Guild Wars Wiki talk:One-revert rule/Archive 1

I think this is generally a good step from 1RV, but still has its own problems. A per-editor basis will greatly shift the bias away from "who was here first" or "who is more willing to revert war" and towards the majority opinion, which is much better. One particular problem though, is that relying on this to determine some final content can quickly lead us into the trap of the tyranny of the majority, usurping consensus. There's often not much incentive towards discussion when you represent a minority burdened with convincing the bulk of an opposing majority to change their minds and join your side. It also seems to break down quickly whenever a third "side" gets involved (and you don't even end up with the plurality opinion).
I also think this still suffers from lack of rigor, like 1RV. For revert restrictions to play such a huge part in determining our final content, they need to be pretty unambiguous to avoid big problems, right?

1. Does an initial partial deletion count as a partial revert?
2. If I disagree with some new content, and exercise my one revert in that section, does it count against me months later when someone adds some other content I disagree with in the same section?
3. Is this really to be enforced in every case equally with no exceptions, or only when sysops feel that someone is breaking their idea of the spirit of it?

The change to per-section basis makes sense ideologically (different section can be different disputes) but creates ambiguity for cases where section boundaries are crossed. I (of course) suggest adding a hard time limit and going to a per-page basis -- not because they fit the natural shape of disputes, but because they help all these other issues avoid being critical.. --Rezyk 21:12, 21 June 2007 (UTC)

I dislike the hard time limit as I've mentioned before, simply because it has its own downsides which I find to be worse than the alternative. If you set a time limit which is somewhat short in the grand scheme of things (such as 24hrs), then you get revert wars that simply expand over a larger timeframe, waiting for the limit to expire. If you set a time limit which is more on the long end (such as 1 month), it pretty much doesn't serve any real purpose to have the limit there at all.

I dislike per-page because it's quite possible for there to be multiple disputed sections, and a dispute over one section really shouldn't be allowed to stall out another.

In regards to "incentive towards discussion", I think that's a rather irrelevant point: if you're both the minority opinion, and don't believe enough in your viewpoint to put in the time to defend it, then chances are the majority opinion should hold as your viewpoint probably isn't enough to justify changing it. Bias towards the majority opinion is, to some extent, a good thing, as majority opinions are usually that way for a reason, and the proposal as is does not make it impossible for a minority opinion to prevail.

Rigor can help, but at the same time, sacrificing the integrity of the policy in the name of rigor is worse than leaving it slightly ambiguous, in my opinion: when in doubt, discussion should be favored over policy rulings, because no policy is perfect. Sure, "common sense" isn't really common, but in the worst case there are methods of arbitration that can deal with such problems.

In regards to your numbered points: as I see it, (1) would be "yes" - if you remove part of an edit, that's a revert, whether in whole or in part. (2) assuming that addition is not related to the previous revert, no. (3) I'd say it should be enforced in any case where an editor decides to make an issue of it. (Aiiane - talk - contribs) 02:36, 22 June 2007 (UTC)

maybe this isn't the place to debate this... i dunno. but i gotta say, a 1 revert rule seems extremely overbearing to me. it almost doesn't even give a chance for the debate to develop before it's stopped. it seems to me as if it would bubble up every insignificantly unimportant minor quibble one editor had with another to the admins. i can't believe that admins really want to deal w/ that kinda crap. especially on subjects they have little interest in. maybe it's too late to voice this, but i'd much prefer a 2 or 3 revert rule. as an example of a situation i had, i entered some trivia which another editor removed. i reverted his edit and he reverted it again. so i took a different tactic and sourced and explained my trivia entry in detail. he then made no more reverts because it was obvious even to him i was right or very close to right. but if he hadn't reverted me the 2nd time, i wouldn't have gone to the trouble of detailing the entry. and the trivia was just that, unimportant trivia that admins would likely find a waste of their time to deal w/. the situation resolved itself and ended up better than it began. the 1 revert policy would not have allowed it. --VVong|BA 07:07, 24 June 2007 (UTC)

Hmmm, why do you feel that allowing more reverts will foster debate and allowing less will not give a chance to debate? I feel it is totally the other way round: The more reverts allowed, the longer it takes till people start talking - 1RV forces them to explain their actions earlier.

In your example, with 1RV in place the second revert would not have taken place, so the matter would have gone to the talk page/explanation earlier, not later. And no need to involve admins either. --Xeeron 13:11, 24 June 2007 (UTC)

the trivia issue was so small i didn't feel like talking about it in a talk page. explaining it via the summary lines and actual entry was enough. if i was only allowed 1 revert, i wouldn't have put further time and research into it b/c i had already reverted once. if i had to spend more time on it on a talk page, i would have just said, screw it b/c it's not worth my time. true, this issue wouldn't have involved admins b/c i would have never taken it up to them, but it did get resolved in a better fashion b/c i was allowed more than 1 revert. --VVong|BA 17:13, 24 June 2007 (UTC)

Actually, per this proposed policy (1RR), you'd be able to revert his revert, because the original addition is not a reversion. So essentially, you'd add the trivia, another editor would remove it (his revert). You'd believe his removal was incorrect, so you'd revert his revert (your revert). Then, if he still thought it should be removed, he'd have to explain his reasoning on the talk page - and perhaps request that your source it, et cetera. (Aiiane - talk - contribs) 01:56, 25 June 2007 (UTC)

A few edits

A little late to the party, apparently, but I thought I'd give the draft some love since I noticed it in the RFC and it looks more palatable than 1RV or 3RR. My edits were intended as clarification and elaboration, but if I've somehow missed the spirit of the proposal, feel free to revert such portions and explain how wrong I was. I'd love to see it out of draft and into proposal. - Tanetris 22:22, 16 July 2007 (UTC)

I still support 1RR over 1RV and 3RR, hopefully some more people can move back into this discussion and we can attempt to finalize something. (Aiiane - talk - contribs) 21:51, 19 July 2007 (UTC)

In my mind, 1RR combines the best points of 1RV and 3RR, but we have to convince more people of that. Last time this debate came up, the support for 1RV and 3RR was very staunch and the discussion was far from consensus. --Xeeron 22:12, 19 July 2007 (UTC)

Well, everyone here so far seems to like it, at least in general (does anyone have any details you think need tweaking?), but still not exactly a lot of people here... It's been on GWW:RFC forever, I posted on GWW:POLICY's talk page, and I even pointed Rezyk over here on his talk page, since he'd expressed concerns before... Any other ways to draw community attention? As for the last debate... Feel free to correct me if I'm wrong, but the only debate I can really see is on 1RV's talk page, most of it taking place before the 1RR draft was even created, and the various concerns raised by both sides seem to be addressed here. Was there further debate somewhere else? - Tanetris 20:51, 23 July 2007 (UTC)

Some concerns

I like that this draft has been given more rigor/clarity, but here are some concerns:

• It can give too much leverage/support to a very unpopular side. Suppose that 5 users prefer A, 5 users prefer B, 1 user prefers C, and they each do their one revert along those lines. This policy would end up supporting C if it's the last revert done. Or, here's another example with just 2 sides: Suppose 4 users prefer A and 5 users prefer B, and they each do their one revert along those lines. This policy would end up supporting B...even if after some discussion, A is preferred over B, 7 to 2.
• It can encourage revert warring too much. Suppose that all the users who would, have each "spent" their one revert. Then one of those users who dislikes the final state has a strong incentive to revert again to his/her preferred state (because then policy is against the other users changing it again).
• The usual "consensus, not majority" spiel.

