On 7/24/2020 1:38 PM, ais523 via agora-discussion wrote:
> On Fri, 2020-07-24 at 08:44 -0700, Kerim Aydin via agora-discussion wrote:
>> On 7/24/2020 8:05 AM, Kerim Aydin wrote:
>>> On 7/24/2020 2:40 AM, N. S. via agora-business wrote:
>>>> So there's precedent that you can basically take a shortcut to do many
>>>> actions (a loop of them) as long as it is reasonable that you could perform
>>>> those actions by writing them out.
>>>>
>>>> Therefore, I call N CFJs with the statement "This is a CFJ" where N is the
>>>> number of CFJs that it would be reasonable for me to fully write out.
>>>
>>> I do not consider this a clear specification of an announcement for the
>>> starting of any cfjs.
>>
>> Just adding: the main reason is that citing a memory of a subjective and
>> possibly context-specific precedent/custom is insufficient to determine N.
>
> There have been a few situations in the past where I considered it
> borderline whether or not an abbreviation was reasonable, so I expanded
> it to make sure that it would be acceptable. (This may have lead to
> unreasonable emails, but the list accepted them.)
>
> One related issue that I've been noticing a lot recently is players
> attempting to take actions conditional on gamestate (via saying "if X,
> I do Y"); I'm not sure if that's ever been tested, and it intuitively
> feels to me like it shouldn't work (I prefer the construction "I
> attempt to do Y, this will fail unless X", but of course it isn't as
> generally applicable). However, I think I'm in a minority on this.
It's actually an interesting slippery slope/random walk. If you take
indirect references like "all my coins" as conditionals, this is something
that failed in 2001 in CFJ 1307[0], and generally succeeds today, all
based on interpretations of the common definition of 'specify' (and no
actual rule changes). I think it's been a history of gradual loosening
("if attempt X succeeded, why shouldn't only slightly less clear X'
work?") accompanied by occasional hard checks when someone pushes it (I
can't find the "large but finite loops fail" CFJ, does anyone know which
one that was?)
Recently, several of the loopy contract things would be things that I
would (as an officer) intuitively fail, but the officers recording it have
been fine puzzling out fairly complex conditionals. However, there are
still limits, Judge R. Lee found in CFJ 3819[1] (again based on a
definition of 'specify') that bidding "N+1 coins where N is the previous
high bid" failed to 'specify an amount' of a currency.
We also rejected hard checks - sometime in 2017-2018 there was a proposal
(Alexis I think) to explicitly ban all conditionals, which was voted down,
so that shows legislative acceptance/preference for at least some level of
conditional allowance.
https://faculty.washington.edu/kerim/nomic/cases/?1307
https://faculty.washington.edu/kerim/nomic/cases/?3819
-G.
