On Sun, Jun 7, 2020 at 8:56 PM Alex Smith via agora-discussion <
agora-discussion@agoranomic.org> wrote:
> On Monday, 8 June 2020, 04:43:19 GMT+1, Aris Merchant via
> agora-discussion <agora-discussion@agoranomic.org> wrote:
> > * EQUATE {X}, where X is a body of text, appropriate if the situation
> > is inequitable and X includes changes that would make the situation more
> > equitable, and only such changes. After the judgement has been
> > in effect for a total of 7 days, the changes specified by X are applied,
> > except to the extent they are retroactive, would directly change the
> rules,
> > or would directly create, modify, or destroy an Agoran decision.
>
> "In effect" isn't defined. This probably doesn't matter much for Moots
> (the only place where the phrase appears in the rules currently); however,
> I'd prefer something less likely to inspire disputes if you're going to
> allow the judgement to make almost arbitrary power-2 gamestate changes!
> (Note that historically, power 2→3 escalations have been very easy to come
> by.)
Agreed.
I also don't like the passive "is applied". When it comes to making changes
> that may need Power, it's important to specify who or what is doing the
> changes, so that precedent and conflicts can be easily involved. (For
> example, you could allow the judge to apply the changes by announcement if
> the judgement were assigned continuously for 7 days; then, the changes
> would happen with the power of the judge, so only unsecured gamestate could
> be affected.)
I'll go with the rule applying the changes; I want it to be able to do
power 2.0 secured things. I can however add a requirement for someone to
trigger it, so that we don't get delayed gamestate changes everyone has
forgotten about.
Last time we did equity, we had the judgement create a contract, rather
> than perform direct modifications to anything. That makes it much clearer
> what is and isn't possible, and also allows for most of the types of equity
> restoration you'd want (by forcing SHALL requirements onto people). In a
> way, it was a bit of a flop, due to being surprisingly unscammable (not
> that people didn't try, but I don't think anyone succeeded).
I considered doing that again, but decided it was less interesting,
required more obliqueness and delays, and would likely require disabling
anti-mousetrap protections I'd rather not touch.
-Aris