--Rezyk 21:48, 23 July 2007 (UTC)

In regards to your first point: I'd say that such is a problem with almost any hard limit on reverts - under 3RR, if everyone who would "spend" their 3 reverts has done so, and someone else makes a change, then we end up with the same situation. Likewise, under 1RV, if the 1 revert done is very unpopular, it's still "stuck" until consensus. I can't see a way to get around this, because the whole point of moving to discussion is to determine precise popularity - and in the average case, 1RR actually gives more support to a highly popular side - because there's a much higher chance that all of the people supporting the popular side will not have "spent" their reverts.

As for the second point, are you speaking of a user who has already reverted once, then reverting it again? That's a simple violation of policy, and the very fact that they're breaking policy means that it's actually against them, and not other users - that would be the point for a sysop to step in say where the reverting needs to stop (changing it back to the "final state" while discussion takes place). If you're not speaking of a user who has already reverted once, then your premise ("all the users who would, have each "spent"....") isn't satisfied.

As for the third, that's the purpose of discussion. If it were possible to embody consensus in an objective policy, we'd have no need for discussion, but it's not. The point of any revert-limiting policy is to force discussion at some point so that consensus can be reached, attempting to factor in consensus before it has been reached is both an unobtainable goal, and a nonsensical one. That said, majority is 'on average the closest thing to consensus prior to discussion, so if forced to pick a starting point, it's generally the best to go with. (Aiiane - talk - contribs) 22:03, 23 July 2007 (UTC)

With regard to a page in a revert war "supporting" the version at any given time: what if we had a {{disputed}} tag? Any user who has already used their revert may add the tag to the article or section in question, basically saying that it's in dispute, to see the talk page for details, possibly reminding folks to follow 1RR, and once the tag is added, it doesn't stop further reverts from happening (as long as the reverting policy is followed), but no one can revert the {{disputed}} tag until a consensus is reached? This is a very rough idea off the top of my head, as I only have a few minutes, and I may be missing an obvious flaw, but could this help? On the other points, I think Aiiane's already covered it. - Tanetris 22:24, 23 July 2007 (UTC)

I think a {{disputed}} is a great way to try and soften any really heavy revert war. Not sure if it should be written into policy though. --Rezyk 05:00, 25 July 2007 (UTC)

1. I agree that it's a very common problem when designing revert limit policies, but it's still a problem.. I won't insist that it has to be completely fixed, but I believe there are ways to make the potential "excessive leverage" much smaller than this.
2. Yeah, I mean someone reverting before, and then reverting again against policy. Would we always negate all edits to that section since that point, no exceptions? If not, there can still be decent incentive. If so, how that works is not really clear in the draft.
3. I understand that, really -- I mean, I even talked at length justifying this kind of bias towards the majority in 1RV's talk page. And as a starting point, it's fine by me. But I don't see how this avoids the majority also becoming the default ending point whenever consensus is not reached. And then what real difference is there between that and majority-all-the-time?
4. New: This policy doesn't stop stupid quarrels from happening on the wiki. (<-- not really a complaint, just trying to make a point about 3RR)

--Rezyk 23:32, 23 July 2007 (UTC)

In regards to 1: the only real way I can see to not favor whoever's revert is last is incredibly awkward, as well as somewhat defeating the point of the policy - that would be making it so that each editor can revert each other editor once. That would guarantee that a single editor's unpopular view wasn't left as the "de facto final" version, but it'd also allow even more group-warring for reverts, and be even more difficult to track. As it is, I think it is mostly fixed - the chance that an unpopular change would be the last in a set of reverts is extremely low due to the fact that an unpopular view will not have many editors reverting to it, and a popular view will.

For 2 - I'm not sure if that's something that can be objectified, beyond the general layout of how to deal with violations of this policy. I'd say that's probably something best left to sysop discretion, as the circumstances can vary widely, and the main purpose of this policy is one of stabilization, not regulation of content.

Addressing 3 - it doesn't. The thing is, you can't avoid something becoming the "default ending point", unless you want to blank the article until consensus is reached (which I think is even more counterproductive), and again, on average the majority opinion is the best default. It may not be the best option in 100% of cases (far from it), but until you can determine what is the best option, you have to go with something which can be objectively determined.

Not sure what you were going for with 4, since I don't see how that differentiates this from 3RR. (Aiiane - talk - contribs) 01:43, 24 July 2007 (UTC)

For 1 and 3, it seems like you're saying "we shouldn't be concerned with these problems because we expect them in almost any non-broken policy". But I don't get why that should alleviate any concern -- even if they were the same in every revert limit proposal, it makes a difference in deciding whether or not to accept this proposal instead of say, continuing to rely on page protection to limit heavy revert wars. For 2: Regardless of purpose, it's going to be a practical problem if some expect sysops to have a special revert power/decision that isn't covered (and is sometimes forbidden) in written policy. For 4: Sorry, I should have said that I was trying to lead up to a point, not making one yet. Someone is supposed to ignore the parenthetical note and counter it, and then maybe I can make the point. =) --Rezyk 05:00, 25 July 2007 (UTC)

I'm not saying we shouldn't be concerned with them, I'm saying that we shouldn't go overboard trying to manipulate the policy to achieve a goal that's unattainable. It's like trying to make a car that can't get into an accident: there are some factors you simply can't prevent (such as, say, other drivers). That doesn't mean you shouldn't acknowledge that those factors exist, or do what can be done to buffer against them, but at the same time you shouldn't spend 10's of years of research and development to try to figure out a way to defy reality. I've already explained above why in most cases, 1RR will result in the "best" outcome, you've countered by questioning the occasional time when a less favorable outcome might result. My counter to that is that there will always be a chance for a less favorable outcome to result, but that as it stands, 1RR (in my view) makes the best attempt it can to make that alternative outcome less likely than all others, and does so at least as well, if not better, than 1RV and 3RR.

