How about "a player CAN object to an intent by announcement"
Then, we could define Without N Objections as "if less than N players have objected"
Likewise, "a player CAN support an intent by announcement"So that With N Support could be defined as "if N or more people have supported"
Defining objecting and supporting as *actions* instead of a *state* clarifies this quite a bit.
On 11/03/2018 12:41 AM, Gaelan Steele wrote:
I think we’ve had a lot of trouble with determining exactly how objecting and supporting work, because unlike many parts of the rules, they are based on some announcement having happened in the past instead of the actual state in the present. By splitting the act of having objected and the state of being an objector, we prevent other things from having to deal with that complexity, and we don’t have to deal with situations like that CFJ we just got where it’s not clear whether the rule cares about having objected or actually making a valid objection—you just have to check the opinion switch. That said, I do think the switch is a bit awkward—is there a good way to keep the “objecting is an announcement that changes state” structure without a switch? GaelanOn Nov 2, 2018, at 9:37 PM, Aris Merchant <thoughtsoflifeandligh...@gmail.com> wrote: I agree. The fact that a feature exists is not per se a reason to use it. Gaelan, what advantages do you see in your revised implementation? -Aris On Fri, Nov 2, 2018 at 7:31 PM Reuben Staley <reuben.sta...@gmail.com> wrote:Switches? Please no. Switches are multi-purpose, but they are not all-purpose. I would not vote FOR this proposal. The way we currently define it already works well enough. On 11/02/2018 06:07 PM, Gaelan Steele wrote:Huh. A proto: DEPENDENT ACTIONS A dependent action is an action that a rule states can be performed by one of the following methods: - Without N Objections, where N is a positive integer. ("Without Objection" is shorthand for this method with N = 1.) [removed the cap of 8—there’s no reason for it] - With N Support, where N is a positive integer. ("With Support" is shorthand for this method with N = 1.) - With N Agoran Consent, where N is an integer multiple of 0.1 with a minimum of 1. ("With Consent" is shorthand for this method with N = 1.) - With Notice. - With T Notice, where T is a time period. A person (the initiator) CAN, by announcement, create an intent (syn. “intend”) to perform a dependent action, unambiguously specifying the action and method (including the value of N or T, if applicable). The announcement SHALL be made conspicuously and without obfuscation; otherwise, it is INEFFECTIVE. [Currently, it’s just IMPOSSIBLE to hide an intent. I make it ILLEGAL too—not sure it should be.] If an intent is more than fourteen days old, it ceases to exist. OPINIONS ON DEPENDENT ACTIONS (new rule. Power?) Opinion is an untracked [active?] player-intent pair switch, with possible values Neutral (default), Supporting, and Objecting. A player CAN flip any of eir opinion switches to any value by announcement, except that they CANNOT flip it to Supporting or Objecting if it has previously held that value. For a player to support an intent is for em to flip eir opinion switch for that intent to Supporting. A player whose opinion switch for an intent has a value of Supporting is said to be a supporter of that intent. For a player to object to an intent is for em to flip eir opinion switch for that intent to Objecting. A player whose opinion switch for an intent has a value of Objecting is said to be an objector to that intent. For a player to withdraw support or objection is for em to flip eir opinion switch for that intent to Neutral. RESOLVING DEPENDENT ACTIONS (retitle 2124, Agoran Satisfaction) A player (the performer) CAN, by announcement, perform a dependent action for which an intent exists, subject to the following conditions: - If the method is Without N Objections, With N Agoran Consent, or With Notice, the intent is at least four days old. - If the method is With T Notice, the intent is at least T old. - At least one of the following is true: - The performer is the initiator. - The initiator was authorized to perform the action due to holding a rule-defined position now held by the performer. - The initiator is authorized to perform the action, the action depends on support, the performer is a supporter of the intent, and the rule authorizing the performance does not explicitly prohibit supporters from performing it. - Agora is Satisfied with the intent, as defined below. - If a set of conditions for the performance of the action was given in the announcement of intent to perform the action, all those conditions are met. Doing so called “resolving” the intent. When an intent is resolved, it ceases to exist. The conditions for Agora to be Satisfied with an intent depend on the method of the dependent action: - Without N Objections: Agora is satisfied with the intent if it has fewer than N objectors; - With N support: Agora is satisfied with the intent if it has N or more supporters; and - With N Agoran Consent: Agora is satisfied with the intent if the ratio of supporters to objectors is greater than N, or the action has at least one supporter and no objectors. - With Notice or With T Notice: Agora is always Satisfied with the intent. This shouldn’t change any function—I just changed some things to take advantage of modern Agoran concepts. GaelanOn Nov 2, 2018, at 11:53 AM, Kerim Aydin <ke...