On 6/12/2022 12:43 PM, ais523 via agora-discussion wrote: > On Sun, 2022-06-12 at 12:22 -0700, Kerim Aydin via agora-discussion > wrote: >> The Winds Died Down recently, beginning the procedure described in >> R2658. >> >> It's pretty clear to me that all of the repeals of (1) went through, >> and I can't think of an argument that those repeals failed. > > The rule says "the following happen in order", referring to a list. If > any of them can't happen, does that mean that none of them happen? > > I think our normal precedent is to do as much as possible, but I'm not > sure why I think that, nor whether there's a rule about that, and > there's more than one reasonable interpretation. >
That's a good point. I'll definitely address it on the next draft. My first sense is that, since R105 makes rule changes rigidly sequential and therefore severable, you have to follow the "text of the rules" as long as that text exists, so if you get to a step that fails, undoing the previous steps is re-amending the rules back to what they were, which *definitely* isn't rule 105 specified. This would be the default assumption, if the rule said explicitly "if one of these fails they all fail" it would specify what happens well enough to be atomic, perhaps? I've also *generally* felt uncomfortable with any conditionals that say "if any of these steps fail they all fail" because it's a conditional based on a future state (even if it can be evaluated hypothetically in the present without any doubt), which we've generally disallowed as being ambiguous, and if interpreted with rules-text level deference can lead to more paradoxes - but that's sort of a last-ditch "good of the game" argument. -G.