As for your response to 2, I was more speaking towards sysop discretion in determining what edits were part of the revert war - subsequent edits that weren't part of disputed content (but were still in the same section) should generally (imo) be kept - for instance, if two people are having a revert war over a section that happened to include a link, and someone made a maintenance fix on that link in the middle of their revert war, but not involved in it, the link fix should not be reverted. Any edits which do violate the revert policy, however, should. (Aiiane - talk - contribs) 11:49, 25 July 2007 (UTC)

Aside from generally agreeing with what Aiiane's already said, I think the strongest method of dealing with majority vs consensus and such is simply fostering a community that values consensus, and I certainly see that written into this policy, as well as many of the other policies on the wiki, and in the spirit of most of the discussions I've seen around the wiki. There will always be people who don't quite get the concept, but I think the community at large has it down pretty well, and if it is someday lost, I can't imagine finding the fault to be the policies.</soapbox>

Now then, I did write up a little piece on the {{disputed}} tag and its uses, more as an advisory that it exists (or will exist, anyway) than as part of the policy per se, so I tried to keep the language soft. I do think it should be included, just so people can find it, but if you do find it objectionable, I'm always open to further discussion. I imagine the tag to say something like "Parts of this article or section are currently under dispute. See the talk page for details. Please do not remove this notice until a consensus has been reached." but someone else is going to have to create it, as I'm quite useless at such wiki code.

On a final note: do you have a policy that would stop stupid quarrels, Rezyk? If so, quit holding out on us. ;) Else, I'll settle for the point you were trying to get to. - Tanetris 17:27, 25 July 2007 (UTC)

I'm starting a separate section below for discussion of a {{disputed}} tag. (Aiiane - talk - contribs) 20:59, 25 July 2007 (UTC)

What would you guys think about a hypothetical proposed policy: "Discussion is encouraged and you should abide by consensus. Also, any dispute may be decided by a majority vote."? --Rezyk 18:08, 26 July 2007 (UTC)

I'm guessing that you're trying to lead into a point that any revert war which doesn't manage to reach a consensus is essentially "decided" by who got off the last revert before the limiting policy kicked in (correct me if I'm wrong?). However, I'd like to ask you a question: can you think of an alternative to that, aside from the option of it being "decided" by who is willing to revert repeatedly over a longer period of time (the effect of having a "reset period" such as 3RR's 24-hour span)? (Aiiane - talk - contribs) 21:35, 26 July 2007 (UTC)

Well, I think there are a number of people who would die a little (or more) inside at the proposal... I also think for matters that are fact rather than opinion, truth cannot be determined by poll, and for matters of opinion, reasonable opinions should be presented equally, and unpopular does not mean unreasonable. I think that it's open to abuse, and I think it's unenforcable as policy: even if I and the rest of the wiki agreed with it, I think the most it could be would be a guideline. On the other hand, I think if you ditched the second sentence, you'd have a great start to a "Guild_Wars_Wiki:Consensus" article. Isn't it funny how often the word is used, yet we don't have an on-wiki explanation of what it actually means? Sorry, drifting... Anyway, I take it this is a continuation of the stupid quarrels point, so: I don't think it would do anything to stop stupid quarrels. I suspect that, if anything, it would shift the focus of the quarrels from the actual pertinent information over to the validity of votes, and down that path leads to sockpuppetry witchhunts and violations of NPA. So I would pretty much oppose that one. I'd vote against it even. ;) - Tanetris 22:35, 26 July 2007 (UTC)

(reset indent)

I'm dropping the "stupid quarrels" thing. (I really shouldn't have mentioned that I was leading to a point, sorry =/ )

Tanetris, I agree with that. So changing it a bit: "Discussion is encouraged and you should abide by consensus. Also, any dispute may be decided by a majority vote. Voting is done in the form of a revert (and only your first vote counts)." Does that fully solve all your concerns?

I also can't fully understand your soapbox'd stance toward not faulting policies. I agree that fostering a culture/spirit that values consensus is the best method..but the mechanics of policy are important too (and are a big influence on culture in the long run). It's like the "sysophood isn't power/sway" ideal. Just because we accept or want it doesn't make it true; it's the policies and practices that really determine it, and we need to maintain or change them to match. If we just rely on cultural mantras like "admins are just regular users" without enough of a watchful eye on our mechanics, one day we'll wake up to a wiki where that mantra is nothing but a forgotten memory and every administrator has near-unlimited authority (can ban users at will, etc). For this policy, I'm just basically saying: I know that revert wars are the focus here, but watch out, for it steps around our consensus ideal by deciding things without it.

Aiiane: Yeah, you can take that as my point. (I actually thought I made it earlier with "default ending point") I also tend to think that any answer to your question is besides the point. I'm not bringing this stuff up to favor another policy (although it does lead into some reasoning behind 3RR's reset period) or to say that it's unacceptable, but because I simply consider it an inherent, major concern with this policy that we should understand (or attempt to understand) before accepting. Alternatives may be better in this aspect and worse in others..but it shouldn't counter this concern if we can't find an alternative that is strictly better.

To answer your question anyways, some alternatives besides simply 3RR-type consensus:

1. rely on page protection to limit heavy revert wars (current state?)
2. push the community to drive harder to find consensus to decide disputes
3. adopt a "removal of disputed strategy/advice during its discussion" stance to deal with the 90% of cases that 1RV is supposed to handle well, with the same effect
4. rely on some admin discretion in dealing with excessive reverters even without fence-crossing violations (also current state?)
5. some combination of the above, with or without also having a strict fence like 3RR

--Rezyk 20:10, 27 July 2007 (UTC)

I agree that we should understand all ramifications of a policy before we decide whether or not to implement it. It's been stated already, but I have no problems with making it clear again: if consensus isn't reached, then the revert policy will "decide" the issue anyways. I can't stress that enough, as making such clear is one step closer towards working out any consensus on a policy to adopt (and, while I hardly wish the process to be shortsightedly rushed, I do think we've spent an large amount of time discussing this issue and anything that moves us more quickly towards consensus is of merit).

That said, given the ideals that seem to constitute what we're working towards on this wiki (sysophood being "no big deal", et cetera), I believe that this policy reasonably addresses its goal in the spirit of the wiki, while doing so with faults that while present, are both manageable and reasonable.

