On Sun, 9 Aug 2009, Ed Murphy wrote:
>> JUDGE'S PROTO-ARGUMENTS:
>>
>> Let's start with the authorizing agent: R1728/24 (power=3) reads in
>> part:
>> A person CAN perform a dependent action authorized by a contract
>> as if that contract were a rule, provided that the above
>> requirements are otherwise met, and that the effects of that
>> action are restricted to altering entities and/or attributes
>> whose existence depends on that contract.
>
> I think this can actually be shot down based on that last restriction,
> as the new rule (once created) would continue to exist independently
> of the contract.
I think this holds some water, perhaps more than my argument. And
it's much simpler.
R1728 allows Contract-actions as long as:
the effects of that
action are restricted to altering entities and/or attributes
whose existence depends on that contract.
How shall we determine what attributes/entites' existence "depends on"
a particular contract? I think it's a simple test, but not as simple
as the caller claims.
1. If something in a contract is not regulated by Agora (see Regulation
Regulations) then it meets the test of being contract-dependent.
2. If something, in general class, *is* regulated by Agora, then Agora
must explicitly delegate all reasonable aspects of governing an instance
of that class to the contract (e.g. a contract-defined asset).
3. Taking the contract as a "natural" contract that consents to be
governed by Agora (historically what current contract rules evolved from)
if the entity in question could have a generally independent existence
using a minimal set of "common definitions and language" if Agora ceased
to exist, then it could also be contract-defined.
An Agoran rule to be created by a contract does not meet any of these
tests. For (1) it is obvious that rules are Regulated. For (2), even
though the proposed rule would cease to exist if the contract ceased to
exist, while it exists it is strongly *directly* governed by multiple
rules. (3) differentiates Rules from, for example, Asset classes. For
an asset class defined by a contract to exist outside of the ruleset, the
contract would have to import a minimum of assumptions that are common-
language assumptions (what an asset is) so a contract-defined asset makes
sense. While for a rule, with no real meaning outside the ruleset, a
contract-defined rule does not.
This is specifically a gray level involving "level of delegation"; how
much of a Rules-regulated entity's regulatory authority must be delegated
to a contract before the entity is "wholly defined" by a contract?
A contract-defined asset meets that test as the rules specifically
delegate authority for such to contracts.
An Agoran rule that a contract would create does *not* meet the test.
