On Thu, Jul 23, 2026 at 4:20 AM secretsnail9 <[email protected]> wrote:
> On Thu, Jul 23, 2026, 6:04 AM Aris via agora-discussion < > [email protected]> wrote: > >> On Thu, Jul 23, 2026 at 2:57 AM secretsnail9 via agora-business < >> [email protected]> wrote >> >> > >> > The following statement is hereby ratified: >> > { >> > Each attempt to grant a player a number card that occured prior to the >> > adoption of this proposal took effect as if the text "CAN once," in Rule >> > 2713 instead read "CAN once by announcement,". >> > } >> > >> > [This should work for each version of Rule 2713 that has existed. I'm >> doing >> > this instead of changing the past text of the rule because doing that >> > retroactively would have caused proposals that amended that specific >> text >> > to fail.] >> > >> >> The scope of the retroactivity here makes me nervous. I'm not saying that >> it's the wrong call, it very much might be the right one if everyone >> has been playing as if this worked and it's propagated to a wide enough >> area of the gamestate. I don't have a solid understanding of everything >> cards can affect and what sorts of conditional actions might have been >> predicated upon them. But I wanted to at least ask if any more >> conservative >> approach might be viable. Could you help me understand why you preferred >> this to an approach that focuses on ratifying a record of the current >> gamestate (potentially with card and radiance values along with any recent >> wins)? >> >> -Aris >> > > Mainly it's to prevent needing to look at 7 months of emails. > > I would prefer the simpler approach of "pretend like everything has always > been fine" because then it should catch absolutely everything. If we > instead tried to craft a new record of the gamestate by hand, using > reports, we may miss something in the 7 months of emails that was relevant > in an odd way. > > I prefer the fullness of this solution, even if the scope is a bit scary, > to the safer method that's less likely to catch every little thing. > > At worst there's some sort of card related scam that could happen in the > future, where we can see and deal with it, whereas missing something in a > 7-month period in the past would cause more invisible issues instead. > If the alternative is looking at 7 months of emails, I agree. I'd hoped that the Numerator and Illuminator would need to calculate those records anyway, so it wouldn't be redundant work. Again, I haven't been following well enough to have a great idea of the current state of affairs. My fear is that this could inadvertently invalidate what we believe the current game state to be. The type of hypothetical scenario going through my head is: 1. Player A voted "conditional: FOR if player B has given me a number card" on a referendum (or some similar arrangement involving promises). 2. Player B then attempted to give Player A a number card that was actually nonexistent because of the bug. 3. The Assessor missed the transfer and recorded Player A's vote as evaluating to PRESENT, which resulted in the referendum failing. 4. The Assessment self-ratified. The Assessor's analysis was actually correct in the old timeline, since the number card in question didn't exist and thus Player B didn't give it to Player A. So it seems like the new timeline, where Player A did obtain the card, should replace the old one, which would mean that in the new timeline, Player A's vote is different. This changes the result of the referendum in the new timeline. [1] In this case, it seems like the self-ratification of the Assessment wouldn't protect us, because we're ratifying a new thing that overwrites it. I know that specific case is unlikely to happen. It's pretty contrived. But I'm a bit worried that something else could break, or that there might be less contrived concerns I'm not thinking of. [1] This can't change the rules, since they're excluded from implicit ratification (R1551/24), but it could change the result of the decision itself. -Aris

