On Tue, Sep 30, 2008 at 8:30 PM, Ian Kelly <[EMAIL PROTECTED]> wrote:
> On Tue, Sep 30, 2008 at 6:15 PM, comex <[EMAIL PROTECTED]> wrote:
>> 3. bayes.py is a script whose purpose is to act on behalf of Bayes in
>> a generally autonomous way, controlled by the parties to this
>> contract.
>> 4. bayes.py CAN cause Bayes to act by sending a message from [EMAIL
>> PROTECTED]
>
> How can a non-person act on behalf of a person?
Although Goethe seems to think that non-persons unilaterally cannot
send messages, scripts are definitely capable of sending messages "on
their own", and I don't know why we should assume the legal fiction
that they can't. Can you suggest a better wording? The PerlNomic
clause:
4. The PerlNomic Partnership shall act by using the mechanisms of the
PerlNomic game to send messages to the appropriate Agoran fora. This
is the only mechanism by which the PerlNomic Partnership may act.
also requires that some non-person ("the mechanisms of the PerlNomic
game") be legally able to send messages.
Which reminds me...
Bayes is, in fact, the realization of my comment in that discussion:
> Heh, maybe I should set up a script that plays on behalf of a
> partnership without any human intervention, voting on proposals in
> some odd manner (FOR if it contains an odd number of lines or the word
> "repeal"). Even better if the partnership were, say, elected CotC...
although it's mostly ehird's code. While I don't think it's going to
try being the CotC anytime soon, it has a highly interesting method of
voting on proposals. Can you guess what it is?
Now, when person X makes the PNP distribute a set of proposals, and a
cron job makes Bayes vote on them, who is the Executor of the voting
message? X, ehird, me? ehird owns the server on which bayes.py runs,
but I have a shell account there with permission to modify it (well,
will be able to tomorrow).
I think we should just prohibit non-first-class persons from
performing dependent actions. There are no circumstances I am aware
of in which they need to perform them...