I dislike the "removal of disputed content" approach, as it seems entirely contrary to the purpose of the wiki, in that it discourages contribution rather than encouraging it - the same goes for page protection. I would not be averse, however, to including language that encourages sysops to publicize discussions where it seems consensus is needed and lacking as part of their maintenance of the wiki. (Aiiane - talk - contribs) 20:33, 27 July 2007 (UTC)

First off, let me clarify my soapbox point on policy, as I was a bit inexact: I am not denying that it is possible for policies that would degrade the spirit of consensus to exist or that policy has a significant affect on culture. What I am saying is that none of the policies I've seen that are currently in effect or currently proposed could be blamed for a shift away from consensus, and I think that such a shift would have to happen first before any policy that would degrade it further could be passed. I apologize if I seemed overly dismissive, as it is something to watch for in general, but I don't feel the concern is justified here.

No, it doesn't alleviate my concerns, but I think I see better where you're coming from. I'm coming at the policy from the angle that it forces users to stop relying on the revert button and go talk to each other so that reaching a consensus is possible, whereas you're coming at it that without letting people continue to revert indefinitely someone wins and someone loses automatically if a consensus can't be reached, weakening consensus. In a way, we want exactly opposite things to accomplish the same thing.

The question then is what should happen if consensus can never be reached? I'm against revert wars stretching into eternity (a large part of my opposition to 3RR), and I'm against outright removal as a rule for all cases. I think the disputed tag is helpful to keep someone from being declared the default winner, but I can't really imagine leaving it in an article forever... What if we said that if all sides agree that a consensus cannot be reached, the article should be editted to neutrally reflect each viewpoint? For example, if it's informational (lore, mechanics, etc, due to ambiguous or contradictory text/findings), make a simple statement that it is currently uncertain whether or not x, or whether x or y, etc. If it's opinion (disagreement over effective tactics, etc), present both options and state that some players prefer one while others the other. Of course there are other disagreements like what belongs in an article... I think in that case, where a neutral version of the article simply isn't possible, it's going to have to default to the side with more reverters. Perhaps it's not perfect, but all in all I'd be pretty hard-pressed to call it unfair, and discussions for consensus can always continue. That's the best I've got, anyway. - Tanetris 22:21, 27 July 2007 (UTC)

I disagree with you both on some crucial points -- I don't see the faults as really manageable, and would see this as a big step in an eventual breakdown of consensus. Frankly, I see the hypothetical "majority vote decides" policy as almost strictly more favorable and think both would end up in the sort of problems that Tanetris mentioned about that.

On the other hand, I am rather convinced now that you both understand my main points. Thank you for that. Our disagreements seem to be strongly subjective, so I don't know how productive further debate would be (although I'm willing if desired). If the community at large still wants this despite my concerns, I generally won't stand in the way of a consensus.

Tanetris, that "try to find a third version that accommodates both sides" path is very much where I would like to see us head, and I think it should even generally be the first thing that we look for -- even before counting sides. Having everyone strongly driven towards such (even when painful/slow) is, I think, a key point of being a consensus-driven community, and something that I don't like deterministic solutions detracting from. (The side who is set to "win" has little incentive to compromise.)

What if we can't find a consensus and can't find an agreeable solution? Try harder! And when that doesn't work, try even harder! Okay...those aren't real answers but I think our culture could use a good kick in that direction. =) Real answer: tough question, and I have no easy solution. One yucky but workable way is to allow arbcomm to interfere and declare a consensus when they agree that it is warranted. I'm also not particularly opposed to "consensus required" solutions and am not so strongly opposed to "majority decides" solutions if they're limited to strategy/advice decisions (or similar naturally widely subjective stuff). --Rezyk 04:53, 31 July 2007 (UTC)

Let me be absolutely clear here: counting sides is what I consider the last, usually unnecessary step. In my (perhaps naive) view, in 99% of cases that this policy would apply to (i.e. not vandalism and such) can be cleared up with both sides happy just by one or both taking the time to explain to the other why they put in/removed what they did. Relatively few people on the wiki know each other, and even fewer are mind-readers, so misunderstandings are inevitably going to crop up, and I think that's the primary thing we're going to have to deal with, hence my insistance that a policy that sends the user to the talk page rather than letting them rely on the other person getting tired of reverting first.

Now let's say it doesn't work because, as you've mentioned the possibility a few times, the side that has their version on the page feels disinclined to discuss it. Personally, if I see one person trying to get another to explain a revert while the other ignores them, unless the person being ignored is blatantly wrong (in which case I'd chime in with an explanation myself), I'm going to side with the ignored person and toss my revert in, even if it's just to get talks going. I'd imagine there are plenty of others around the wiki who'd do the same, many of whom keep an eye on Recent Changes and the RFC page. Again, perhaps naive, but I trust the community to keep anyone from trying to use the policy to dig in in trenches.

I do agree that our disagreements are strongly subjective, certainly you're not "wrong" in your concerns, and I both appreciate and respect that you've kept at this, helping hammer out something that, while not strictly perfect, at least recognizes potential problems. So thanks for that. I'm going to go add a bit to the resolving disputes section (actually just did), and I welcome comments/criticism on that, or anything else really. After that, I think I for one am ready to call this a proper proposal and see who else I have to wear down how it goes. It's certainly been an interesting process. - Tanetris 07:15, 31 July 2007 (UTC)

Some concerns, continued..

Xeeron asked me to elaborate, so here goes...

How this as policy might usurp consensus decision making: I presume we agree that if consensus isn't reached, then the revert policy will "decide" the issue anyways (as previously noted). In the simple case, this "1RR result" should be the version supported (at the time of revert) by the majority of those who opt to exercise their one revert.

• Suppose that the 1RR result favors version A, but there is no consensus yet. Users who just favor version A will be naturally less inclined to work as hard towards a consensus, as they don't have much to gain from it.
• Each user is less inclined to discuss arguments according to the amount of effort they think it would take to sway the 1RR result. For example, if the issue was roughly split 40-60 and one believed that there was no hope of changing enough minds to change the 1RR result or reaching a overriding consensus, there is little incentive to even put in the time and effort to bother explaining.
• As we get more accustomed to this process and these practices, more issues are left discussed less and generally decided just by the reverts.

