==============================  CFJ 2324  ==============================

    Craig is a player.


Caller:                                 Murphy
Barred:                                 Wooble

Judge:                                  ais523
Judgement:                              TRUE



Called by Murphy:                       03 Jan 2009 15:28:21 GMT
Assigned to ais523:                     08 Jan 2009 06:54:26 GMT
Judged TRUE by ais523:                  12 Jan 2009 11:05:19 GMT


Caller's Arguments:

Craig mentioned on IRC the other day that e is the same Craig listed in
the Registrar's report as having been a player from 2002 to 2003.  By
implication, e hasn't attempted to register since then, so the 30-day
limit isn't an issue.

It's plausible that Craig intended to register if and only if eir
contract with Warrigal succeeded in assigning a referent to "Markos
Sophisticus Maximus".  It's also plausible that that contract failed
to do so; it attempts to define MSM as an office, which Rule 1006
only allows rules to do.

Warrigal has questioned Rule 1006's restriction, in light of the
general perception that contracts can successfully define dependent
actions pertaining to their internal state, despite Rule 1728's
similar restriction.  I can think of a few plausible resolutions:

  1) Dependent Contract Changes are backed by the first paragraph of
     Rule 2198, so those should be okay.

  2) Other contract-defined dependent actions:
     a) don't work, but we didn't realize it until recently.
     b) do work, because internal contract state is pragmatic based on
          whatever the parties accept.


Caller's Evidence:

Caller's evidence, part 1 of 2:  Text of the contract in question

> This is a public contract and a pledge. Warrigal CAN terminate this
> contract by announcement.
> The Badge of Markos Sophisticus Maximus is a singleton asset whose
> recordkeepor is Markos Sophisticus Maximus. Ownership of the Badge of
> Markos Sophisticus Maximus is restricted to parties to this contract.
> Initially, the Badge of Markos Sophisticus Maximus is owned by
> Warrigal. Markos Sophisticus Maximus is an imposed office held by the
> owner of the Badge of Markos Sophisticus Maximus.

Caller's evidence, part 2 of 2:  Craig's purported registration

> I suspect that MSM couldn't register before, being a player already.
> But now that it's me, and I'm not a player...
> Markos Sophisticus Maximus registers.


Gratuitous Arguments by ais523:

I always interpreted contract-defined dependent actions as being
shorthand. The Gnarlier Contract defined a switch-like object, which is
generally believed to have worked. It wasn't a switch due to not being
defined by the rules and not having a recordkeepor, but describing it as
like a switch was the easiest way for everyone to understand how it
worked. As far as I can tell, contract-defined dependent actions are
similar; they aren't actually rules-defined dependent actions and aren't
backed by the rules, but it's an unambiguous abbreviation for, for
instance, "This contract can be changed if all the same conditions are
met as would be met for a dependent action without 3 objections." So I
don't think there's a crisis here. As for whether the registration
worked, I have no idea; I don't really understand what the scam in
question is trying to do.


Judge ais523's Arguments:

This case hinges around what exactly happened when Craig sent the
message in question. The sentence in question was
{{Markos Sophisticus Maximus registers.}}

The only definition of Markos Sophisticus Maximus that has been
suggested, or that seems plausible in context, is the definition that is
given by a pledge by Warrigal (of which Craig is obviously aware, being
a party to it):
Markos Sophisticus Maximus is an imposed office held by the
owner of the Badge of Markos Sophisticus Maximus.

