Index ← 3326 CFJ 3327 3328 → text
==============================  CFJ 3327  ==============================

    It was POSSIBLE for me to cash 'ZipZop Series G-002' when I
    attempted to do so.

========================================================================

Caller:                                 omd

Judge:                                  Ienpw III
Judgement:                              


Judge:                                  woggle
Judgement:                              TRUE


Judge:                                  woggle
Judgement:                              TRUE

========================================================================

History:

Called by omd:                          21 May 2013 01:13:52 GMT
Assigned to Ienpw III:                  27 May 2013 22:33:11 GMT
Ienpw III recused:                      09 Jun 2013 18:20:49 GMT
Assigned to woggle:                     09 Jun 2013 18:21:30 GMT
Judged TRUE by woggle:                  15 Jun 2013 23:02:21 GMT
Reconsideration requested by Murphy:    19 Jun 2013 17:05:18 GMT
Reconsideration requested by Roujo:     19 Jun 2013 17:07:33 GMT
Reconsideration requested by omd:       19 Jun 2013 19:03:58 GMT
Assigned to woggle:                     19 Jun 2013 19:03:58 GMT
Judged TRUE by woggle:                  23 Jun 2013 18:20:29 GMT

========================================================================

Request for reconsideration by :

I intend (with 2 support) to request reconsideration, suggesting
REMAND so we can get some arguments and/or evidence.

========================================================================

Request for reconsideration by :

I support.

========================================================================

Request for reconsideration by :

There is nothing out of the ordinary about this promise, but consider
how the logic actually works.  The rules state that the transfer from
the Tree succeeded if the cashing succeeded and that the cashing
succeeded if the transfer succeeded (because otherwise it wouldn't be
in my possession): c -> t ∧ t -> c.  Both "c = t = T" and "c = t = F"
are consistent, and there is no actual guidance on which way the
computation should be biased, so a priori the rules do not define c or
t as either true or false.  That is, except for game custom and
perhaps common sense as guidance: is that enough?

Does the clause about indeterminacy caused by cashing come into play?

========================================================================

Request for reconsideration by :

      If cashing a promise would lead, through its own actions or
      actions directly caused by its cashing, to a value being
      indeterminate an instant after the promise is cashed, then
      (other provisions of this or other rules notwithstanding) it
      CANNOT be cashed.

      If a promise is possessed by the Tree, any player except the
      promise's author CAN transfer it to emself by announcement, if e
      cashes the promise in the same message in which e transfers it
      to emself.

========================================================================

Judge woggle's Arguments:

I rejudge CFJ 3327 TRUE.

The relative ordering of actions in a message in Agora exists from game
custom, where we generally evaluate the clauses/sentences/etc. of a message in
the order they are written. This has been preferred to the strategy of
treating all parts of a message as simultaneous, which can easily lead to
indeterminate gamestate. The cashing promises rule (at the time this CFJ was
called), however, was clearly written with the strategy of treating a message
as a simultaneous multi-action. We can see that it prefers this interpretation
because it references what would ordinarily be other actions in the same
message. So, for the special case of interpreting these rules, we will use
this usually unAgoran theory of message timing.

Given the idea that the actions in a message are really one combined action,
the requirement for C (cash) only if T (transfer) in same message and T only
if C in same message can be read as creating a combined action of T+C for
certain types of promises. This combined action is performed by announcing T
and C somewhere in the same message (and runs into nasty issues if those
announcements of T and C are conditional, etc.).

Alternately, if we do not treat T+C as a single action, we should treat the
combination as possible because the rule's "CAN ... if" clearly indicates that
there should be some circumstances under which the cashing is possible. This
preference tells us to prefer the "true" interpretation to the "false" one.
Since the rule imagines that the message occurs simultaneously, there is no
issue about what must be evaluated first.

With the strategy for timing in mind, the rule's "CANNOT be cashed" clause is
no issue, because the "instant" after the promise has been cashed does not
occur until after the entire message is evaluated.

========================================================================