Some of my internal views of how this "would be worse for the wiki than revert wars" (please don't take these as trying to conclusively prove anything; they are subjective):

1. Actually, I reject that this is some choice about having revert wars or not. Making something a violation of policy doesn't automatically stop it -- it just makes it a violation of policy. (NPA doesn't eliminate all personal attacks from happening.) This is an important distinction here because I worry about how restrictions may encourage revert warring in some cases.
2. I worry that the real goal is being oversimplified here. This is all really because we're familiar with (and want to reduce) the often excessive/disruptive strife that follows revert warring, right? If we eliminate revert-warring strife from a case, but cause more strife in another form, we haven't necessarily improved the case, right? I see this proposal as decreasing revert-warring strife, but probably increasing I-demand-you-stop-reverting-my-allowed-revert strife, I'm-being-oppressed strife, that-guy-is-a-sockpuppet-no-he-is-not-yes-he-is-no-he-is-not strife, etc. I don't even take it as a given that revert warring must be stopped, if the only solutions may lead to worse strife.
3. Less discussion tends to leave our content worse off. At least with a revert war, increased visibility tends to lead other users to weigh in or help find a good middle ground.
4. Our whole system is built around and theoretically based on consensus. Damage to the consensus decision-making model in one aspect can undermine various other aspects. It's hard to be specific since it depends so much on how things play out, but I wonder if this will eventually force us to .. (wrote a few things here but decided to omit because it seems pretty speculative and alarmist). Anyways, this is not automatic rejection for me but I take the consensus ideal as something that should be carefully guarded, as it is not trivial to maintain. As far as I know, there isn't anything that is dependant on "no revert wars". Though potentially explosive, revert warring incidents still tend to be relatively isolated, and not without controls. Stopping them cold now is not vital enough compared to the risk towards the consensus model (but this is based rather subjectively on my own possibly-overreactive assessment of that risk).

Maybe it would be easier to understand me (or for me to understand the opposite view) if someone could try to explain: How is 1RR so much better than this hypothetical proposal? -- "Discussion is encouraged and you should abide by consensus. Also, any dispute may be decided by a majority vote. Voting is done in the form of a revert (and only your first vote counts)." What should I depend on to stop 1RR practices from devolving into virtual voting? --Rezyk 10:15, 22 October 2007 (UTC)

Consider this, Rezyk: the difference between your proposal, and 1RR, is that, if we assume that at least some portion of the wiki populace believes in working for consensus (which, if we can't assume that, means that a consensus-driven wiki is doomed... so let's assume that), then unlike a majority vote, there is a good chance that at least some of the people who would have formed the majority will elect not to "utilize their allowed revert" and instead take the matter to discussion. Whereas with a vote, the majority will win no matter what, best case, or worst case, with 1RR, only the worst case is equivalent to a vote, and the very assumptions we base the entire operation of this wiki on lead towards an outcome that isn't based on a worst-case scenario.

In short, voting and anarchy are on two ends of a spectrum, with anarchy being revert wars. Though the outcome of 1RR can potentially be anywhere along that spectrum, if one takes either worst case to an extreme, the aim of 1RR (and something I believe it will effectively accomplish is to make it much, much more likely to have an outcome somewhere near the middle of the spectrum, and centered around the consensus we wish to foster.

As I've noted on other pages, I believe that any option we might choose to implement will have its downside, but not implementing any policy has its downsides as well, and I personally believe this to be the best option among them. (Aiiane - talk - contribs) 12:25, 22 October 2007 (UTC)

For my hypothetical, users could also discuss while voting, or while electing to not vote. So only the worst cases would not be decided by consensus, too.

Is it about a soft difference in labels leading users to tend to act differently between the two proposals? That calling it a vote would encourage users to follow the votey/reverty mechanism more instead of discussion? --Rezyk 07:51, 23 October 2007 (UTC)

There's still a difference, though, Reyzk - by choosing not to exercise their revert, they prevent those supporting an opposing view from exercising any further reverts, as well - specifically, in your example, if someone opposing a change decides not to vote, that doesn't stop 5 more people from voting in support, thus making the situation seem even more one-sided. Essentially, a vote can have a margin that is larger than 1, whereas a sequence of reverts cannot. 1RR encourages users to work towards a version that everyone can agree to not revert on, because they can never really be sure whether there will be enough users to outright "counter" a revert, whereas a vote encourages users to work towards a page that the clear majority can agree on. With a vote, if you see "8 in favor, 1 against", you don't bother to continue to adapt it because there's no reason to, the majority will clearly support it and it's unlikely that will change. With 1RR, all you see is that there's a disagreement between some people, and thus you go to the discussion and work from there. (Aiiane - talk - contribs) 08:39, 25 October 2007 (UTC)

Hmm so many good points and so much I want to respond to, I don't know where to start, so let me say this first: I didn't get it at first, but now I see exactly where you are comming from Rezyk. And in part, I share your concern. My biggest problem with this proposal here is the possibility that by making it policy, people who previously never thought about reverting might be made aware of that technic.

However, your main fear seems to be one of a "majority vote by stealth", not of people getting aware of the possibility of reverting. There is one flaw in your argumentation, which I will be sure to "exploit" if this policy proposal ever passes: It really pays off for members of the minority NOT to use their reverts early. If you know from the beginning that you will "lose" a revert war and after that the other side's willingness to discuss will be reduced, don't revert. The better strategy is to go to the talk page and trying to convince other people while you still have the possibility to revert. All your "getting stuck because all reverts are used up" scenarios depend on editors being stupid and indeed using all reverts up front. Why should they? Why not first try to convince the others on the talk page before burning the topic by starting a revert war you will lose? It is a strictly better strategy to start the discussion first. And that is exactly what we want to achieve: To get people to discuss. --Xeeron 18:15, 22 October 2007 (UTC)

True enough. I'd like to make it clear that I'm not worried about every sort of discussion being stifled, and I know there will still be good incentive for a currently-losing-but-has-reasonable-hope side to try putting forth convincing arguments (I tried to word my explanation to exclude that case). I'm more worried about the direct effects on the "end part" of discussions (how they resolve), and only indirectly from that onto the start and middle parts. --Rezyk 07:51, 23 October 2007 (UTC)

Reverting is one of the first things we tell new users about. (GWW:WELCOME links to Help:Reverting). Backsword 05:03, 29 October 2007 (UTC)

Cont.

