On Wed, 1 Jan 2020 at 07:03, Aris Merchant
<[email protected]> wrote:
> On Tue, Dec 31, 2019 at 9:41 PM James Cook <[email protected]> wrote:
>>
>> On Sun, 29 Dec 2019 at 18:59, Aris Merchant
>> <[email protected]> wrote:
>> >
>> > I intend to ratify without objection the following ~~~-delimited
>> > document (see further the notes at the bottom of this message):
>> > ~~~
>> > Effective date: Dec 14 00:15:01 UTC 2019
>> >
>> > At Dec 14 00:15:00 UTC 2019, the fora agora-official and
>> > agora-business became discussion fora.
>> > ~~~
>> >
>> > I intend to ratify without objection the following ~~~-delimited
>> > document (see further the notes at the bottom of this message):
>> > ~~~
>> > Effective date: Dec 28 01:45:01 UTC 2019
>> >
>> > At Dec 28 01:45:00 UTC 2019, the fora agora-official and
>> > agora-business became public fora.
>> > ~~~
>> >
>> > I pledge not to ratify either of those documents without ratifying the
>> > other one. I note that the documents above are technically incorrect,
>> > but that ratifying them would reduce ambiguity about what messages
>> > failed to be public under CFJ 1905.
>>
>> I object to both intents.
>>
>> Sorry to prolong this, but I'm not convinced this gets around Ørjan's
>> objection. Here are two modifications to the gamestate that could be
>> made at 00:15:01 on Dec 14 that would make the first document true:
>>
>> a) Insert two events into the historical record: a-o and a-b become
>> discussion fora. Flip both publicity switches to Discussion.
>>
>> b) Insert four events into the historical record: a-o and a-b became
>> discussion fora, then immediately after, became Public fora again.
>>
>> Both of these involve four changes (either two additions to history
>> plus two changes to Publicity switches, or four additions to history).
>> The first one is what we intend, but I'm not confident that it is the
>> unique minimal modification.
>>
>> Is there anything wrong with passing a proposals that says "Change the
>> gamestate to what it would be if a-b and a-o's publicity had been
>> switched to Discussion at time X and then switched back to Public at
>> time Y, so that none of the intervening messages on either list were
>> sent via a public forum"?
>>
>> Alternatively, I wouldn't be averse to just fixing the uncertainties
>> one by one. I don't think there are that many. A few Master switches,
>> some income earned, the state of the PM election, and whether a
>> proposal was distributed. Anything else?
>>
>> --
>
>
>
> *Sigh.* I see your point, but I’m not buying it.
>
> I CFJ, barring Falsifian, the claim “If the two documents quoted above were
> ratified, the first one would have the effect of modifying the historical
> record twice and the publicity switches of the relevant fora twice, in the
> manner stated as interpretation a above.”
>
> Arguments:
> The relevant paragraph of rule 1551:
> “When a document or statement (hereafter "document") is ratified, rules to
> the contrary notwithstanding, the gamestate is modified to what it would be
> if, at the time the ratified document was published, the gamestate had been
> minimally modified to make the ratified document as true and accurate as
> possible; however, if the document explicitly specifies a different past time
> as being the time the document was true, the specified time is used to
> determine the minimal modifications. Such a modification cannot add
> inconsistencies between the gamestate and the rules, and it cannot include
> rule changes unless the ratified document explicitly and unambiguously
> recites either the changes or the resulting properties of the rule(s). If no
> such modification is possible, or multiple substantially distinct possible
> modifications would be equally appropriate, the ratification fails.”
>
> I don’t think any reasonable reader would say that the minimal change
> necessary to make a document true involves inserting a change cancelling out
> any changes made by the document. That’s simply not what minimal change means
> in natural language. Also, if there is any interpretation in which this view
> is correct, the best interests of the game and common sense mandate that it
> be selected.
>
> End arguments.
>
> I request expedited assignment for this case. I pledge the Arbitor 5 coins if
> e assigns this case within 2 days. I pledge the first judge of this case 5
> coins if e judges it within 2 days of it being assigned to em.
>
> -Aris
Gratuitous arguments:
Repeating the two possibly gamestate changes I mentioned earlier, for context:
a) Insert two events into the historical record: a-o and a-b become
discussion fora. Flip both publicity switches to Discussion.
b) Insert four events into the historical record: a-o and a-b became
discussion fora, then immediately after, became Public fora again.
First, I agree that (b) probably isn't the (unique) minimal change,
but I do suspect it triggers this clause from R1551 "If ... multiple
substantially distinct possible modifications would be equally
appropriate, the ratification fails.".
Second, here's an even smaller change that could be made:
(c) Insert two events into the historical record: a-o and a-b become
discussion fora. Make no other changes (so their publicity switches
are still Public).
This introduces an inconsistency between the sequence of historical
events and the current value of switches, but I don't see why that
would rule it out. R1551 says the change "cannot add inconsistencies
between the gamestate and the rules", but the rules are silent on the
question of whether the current value of a switch should be consistent
with the sequence of past changes to that value.
Now you may say: "But come on! Even if the rules don't rule that out,
common sense tells us that this gamestate makes no sense and must be
ruled out." But consider the common-place effect of ratifying an
officer's report containing an incorrect value for a switch: in that
case, I think we all casually assume that the resulting minimal change
to the gamestate is that the switch is flipped to the reported,
incorrect value, even though that value is inconsistent with the
history of changes to that switch. The only difference here is that
it's the other way around: rather than the switch being flipped
without a change to its history, the history is changed without a
change to the value of the switch.
--
Side note: maybe the documents to ratify should simply say "This
document is true as of X. The publicity a-o and a-b are
{Discussion/Public}." No need to talk about changes in the ratified
document.
--
- Falsifian
_______________________________________________
Agora mailing list
[email protected]
https://listserver.tue.nl/mailman/listinfo/agora
