I'm considering a motion to reconsider this case, and would like comments.
Reasons: ID Numbers are nowhere explicitly defined. So what's an ID
number? It's a number used to ID a rule. Once a SLR has been published a
few times, it's clear that there's only one number that identifies a
particular rule. Even if a rulekeepor hasn't assigned a rule ID, a
statement "amend Rule XXXX" would meet R105's "Any ambiguity in the
specification of a rule change" clause, as long as SLRs around the time of
the change have listed the number and nobody's questioned it. Therefore,
those "unassigned by the rulekeepor" numbers are still numbers that ID a
proposal, and meet the common definition of ID numbers.
Without citing this a strict precedent, CFJ 1358 has a similar principle,
that the "name" of something, if not strictly set, can be what we in
common law come to call something, if there's strong evident consensus on
that.
Further, except in rare circumstances, the process is entirely
deterministic. Anyone can look at a proposal and the current SLR and
*know* what the next number would be by common convention, before the
rulekeepor acts. If the rulekeepor can't officially "assign" a number, e
can still in a practical sense confirm the convention by publishing an SLR
with a new number to confirm our common law consensus. This actually is a
better regulation of rule numbers. If the rulekeepor attempts to deviate
from the known pattern, it doesn't reach consensus/ambiguity standards and
fails.
This implies that ID numbers can't be set by the rulekeepor not because
they don't exist, but because they're predetermined by common law (except
in rare cases) and the rulekeepor can't affect that.
This would lead to TRUE when asking whether a specific rule ID (that came
about in the expected sequence) worked, but FALSE when asking whether it
was assigned by the rulekeepor.
On 2/4/2021 4:24 AM, Gaelan Steele via agora-business wrote:
> The question at hand is whether the Rulekeepor assigning rule numbers works.
> The caller argues that it doesn't, because the rules specify that e CAN do so
> but does not provide a method.
>
> This appears to last have been litigated in CFJ 2981.
>
> It was judged TRUE (i.e. rule numbers work) on a loophole in the wording. A
> follow-up proposal. P6992 by Murphy and omd, removed that loophole and
> attempted to make rule numbers work without it. The relevant parts of rule
> 2141 (now /14) haven't changed since.
>
> So legislative intent is very explicitly for this to work. Presumably, the
> authors expected this to work as follows (quoting from 2141/14):
>
> - "However, rules to the contrary notwithstanding, the Rulekeepor CAN set
> rule aspects as described elsewhere in this rule."
> - "Every rule shall have an ID number, distinct among current and former
> rules, to be assigned once by the Rulekeepor."
> - Therefore, the Rulekeepor CAN "assign" ID numbers to rules.
>
> The question, then, is whether this sufficiently specifies a method for
> setting the ID number. There's an argument to be made that the Rulekeepor
> "assigns" the ID when e publishes a ruleset containing that number. It's
> certainly tempting to go with the interpretation that makes the rules work,
> but I don't think I can justify this interpretation, for a few reasons:
>
> First, it's far from the only interpretation. Other reasonable readings of
> the ruleset would be that the rule gained the ID number as soon as the
> rulekeepor decides on the number, or that e must explicitly give rules
> numbers by announcement. It's hard to justify picking this interpretation
> over the others.
>
> Second, and more importantly, Agora has moved in recent times towards
> assuming CANs without methods do not work. This is in many ways a matter of
> game custom, but there have been some attempts to codify this custom, so I'll
> be considering legislative intent heavily here.
>
> This shift was codified in P7928, which caused Rule 2125/10 to read, in part:
> {
> A Restricted Action CAN only be performed as described by the Rules, and only
> using the methods explicitly specified in the Rules for performing the given
> action.
> }
>
> In a comment, the proposal explains that, after it's passing, "in general,
> "by announcement" is NOT implied". Sounds like we've got pretty clear
> evidence that this is the policy, right? Wrong. Enter everybody's favorite
> proposal, Statutory Instrumentation (8354). It rephrased the relevant
> provision to "including by limiting the methods to perform that action to
> those specified within it", notably dropping the word "explicit". The
> question then, is whether this was an intended weakening, or simply an
> insignificant rephrasing.
>
> Alexis, the author of 8354, was certainly aware of the significance of that
> word "explicit"; right in the middle of drafting of that proposal, e issued a
> judgement in CFJ 3793 that discusses at length the meaning of that word in
> 2125 (finding that, in a similar situation to the one we're dealing with
> today, an implied method was not enough).[^1] There was also a lot of
> discussion of this issue by other players at the time. However, there seems
> to have said anything about changes to this provision in Statutory
> Instrumentation, so it seems unlikely that Alexis intended to make a change
> here.
>
> Therefore, I find that game custom and the rules are clear that CANs without
> fairly explicit methods don't work. I find FALSE.
>
> So, what are the implications of this?
>
> Rule IDs are significant for two things: referring to rules, and last-resort
> precedence. For the former, I don't think there's an issue - each rule only
> has one purported number, so the references remain unambiguous and clear. For
> the latter, it's more of an issue, but I'm not aware if we've had any
> situations where this is relevant lately.
>
> [^1]: also, that CFJ has this magnificent sentence, which I can't not quote:
> In conclusion, this is a typical example of the rules say I do without
> saying how, therefore I do, which has plagued Agora for a long time
> but possibly not for as long as I say I do, therefore I do has.
>
> Gaelan
>
> ---
>
> Evidence:
>
> Proposal 6992 (Democratic, AI=3.0) by Murphy
> (coauth: omd)
> Fix rule numbers
>
> Ratify all rule ID numbers in the document purported to be the
> Short Logical Ruleset and published on or about Thu, 3 Mar 2011
> 14:40:44 -0500.
>
> Amend Rule 2140 (Power Controls Mutability) by replacing "modify"
> with "set or modify".
>
> Amend Rule 2141 (Role and Attributes of Rules) by replacing this
> text:
>
> Rules have ID numbers, to be assigned by the Rulekeepor, and are
> strictly ordered.
>
> Every rule shall have a title to aid in identification. If a
> rule ever does not have a title, the Rulekeepor shall assign
> a title to it by announcement as soon as possible.
>
> For the purposes of rules governing modification of instruments,
> the text, power, ID number, and title of a rule are all
> substantive aspects of the rule.
>
> with this text:
>
> Rules have ID numbers, to be assigned by the Rulekeepor.
>
> Every rule shall have a title to aid in identification. If a
> rule ever does not have a title, the Rulekeepor SHALL assign
> a title to it by announcement as soon as possible.
>
> For the purposes of rules governing modification of instruments,
> the text, power, ID number, and title of a rule are all
> substantive aspects of the rule. However, rules to the contrary
> notwithstanding, the Rulekeepor CAN set rule aspects as described
> elsewhere in this rule.
>
> ---
>
> Proposal ID: 7928
> Title: no we can't
> Adoption index: 3.0
> Author: G.
> Co-author(s):
>
>
> Amend Rule 2125 (Regulated Actions) by replacing:
>
> Restricted Actions CAN only be performed as described by the Rules.
>
> with:
> A Restricted Action CAN only be performed as described by the Rules, and
> only
> using the methods explicitly specified in the Rules for performing the given
> action.
>
> [this over-arching protection means in general, "by announcement"
> is NOT implied. I just want to put this in place and absolutely
> clarify the ruleset if it passes, and we can add the MMI change later
> if desired].
>
> ---
>
>