I find all these discussions to be very theoretical and I personally don't understand half of it. Rather than each of you trying to convince the others why your preference is better, shouldn't you guys take a step back and consider what's best for the wiki instead? What I mean is, to take into consideration the sorts of revert wars that have already happened in this wiki and then see which method would be sufficient enough to take care of it. -- ab.er.rant 02:24, 24 July 2007 (UTC)

I think the problem is that all 3 would be sufficient to take care of it. ;) Thus again it comes down to preference, hence the more "theoretical" (as you put it) discussion comes up, since we all want to try to aim for something that's lasting and won't need to be constantly rewritten. (Aiiane - talk - contribs) 03:26, 24 July 2007 (UTC)

I still can't really wrap my head around all the revert here revert there and but can't rever this and that stuff :P Must read more slowly and carefully. Anyway, how about going with the simplest and easiest solution instead? The whole point is to stop revert wars, and isn't something that's easily explained be more understandable by general users? -- ab.er.rant 04:47, 24 July 2007 (UTC)

What about starting with a loose policy against revert wars which can be amenderd later? -- Gordon Ecker 05:06, 24 July 2007 (UTC)

I've put together a proposal here which I believe covers all the common elements of the three proposals without any of the exclusive elements. -- Gordon Ecker 05:46, 24 July 2007 (UTC)

Unfortunately, I don't see that as being rigorous enough to enforce as a policy. It could work as a guideline, but policy needs to be more explicit, generally. At what point does something become a revert war under your proposal? That's exactly what the 3 proposals already made attempt to lay out, because it's necessary. (Aiiane - talk - contribs) 11:49, 24 July 2007 (UTC)

Whenever a sysop decides it's revert-warring. They would have the discretion to block or warn under the other three proposals. Ideally, no one would get blocked for revert warring without getting a warning first, but if a "don't block without warning" clause was added, the proposal would no longer be a compilation of the common elements of the three other policies. -- Gordon Ecker 22:05, 25 July 2007 (UTC)

My personal opinion (let me be clear on that) is that this proposal is essentially the "de facto" policy on the wiki as of current, and I'd rather work on something that work be more structured for a long term policy (which I don't believe NRW would do as well for), and just stick with "de facto" policy until that time, because otherwise (however unintentionally or good intentioned), we effectively give the NRW proposal implicit support as a long term policy simply through its adoption. (Aiiane - talk - contribs) 00:07, 26 July 2007 (UTC)

One problem is that if the currently proposed blocking policy gets passed, the current de facto policy will no longer be enforceable, as you can't have a "clear violation of revert policy" without a revert policy. -- Gordon Ecker 01:59, 26 July 2007 (UTC)

Then I'd say passing that (in its currently written form) is contingent on passing one of these proposals, but that doesn't necessarily say we should adopt something potentially sub-par simply because of that. Certainly we can take it into consideration, but I'd much rather try to come to a consensus in regards to the existing discussion. (Aiiane - talk - contribs) 02:01, 26 July 2007 (UTC)

The quick and simple explanation: You can revert once in any given disagreement. So can everyone else.

Everything beyond that is making sure we're absolutely explicit what that means, what happens next, etc. - Tanetris 05:13, 24 July 2007 (UTC)

My opinion on this is that allowing another person supportive of the same edit to keep reverting is inviting more fuel to be added to the fire. It's one thing if that person is unaware, and it's another if they're involved in the discussion and just partake in tag teaming the reverts. I think it should be one revert per "side" not per person. Reverting is not a "right" to be practiced by people, it is a means to maintain an article. Plus, such reversions are bad faith edits. --Karlos 10:36, 27 July 2007 (UTC)

The problem with allowing only 2 sides in a discussion is that sometimes there aren't just two sides. Due to someone voicing his opinion in a more distinct manner than others, might give the impression that that's "one side" to the argument where more people are supportive of, but some people might still have a different opinion.

Regarding point 4 of the proposal, I don't think the last self-revert should be used or should be much more clear by adding something along the line that the summary of the revert should clearly state the self-awareness of the 1RR violation by the user doing that last revert.

For the rest, I think the {{disputed}}-tag is a great addition to trying to avoid a revert war and I think that this tag should be included as soon as a revert skirmish is likely.

Lastly, I think this policy is a good addition to the current policies we have. --  (CoRrRan / talk) 11:16, 27 July 2007 (UTC)

Hah, way to miss my point. Ok, If there are 14 sides to the debate then allow one revert per side. The number of sides is actually irrelevant. Let's say the article text says: "Prince Rurik is known to be the founder and leader of the Ascalon Vanguard" So, team A believes he is the founder. Enter team B. They believe he is just the leader, so they change it to "Prince Rurik is known to be the leader of the Ascalon Vangaurd" So, semantically, any change of the article to say Rurik IS the founder will be RV number ONE for that side, and any change that removes references to him as founder will be RV ONE for the other side. Now, if team C shows up that says "Prince Rurik is the leader and possibly the founder of the Ascalon Vanguard." Then they can do that and neither of the first teams can changs it back because they have used their RVs relevant to that piece of information. --Karlos 20:28, 27 July 2007 (UTC)

The problem I see with this is that it's incredibly hard to enforce, unless you can objectively define what a "side" is - otherwise, there's the potential to be counterproductive and simply make the situation more difficult, with everyone arguing over whether their or others' contributions constitute a "side" or not.

Additionally, this suffers from effectively the same problem as a 2RV rule would: it's highly biased towards the second revert (generally keeping the additional content as was originally added). (Aiiane - talk - contribs) 20:35, 27 July 2007 (UTC)

"disputed" tag

I've prototyped something here, any feedback, suggestions, or other mudslinging? (Aiiane - talk - contribs) 21:00, 25 July 2007 (UTC)

I much prefer the tag be placed on a section level than an article level. i.e. if two people are arguing over whether Shiro is an assassin or Warrior I'd rather have this tag over the notes section than on top over the whole article.

I have always felt that the tag on the whole article tells users that the whole article is not to be trusted. Which I think is unfair and incorrect. --Karlos 02:18, 26 July 2007 (UTC)

The reason for the dual wording is that some articles consist of only a single section (oftentimes without a section header) and thus using only the section phrasing would be awkward in those cases. However, I'll go ahead and add a recommendation to the template's Usage section that it should be placed at the section level whenever possible. (Aiiane - talk - contribs) 02:24, 26 July 2007 (UTC)

I would oppose it placed at the top of any article. If the article is very small, perhaps we should make a one line version of this box to be placed in small articles atop the contested part. Like I said, I am opposed to the idea of a big tag at the start of the article telling the user to be suspicious of it. --Karlos 08:40, 26 July 2007 (UTC)

So you'd be opposed to placing it at the top of say, Shutdown? Or for another example, Quit? What about Easily interruptible? There's simply no better place to put such a tag on those articles, unless you're literally wanting to plop a tag in mid-sentence. (Aiiane - talk - contribs) 11:40, 26 July 2007 (UTC)

Waah. It's huge. I'd prefer something much small, perhaps a single line, placed over the dispuded entry. Revert wars may be of more interest to us, but for the normal user, the actual article is key. There is no need to make them doubt the numbers in a skill article because someone can't agree obut the inclusion of some trivia. Nor would I like to see editing stop on other parts of the article. Backsword 10:17, 26 July 2007 (UTC)

