On Mon, 2017-07-24 at 11:12 -0700, Kerim Aydin wrote: > All fine points! > > So do you think we should get rid of the R217 protection all together > rather than amend it? Because I think it's pretty broken right now.
I agree it's probably broken. (Compare it to 1698, our most basic and important protection, which incidentally appears to be just a couple of pips of Power too low due to recent Power creep. That goes into a lot of detail about what happens if a gamestate change, e.g. the loss of all assets required for pends without any way to gain them, would make it impossible to change the rules.) > First, because protecting the "initiation" is useless if you're trying > to ensure some form of "justice". And if we're not trying to protect > justice - and I'm not saying that we have to - what are we trying > to protect exactly? IMO, this is all a probably accidental side effect of the Great Rule 101 Experiment (which I don't think ever actually failed; we halted it because we felt like a change, not because it wasn't working), rather than anything that really happened deliberately. I know you know the history behind the experiment in question, having been involved both in how it started and in how it ended, but because there are a bunch of players who've joined since, I'll summarise it so that other people reading the thread can have context. The basic idea was to take a bunch of anti-scam protections in the rules that already existed (things like "players can't be added to a contract without their consent", that sort of thing), compile them all together into a single rule, then put its power up higher than would normally be wise (in addition to numbering it 101, thus causing it to automatically win any Power ties with Power 3 rules and incidentally moving it to the top of the ruleset). I'm not sure if it was deliberate or not, but the inevitable effect of this was to allow people to make "constitutional" arguments, ranging from solid to dubious, to screw up the activity of other rules because they were infringing on eir rights. (In addition to modelling a part of real-life legal systems that isn't often simulated in nomics, it also added a very interesting layer of protection relating to scams; from my point of view as a Scamster, it was relevant that it didn't say "scams are disallowed" or anything like it, but rather banned several things which are common parts of scams but not /necessary/ parts of scams. Thus, dodging around it was pretty fun.) "You can't do that because of rule 101" was a fairly common thing to see in anti-scam attempts. "I can do this because of rule 101" was also fairly common for people on the CuddleBeam end of the scam spectrum. Attempted rule 2125 scams are common enough even nowadays. You can probably imagine what they were like when the relevant rule looked like this (this quote is from proposal 4866, distributed August 20, 2006, i.e. the very first part of the experiment): > Rename Rule 101 to "Agoran Rights and Privileges" and amend it to > read: > > The rules may define persons as possessing specific rights or > privileges. Be it hereby proclaimed that no binding agreement > or interpretation of Agoran law may abridge, reduce, limit, or > remove a person's defined rights. A person's defined privileges > are assumed to exist in the absense of an explicit, binding > agreement to the contrary. This rule takes precedence over any > rule which would allow restrictions of a person's rights or > privileges. > > i. Every person has the privilege of doing what e wilt. > > ii. Every player has the right to perform an action which is not > regulated. > > iii. Every person has the right to invoke judgement, appeal a > judgement, and to initiate an appeal on a sentencing or > judicial order binding em. > > iv. Every person has the right to refuse to become party to > a binding agreement. The absense of a person's explicit, > willful consent shall be considered a refusal. > > v. Every person has the right to not be considered bound by > an agreement, or an amendment to an agreement, which e has > not had the reasonable opportunity to view. > > vi. Every player has the right of participation in the fora. > > vii. Every person has the right to not be penalized more than > once for any single action or inaction. > > viii. Every player besides the Speaker has the right to deregister > rather than continue to play. > > Please treat Agora right good forever. When the experiment ended, the various protections were "dispersed" back into whichever rules seemed appropriate. I'm not entirely sure how the "invoke judgement" part of 101(iii) ended up in rule 217, but it's at least not an obviously inappropriate place for it (and it prevents it being accidentally amended out of the CFJ rules when they change, plausible as they change fairly often). Anyway, the point is that the "invoke judgement" part of 101(iii) wasn't really the main focus of the rule, when it was there. 101(iii) and 101(vii) were mostly in practice about maintaining the rights of accused criminals (when seen from one point of view) or giving scamsters a lot of ammunition to try to weasel their way out of punishments for rulebreaking (when seen from the other point of view). (For people interested in this bit of history, the first relevant CFJ that comes to mind is CFJ 2451a: <https://faculty.washington.edu/kerim/nomic/cases/?2451a>.) So when the criminal portion of the CFJ protections ended up being moved to rule 101, an inquiry portion came with it. Oddly, the double jeopardy rule ended up being repealed (possibly accidentally?) during the creation of the Referee office, so the portion of the rules that was actually relevant has now gone, whereas the portion that was there to make rule 101 look more "official" is still here, and still cluttering up rule 217. > Second, because the R217 protection disallowing "rule changes" while > not protecting against "game state changes" is actually dangerous. > In the latest Economic proposal, it's very easy to imagine, say a year > into operation, some bug freezes spending and makes the CFJs unusable. > Either that means the rule change is retroactively nullified (really > ugly), or the protection doesn't work in practice, and is therefore > useless. There's no way to retroactively nullify a rule change at power 3, unless the rule's number is less than 105. (There aren't many rules with number less than 105.) Thus, rule 217 isn't powerful enough to do something like this, being "only" power 3. (Incidentally, this is the reason why rule 1551 has power 3.1, something which took us frustratingly long to notice.) I agree that the protection doesn't work, and should probably just be repealed. I also think it would considerably improve the aesthetics of rule 217 (if I had my way, I'd additionally move the second paragraph of rule 217 into a separate rule so that the original could be just that pair of sentences at the start, although I can't deny that it probably belongs there rather than elsewhere); rule 217 is fairly unique in containing probably the only paragraph of Agora's ruleset that works as a standalone nomic ruleset by itself (and the nomic in question, Nomic 217, survived for quite a while, IIRC). -- ais523