Now, it is clearly impossible for a contract to create offices. The
relevant precedent here is CFJ 1930, which found that when a contract
uses a rules-defined term, it is using the rules-defined meaning not the
English meaning, unless it explicitly says otherwise. An office is
defined by the rules as "a role defined as such by the rules"; so the
contract cannot create an office. So Fight Arena failed, because
creating a "secret rule" is something that although impossible,
certainly made sense. A contract obligating someone to create a secret
rule is much the same as a contract obligating someone to do any other
impossible thing; ridiculous and you'd be a fool to join, but legally
parsable. In the case of contract-defined dependent actions, meaning has
to slip a bit, because the circumstances are different. Rule 1728 allows
people to perform rules-defined dependent actions dependently. There's
nothing preventing a contract saying that something CAN be done w3o, or
whatever; however, rule 1728 doesn't give such actions any special
meaning, so the phrase just sticks around with game-custom meaning (and
I believe that it's sufficiently clear to allow the first paragraph of
rule 2198 to cause dependent-action-like contract changes to work if
they use the same wording as invokes rule 1728).

So what about contracts that purport to define an office? We first need
(per the precedent of CFJ 1930) to expand the definition:

Markos Sophisticus Maximus is a role defined by the rules to be an
office but not an elected office, held by the owner of the Badge of
Markos Sophisticus Maximus.

So we have here a case of a contract which is clearly lying, in much the
same way as a contract which says "This contract has power 4". CFJ 1933
covered the case of a contract which effectively said "This contract is
the backing document for <asset defined by another contract>"; the
judgement there was that each contract had its own internal gamestate,
and they could coexist quite happily, producing crop reports (which
contradicted each other) and disregarding each other (and incidentally,
CFJ 1923 found that in such a case, referring to a "crop" would
therefore be ambiguous unless it was clear which contract was meant from
context). Applying this precedent to the current case would imply that
the contract believed there was a rules-defined office when the rules
believed there weren't; however, it is not at all clear whether
contracts should by default be assumed to have rules-contradictory
internal gamestate. (There have been contracts with internal gamestate
different from the rules, mostly external nomics which somehow ended up
as Agoran contracts; for instance, Flapjack specifically stated that its
gamestate used different definitions from those in the rules, and had a
convention to show which meaning was used where. However, this is
different from defining false facts to be true.) So this precedent would
encourage TRUE; but it is shaky and not entirely applicable.

However, there is a more compelling argument which also indicates TRUE.
If the impossible part of the above definition is deleted, we get
Markos Sophisticus Maximus is a role, held by the owner of the Badge of
Markos Sophisticus Maximus.
The part of the contract's definitions that purports to contradict the
rules - which is whether Markos Sophisticus Maximus is rules-defined -
is irrelevant here. The definition as a role is (per CFJ 1930, and per
common Agoran language definitions) part of the definition of something
as an office; and we should not ignore the entire sentence just because
it contradicts the rules, but instead merely the part of it that
contradicts the rules. So it is clear that, regardless of how conflicts
with the rules are resolved, that from the point of view of the
contract, Markos Sophisticus Maximus was Craig at the time e sent the
message in question. It is very common shorthand (and r754-unambiguous)
to use a term defined only in a contract as shorthand for the definition
used by the contract; for instance, "I transfer a prop to the SoA" would
be taken as expanding to "I transfer a prop to the player that the
Agoran Agricultural Association defines to be the Secretary of
Agriculture", as that is the only plausible referent for SoA in the
current Agoran gamestate; and likewise, "Markus Sophisticus Maximus
registers" expands to "The player which Warrigal's recent unnamed pledge
defines to be the holder of Markus Sophisticus Maximus registers". (It
is a rules-defined property of offices that the name of the office can
be used to refer to its holder, although even without the definition it
would be a pretty clear abbreviation (when recently attempting to
transfer an asset to an office itself, not its holder, the player who
attempted to do so had to officially clarify this in the message); and
likewise, that is the only plausible meaning in this case.) So in other
words, Craig's message expanded to "Craig registers", and it was an
unambiguous abbreviation for that per rule 754. Craig's announcement was
therefore indeed "announcing that e registers" per rule 869, so the
message in question succeeded in making em a player. As I know of
nothing that might have deregistered em in the meantime, I conclude that
Craig is indeed a player (now and at the time the CFJ was called), and
therefore I judge CFJ 2324 TRUE.