Which is why it states to see the discussion page. However, I've temped up a smaller (but less informative) version, here. (Aiiane - talk - contribs) 11:50, 26 July 2007 (UTC)

I think the prototype was meant to keep it consistent with our other notices, but I don't see what difference it makes between having a large notice and a small notice that basically says the same thing. Users are more suspicious the larger the notice is regardless of the wording? -- ab.er.rant 11:08, 26 July 2007 (UTC)

I'd say so. People tend to focus on big shiny things. Relative size will be seen as an indicator of relative importance, intended or not. So I was thinking that a big note is good if the entire article is disputed, but something more like the otheruses template for the normal case. Bit of an formating issue too, as the disputed item will oft be a single line. A big box would look odd placed just above. Backsword 22:38, 27 July 2007 (UTC)

Aberrant, please read above carefully. I said: I want the notice to be only placed on SECTIONS and NEVER on ARTICLES. If the article is just composed of three lines of text, make a small version of the notice and place it atop line # 2. --Karlos 11:36, 26 July 2007 (UTC)

I think I got that, and I wasn't really responding to your comment about where it was located ;) I was just expressing my thought about Backsword's comment regarding the size, which I feel that a big notice wouldn't really be any different from a small notice, except that the small notice is more easily missed. -- ab.er.rant 02:57, 28 July 2007 (UTC)

Draft -> Proposal

Since this has been sitting on RFC for quite a while now, and most of the discussion seems to be trending towards "what we should be aware of if we choose to adopt this" rather than "this part of the policy is inherently broken", I'm going to go ahead and change its tag from a policy draft to a policy proposal. Perhaps we can move a step closer to adopting one of the various proposals put forwards in regards to reverts. (Aiiane - talk - contribs) 18:48, 28 August 2007 (UTC)

This has been formally proposed for over a month and a half now, it'd be nice if people could actually comment on whether they find it suitable to adopt or not. I realize there have been other hotspots of wiki activity, but it'd be nice to at least make a start here. (Aiiane - talk - contribs) 01:17, 16 October 2007 (UTC)

I continue to wholeheartedly support it. - Tanetris 01:31, 16 October 2007 (UTC)

I commented out the reference to blocking policy (we should not link policy proposals that are not policy yet from policies). Apart from that, I support this. --Xeeron 09:01, 16 October 2007 (UTC)

I too think that it is mature enough to be used. --  (CoRrRan / talk) 11:08, 16 October 2007 (UTC)

On the same note, can we replace the middle three exceptions with a statement about reverting content disallowed by other policies? So we don't have to rewrite this when other policies change, thus also avoiding the risk that we forget and end up with contradictory policies. Backsword 10:41, 18 October 2007 (UTC)

I think it's better to leave it as it is - some things are a lot more subjective, such as GWW:NPA, and thus it might be better to not explicitly include them in the excepted cases. (Aiiane - talk - contribs) 10:46, 18 October 2007 (UTC)

I'd say vandalism is a worse offender there. Hard to abuse NPA outside talkspace. Backsword 11:00, 19 October 2007 (UTC)

There's a reason the policy very explicitly only refers to the "most blatant" vandalism. (Aiiane - talk - contribs) 11:15, 19 October 2007 (UTC)

I'm strongly against it, my main reasons being generally summarized above in this edit (even though that's for a slightly different proposal) and that the potential damage to consensus outweighs the benefits. I also encourage seeing if we have a consensus for this despite my opposition. --Rezyk 04:59, 17 October 2007 (UTC)

Er, I think you linked the wrong diff, Rezyk. (Aiiane - talk - contribs) 06:17, 17 October 2007 (UTC)

I guess I also should explicitly note that I am generally willing to "stand aside" for a potential consensus here. (would have said something like "I object" otherwise) Just wanted my frank opinion on the record. --Rezyk 10:53, 22 October 2007 (UTC)

One question for Rezyk: Since you are the one person vocally arguing against this, I read up on your posts regarding this, because I tried to find out what exactly does bother you. Am I right in my guess that it all comes down to the one big problem that you fear that creating this policy might lead to a change in discussion culture/something else that in turn leads to disagreements being solved by the majority imposing its view instead of searching middle ground/engaging the minority?

If I am right about this, can you please elaborate how you think this mechanism would work, why it would be dependant on this policy being made official in contrast to being rejected and, most importantly, how you arrived at your view that this mechanism would be worse for the wiki than revert wars. --Xeeron 10:13, 18 October 2007 (UTC)

Sorry for the late response. Yes, it is mainly that issue of usurping consensus (generally through dispute mechanics here rather than culture), though there is also a separate element of how much this will push us into uglier argument territory (such as bickering over precisely which edits are counted), which is what I meant by linking to Tanetris' post. I'll try and answer your follow-up 3 questions soon. (EDIT: Replied at #Some concerns, continued..) --Rezyk 05:52, 20 October 2007 (UTC)

Consensus

There was a very good point in Tanetris post linked by Rezyk, which I wanted to split off from the proposal discussion: Everyone around here agrees that consensus is the holy grail of the wiki, but noone ever wrote down exactly what consensus is. This might be because writing it down would reveal that we in fact do not even have a consensus about what consensus is ;-)

As far as I gather, opinions about consensus vary from "unanimous" to "majority + a bit more". Some even suggest that this needs to be weighted with the relevance of the arguements put forward (but who is to decide on the relevance?). Most people seem to agree that consensus needs to be well above a majority, but not be unanimous, putting it at some 80-90% of users being in favor, where, of course, the percentages are not clear cut. --Xeeron 10:13, 18 October 2007 (UTC)

Consensus is usually defined as "everyone agrees", but on wikis it tends to work out closer to: "most people agree, and have made a sincere effort to compromise and engage with those who didn't." Regardless of definition, consensus is definitely not about the number that agree with you primarily, but rather the fact that you've attempted to reach a solution amenable to every viewpoint if possible, and if not, you've compromised with those that disagreed so that they can be happy with the end result. --Ari 23:44, 21 October 2007 (UTC)

See consensus is such an elusive animal to catch. If those who disagreed are happy now with the end result, it means that everyone agrees again, so it would be about the number of people agreeing (everyone). As you said yourself just a bit earlier, it usually is handled differently on wikis though. --Xeeron 09:34, 22 October 2007 (UTC)

I find one particular statement on Wikipedia's Consensus to be particularly important: Consensus does not mean that everyone agrees with the outcome; instead, it means that everyone agrees to abide by the outcome. -- ab.er.rant 09:43, 22 October 2007 (UTC)

I don't know about argument relevance, but here are some more things that might be part of the equation:

1. Whether the potential concerns have been discussed enough
2. Number of users weighing in, compared to overall scope of decision
3. Whether there would be a consensus that there is a consensus. (Circular definition...but it should always hold, right?)

