Index ← 3207 CFJ 3208 3209 → text
==============================  CFJ 3208  ==============================

    A rule-defined entity's name generally CAN be initialized to be the
    same as another rule-defined entity's name.


Caller:                                 omd

Judge:                                  ais523
Judgement:                              FALSE



Called by omd:                          05 May 2012 02:16:06 GMT
Assigned to ais523:                     26 May 2012 01:55:49 GMT
Judged FALSE by ais523:                 01 Jun 2012 23:23:08 GMT


Caller's Arguments:

This pair of annotations, which I just realized I forgot to
upload when I wrote them a month ago:

[CFJ 2981 (called 21 March 2011): An instrument is not "modified" when
an initial value is set for an attribute.]

[CFJ 2945 (called 22 December 2010): An instrument is "modified" when
an attribute is removed.]

I think the idea that initialization is not a "modification" or
"change" is bogus.


Gratuitous Arguments by Murphy:

The annotations are not quite right (though, to be fair, my judgement in
CFJ 2981 was also insufficiently clear).  Rule 2140 (c) doesn't talk
about modifying an instrument, it talks about modifying an aspect of an
instrument; and the idea behind CFJ 2981 was that you can't modify
something that doesn't yet exist.  (Rule 2140 now says "set or modify",
but "set" was added after CFJ 2981.)

In the situation described in the statement of this case, if the second
sentence of Rule 1586 doesn't apply, then the first sentence does, thus
negating the "another" part in the process.  I'm not sure what judgement
is appropriate for that.

Incidentally, I could have judged CFJ 2981 the same way for a different
reason, interpreting that the Rulekeepor assigns the ID number to be
set but Rule 2141 actually sets it.  (This is now moot, as Rule 2141 now
declares precedence over Rule 2140.)


Gratuitous Arguments by omd:

Erm, you're right, but that applies to both cases.  In 2945, G.
said "more generally it's worth stating as a precedent that removing a
property from an entity does in fact count as a "change" of the
property for that entity".


Judge ais523's Arguments:

To start with, there's plenty of precedent that modifying an attribute
of something, and creating something in a particular state, are separate
occurrences. The Threat Trombones scam (before the start of Murphy's CFJ
archive so I can't easily give numbers) relied on this fact, although it
failed for unrelated reasons. I had CFJ 2595 in particular in mind,
although it doesn't seem quite directly relevant (because the rules in
question had equal power, so the "changes are secured" phrase was
irrelevant and so wasn't mentioned in the judgement). Likewise, the
argument that creation is the same as modification could have been
brought up in CFJ 1936 (and probably would have been if people thought
it was viable, because it would have been a simple way to block a scam).
As such, the clause of rule 1586 against changing a rule-defined entity
to have the same name as another is irrelevant here.

However, rule 1586 also holds that two attempts to define rules-defined
entities with the same name define the same entity. Attempting to create
a rules-defined entity with the same name as another (i.e.
"initializing" its name to clash with the name of another entity)
clearly triggers this cause, causing the rules to define the same entity
rather than two different entities. As such, the "another" part of the
statement cannot hold; trying to initialize a rules-defined entity's
name to be the same as another rules-defined entity's name fails because
the two resulting entities are, by definition, the same (and if their
definitions contradict each other, then precedence between rules will
come into play). (Possible exception: if a rule that takes precedence
over 1586 insists on creating two different entities with the same name,
it will of course succeed. But I don't think any rules are trying to do
that at the moment.)

So indeed, initialization is not modification, there's plenty of
precedent for that. (You can create a player in a registered state
without registering it, etc.; thus, the Registrar's reports of golems
"registering" are a bit misleading, Golems can't register because they
can't exist in a non-registered state.) But you can't create two
rules-defined entities with the same name either.

I judge CFJ 3208 FALSE, and intend, with 2 support, to make CFJ 3208