On Mon, 28 Feb 2022, Jason Cobb via agora-discussion wrote:
I present the following draft to clean up the ratification rule.
The only intended semantic change is securing all retroactive
modification, everything else is just mean to make the existing text
much more clear.
Title: Temporal Incursion Modification and Exclusion Act
Author: Jason
Coauthors: Aspen
Adoption index: 3.0
Amend Rule 1551 (Ratification) to read, in whole:
{
A retroactive change is one that changes the game's record of
past events. Retroactive changes are secured with power
threshold 3.
Ratification is purposefully defined so as _not_ to do any retroactive
changes in the intuitive sense, but only simulate their effects by
changing the gamestate in the present.
I see it as the continuation of a long tradition of keeping Agora in a
style where platonic and pragmatic interpretations of the rules lead to
the same result.
I am not sure that the above definition corresponds _either_ to what a
retroactive change intuitively is or to what ratification platonically
does. It might correspond to a pragmatic view, which is close to saying
that Agora's gamestate _is_ its records.
When a document or statement (hereafter "document") is to be ratified,
the following definitions apply:
* The publication time is the instant at which the document to be
ratified was published.
* The truth time is the instant at which the document specifies
that it was true, or the publication time if such an instant
is not specified.
* The application time is the instant at which the document to be
ratified is ratified.
Ratification CANNOT occur if the truth time would be after the application
time, or if the publication time would be after the application time.
Rules to the contrary notwithstanding, when a document is ratified,
the gamestate is modified to what it would be if, at the truth time,
the gamestate had been minimally modified to make the ratified
document as true and accurate as possible.
Ratification CANNOT occur if it would add inconsistencies between
the gamestate and the rules.
Ratification CANNOT occur if it would cause past or present rule changes,
unless the ratified document explicitly and unambiguously recites either
the changes or the resulting properties of the rule(s).
I would strongly suggest that this only restrict rule changes in the
minimal modification, and not any later changes following logically from
it. Otherwise the discrepancies between the intuitive interpretation of
retroactively fixing an old error and the platonic result could be
mind-wrecking.
Greetings,
Ørjan.
Ratification CANNOT occur if the required modification to the gamestate
is not possible or if multiple substantially distinct possible
modifications would be equally appropriate.
An internally inconsistent document generally CANNOT be ratified;
however, if such a document can be divided into a summary section
and a main section, where the only purpose of the summary section
is to summarize information in the main section, and the main
section is internally consistent, ratification of the document
proceeds as if it contained only the main section.
Text purportedly about previous instances of ratification (e.g. a
report's date of last ratification) is excluded from ratification.
The rules may define additional information that is considered to
be part of the document for the purposes of ratification; such
definitions are secured with power threshold 3.
Ratification is secured with power threshold 3.
}
--
Jason Cobb
Assessor, Rulekeepor, Stonemason
