I think we might need another cfj about it, honestly.

I’d argue that the reversion *was* correctly performed in that it did revert to a previous version that excluded some of the illegal changes. However, it also took a second step which was to revert an illegal change that had happened prior. This second step is permitted by the rules, up until it becomes disputed and requires a call for judgement.

If a change is first performed illegally, and then later performed legally, does that let you revert it? I think the answer logically should be “no”, and if the answer is “yes” it may be a bug in the core rules (but I am not fully confident that no such bug exists).

Assuming the CFJ itself was legal, the CFJ’s revert to a specified version was legal, even if that version was originally created illegally. As such, the change had been performed legally and couldn’t be reverted (even though it was performed illegally earlier).

So, my logic there is that there’s already specific language that’s supposed to be used, which is that the illegal move has to be “upheld”. However, this did not occur because at the time the cfj was made, the prior change was not known to be illegal.

I’d argue that it would be more of a core rules bug to allow implicit upholding of one illegal change simply by reverting a later illegal change by cfj.

Although honestly, there’s probably not much point worrying about it either way. Either the rule is broken one way, or it’s broken two ways. The rule is bricked until it’s fixed by proposal.

Hmm… I can also see an argument that the CFJ couldn’t be enacted properly because it specified that the ruleset should be reverted to a version that was never actually present, which may be impossible. (In such a situation, AFAICT the CFJ couldn’t be enacted – you’d just have to wait for it to time out.)

This is a good reason why CFJs should use language that doesn’t make assumptions, e.g. “change the ruleset to match [link]” rather than “revert the ruleset to match [link]”.

You can do two things with one wiki edit. I assert that the CfJ was correctly enacted, but the wiki edit included an additional action, which was the subsequent reversing of an additional illegal action.

@Josh: You were trying to revert an action that was peformed prior to the CFJ, but the CFJ set the ruleset to a particular version, overwriting the ruleset before that point. In other words, the CFJ legally added “charactery” to the ruleset (and was the first point at which “charactery” had been present in the ruleset).

You can’t then remove the legally added “charactery” just because it had been illegally added at an earlier time – it’s been added by CFJ and a CFJ is a legal way to add it.

The CfJ didn’t set text, it reverted edits. The CfJ invalidated changes, it did not make a positive change of its own.

CfJs can only affect the ruleset and gamestate by making positive changes: “When a CfJ is Enacted, the Admin Enacting it shall update the Gamestate and Ruleset, and correct any gamestate-tracking entities, as specified in the CfJ.”

Yes, and the CfJ specified “revert”.

Revert means to return to a prior state. It is not logically possible to iterate to a reversion. Revert means “go back”.

I really don’t want to have another “what do words mean” chat. I have explained what I did, if it is contentious then it can be CfJed. I don’t care all that much about the outcome.

I think I understand the point of contention now, though – the CFJ requests to do something impossible (reverting the ruleset to a state it was never in), and you’re considering the closest possible interpretation to be “reject all changes to the ruleset made since [time]”, whereas I’m reading it as “set the ruleset to match the tracking page as of [time]”. I think the latter makes much more sense because the time was specified by a specific link to a version of the tracking page.

The exact wording was “Revert the ruleset back to the version at this link:”; I don’t think that can be interpreted as “reject all changes to the ruleset since the ruleset had this version:” since it never had the version in question.

(I also seem to have an assumption that if a proposal/CFJ doesn’t explicitly say that it’s changing the history of something using words like “uphold”, it is changing only the current version. However, that might be based on rules/custom from another nomic and might not be true at BlogNomic.)