@u.washington.edu> wrote: I think it's just long, long history combined with "it generally works, has gone through a lot of CFJs, and messing with Objections is dangerous" so no one's dared/bothered with a big refactor. That's not bad or good, just no one has tackled it since - I just checked - 1999, when it first came to have this general procedure (it may have been in keeping with rule style in 1999, I dunno). If we dig into it there may be some good reasons to keep some of the things "weird", but the whole structure isn't sacred or anything (I think!) On Fri, 2 Nov 2018, Gaelan Steele wrote:Is there a reason the dependent action rules are so weird? Seems like they need a refactor to use more “normal” mechanisms. GaelanOn Nov 2, 2018, at 10:45 AM, Kerim Aydin <ke...@u.washington.edu> wrote: [The easy one first] On Mon, 29 Oct 2018, D. Margaux wrote:I CFJ barring twg: “If in the last 48 hours the Speaker has objected to any announced intents to Demand Resignation, then Agora is not satisfied with those intents and an attempt to Demand Resignation would be INEFFECTIVE.”CFJ 3679 Judgement The 'Demanding Resignation' procedure in R2472 is entirely by announcement: a player can do it "by announcement" if e "announced intent" in a specific time window. The only connection with R1728 or R2124 is a procedural similarity, and the use of the term "intent" which is used in the common-language sense ("intent" itself is not given an explicit, reserved definition in the rules). As for the procedural similarity, it is not exact: in particular, Agora does not need to be "satisfied" with the intent for the action to be performed, so R2124 does not apply to this procedure. FALSE. [This second one was tough: I started out thinking the opposite to this judgement then realized the implications of that. So this could certainly use some review].I CFJ barring twg: “A player CAN object more than once to a dependent action if e has not ever withdrawn an objection to that dependent action.”CFJ 3680 Judgement R2124 reads in part:An Objector to a dependent action is an eligible entity who has publicly posted (and not withdrawn) an objection to the announcement of intent to perform the action.This actually has two specific actions: 1. publicly posting an objection; 2. becoming an Objector to a dependent action. and you do #2 by doing #1. Later in the Rule, there are consequences for whether or not the Speaker has "objected to" something in the past 24 hours. But does "objecting to" something equate to #1, "publicly posting an objection", that can be done more than once, or does it equate to #2, "becoming an Objector?", which can only be done once? (because after you've become an objector, posting another objection doesn't change that status). At first glance it seems obvious that "objecting to" is equivalent to "publicly posting an objection", just due to straightforward language usage. However, in a parliamentary context, "objecting to" is a strict procedural term that only counts if it changes the legal status (i.e. it only counts if it makes you an Objector, otherwise you're just out of order or repeating yourself or something). So in an outside-of-Agora context, either interpretation is possible. While common-language generally implies the first, and we are biased towards common-lanugage interpretation in general, the context of the Rules in question - a formal approval procedure - means the parliamentary context is also plausible. What it comes down to, inside Agora, is Regulation. Quite simply, "publicly posting an objection" to something is wholly Unregulated. It's a speech act, in the real world, if you say you've objected to something, you've done it. Nothing in the rules limits or constrains your ability to do it, you can do it for anything - you can publicly post objections to proposals, rules, moves, anything you like. And it's VITALLY IMPORTANT that it be unregulated in the current system, if Dependent Actions are to work. Because the rules nowhere explicitly say "a player CAN publicly post an objection". So if "publicly posting an objection" was taking to be regulated, it would be IMPOSSIBLE to do it because the rules don't say how you CAN do it (R2125). However, "becoming an Objector to a dependent action" *is* regulated, because R2124 states how you do it (by publicly posting an objection). So what about "objecting to" something - regulated or unregulated? Well, R2532 says you can't do it for Zombies, and R2124 says you can't do it after you've withdrawn the same sort of response. So it is Regulated. So if we equate "objecting to" with "publicly posting an objection", the thing that needs to stay Unregulated to work, becomes Regulated. And that means publicly posting an objection generally CANNOT be done, because the rules don't supply a method for doing so. And the whole basis of dependent actions breaks down - dangerously. Not a good idea. Equating "objecting to" to "becoming an Objector" doesn't have this issue. They're both Regulated, and everything works fine. So since both readings of "objecting to" are plausible (one reading in common language, one in parliamentary domains), and one reading causes massive danger and breakage, we prefer the reading that doesn't. So "objecting to" in R2124 means "becoming an Objector". Which, once you've done it once, cannot be done a second time. So once a Speaker becomes an Objector, further public postings of eir objections do not force a delay in the action, because e has *already* objected to it (become an Objector) once. FALSE.-- Trigon