 1) The defender can ignore the ko rule and immediately retake any ko, ANY TIME that they want. The attacker can pass or play in order to lift the ko ban.

Scenario 2) The defender can ignore the ko rule and immediately retake any ko, AT MOST N times. The attacker can pass or play in order to lift the ko ban.

In the standard moonshine life position, the defender cannot be captured in Scenario 1, but the defender can be capture in Scenario 2, no matter how large N is. In the mannen ko example however, white can win the mannen ko in Scenario 2 even for N=1.

As long as W can go for the direct ko, it is not moonshine life oc.

But iirc you assumed a case when the double ko is not big enough to go for direct ko with it (at least for W), only leaving the status of W corner ko stone in question.