Maybe we should consider whether 1RR currently has consensus? It's in the right ballpark, percentage-wise. Discussions on potential concerns have renewed a bit, but it was actually already at the point where the only opposer (myself) felt that the remaining disagreements are largely subjective and maybe not worth further debate. Number of users weighing in is a bit low still? What else is there to consider? --Rezyk 10:37, 22 October 2007 (UTC)

Well, going along with the element ab.er.rant mentioned above, would 1RR be something you could agree to work within were it made into policy, Rezyk? Even though you're not sure you agree with it? (Aiiane - talk - contribs) 12:27, 22 October 2007 (UTC)

Yeah. --Rezyk 06:50, 23 October 2007 (UTC)

I hadn't actually read this draft closely until now. I share some of Ryzek's concerns, but not all of them. While I can work with the current draft, I wouldn't mind seeing one element from the now rejected 3RR migrated to here: specifically, a time duration. So it would read something like "Each editor may not perform more than one related revert on any given part of an article, or the entirety of narrowly focused articles, within a seventy-two hour period." That's just a rough rewording while I'm headed out the door - keeping a reference to concensus in there would be useful too, but I can't work on it now.

Like I said, I could work within the current 1RR draft - this is just a change I would prefer seeing. --- Barek (talk • contribs) - 15:15, 22 October 2007 (UTC)

Let me argue the case against any time duration in the policy. See the Some concerns, continued.. section above for the case of "majority voting by stealth" put forward by Rezyk and why I disagree with it, but for the sake of this discussion, lets for a moment asume that the policy without a time limit would indeed end up facilitating majority voting.

What does a time period change on that policy? It gives more power to the minority. With a time limit, even a single person can go on reverting the article back to his/her prefered state. Essentially it changes revert wars into a battle of attrition. It becomes easy to ignore the other side or to put up the facade of a discussion, without any intention of compromising, while hoping that the other side will sooner or later tire of the issue. Another big problem with battles of attrition is that fanatics are usually highly motivated, while "common sense" people will stop wasting their time on the issue. Thus, even in the worst case of a policy without time limits promoting majority by stealth, I still prefer it to time periods.

For examples about how bad time periods work, check out about any controversial wikipedia page. The revert policy is among the most broken they have. Wikipedia revert wars usually involve editors reverting before even making the first discussion page entry, they tend to go on for months and in more cases than not, they are solved by temporary protection of the article (about the worst way to solve any wiki problem imho). --Xeeron 09:39, 23 October 2007 (UTC)

I do have a minor concern with the "each editor" part of this rule- it sort of implies that you can "gang up" on an article and have your friends revert it for you, if you're really determined. I'd rather simply say that an article should never be reverted more than once until a consensus is reached on the talk page. Does that sound acceptable? This should also stop us from needing to put a hard and fast time limit on the policy, too. --Ari  (talk) 18:58, 28 October 2007 (UTC)

That's 1RV, Ari, which is a proposal that this one (1RR) was specifically differentiated from. The reasons for not using 1RV are mentioned both in the discussion above and on the talk page for 1RV (Guild_Wars_Wiki_talk:Only_revert_once). (Aiiane - talk - contribs) 20:40, 28 October 2007 (UTC)

I think the main advantage of this proposal is that there is no risk of accidental violation. -- Gordon Ecker 04:52, 29 October 2007 (UTC)

If there is no further opposition or objection to consensus here, I intend to move the current proposal (version 447687) into policy in 4-5 days. --Rezyk 06:45, 2 November 2007 (UTC)

Perhaps the wording could use some work. It's not clear to me if "strongly discouraged" and "encouraged" defines policy. Backsword 08:56, 2 November 2007 (UTC)

"Note that a different editor can still revert that part of the article once, but tag-team reverting (having a friend do an extra revert for you) is strongly discouraged." - The full sentence when read together makes it clear what is forced and what is not. Note the bolding. The words mean exactly what they should mean: encouraged/discouraged do not force, and neither does recommended. They're in there because they're points worth making, but there's not enough to really put into an entirely separate guideline. (Aiiane - talk - contribs) 09:03, 2 November 2007 (UTC)

I don't see that. And a "but" statement is often used to give an exception to the preceding statement. Backsword 09:26, 2 November 2007 (UTC)

So you don't see the difference between "strongly discouraged" and, say, "disallowed"? (Aiiane - talk - contribs) 09:41, 2 November 2007 (UTC)

I don't see that everyone would make a distinction in this context, with regard to policy, no. Backsword 10:03, 2 November 2007 (UTC)

Care to offer a suggestion, then? (Aiiane - talk - contribs) 09:51, 2 November 2007 (UTC)

Not my forté. Making it 'allowed but discouraged' would be clear, but too much of a recomendation to do something we'd rather not see. Partly this may be due to the original text being intended as 'scarewording'; trying to make it sound disallowed. Backsword 10:03, 2 November 2007 (UTC)

So "allowed" is okay but "can" isn't? (Aiiane - talk - contribs) 10:47, 2 November 2007 (UTC)

Since I'm just joining in here, I think this is more of a Guideline than a policy.--§ Eloc § 13:40, 2 November 2007 (UTC)

How so, Eloc? It lays out a very clear and absolute restriction, and then adds a couple of comments about actions within that restriction. The restriction is still there. (Aiiane - talk - contribs) 23:09, 2 November 2007 (UTC)

Well, to me atleast, policys are considered rules in which you can be punished for if broken, such as GWW:NPA, but guidelines you can't really be punished but they seem to guide the wiki in a positive direction. I just couldn't see people being punished for more than 1 revert.--§ Eloc § 23:13, 2 November 2007 (UTC)

If someone is intentionally breaking this policy, they would be warned, and if such behavior continued they could and would be blocked for it, Eloc. There's a reason the 'Enforcement' section is present in the policy. (Aiiane - talk - contribs) 23:25, 2 November 2007 (UTC)

That's why I find it kind of strange that you can be blocked for just reverting.--§ Eloc § 23:30, 2 November 2007 (UTC)

You can be blocked for repeatedly reverting, Eloc, because it's disruptive of editing the wiki - would you like it if you were trying to add something to an article, and I just kept removing it with no explanation whatsoever? That's why this policy was created, to stop so-called revert wars. (Aiiane - talk - contribs) 23:32, 2 November 2007 (UTC)

Well, when you put it that way, it makes sense. Damn you Aiiane and you making sense.--§ Eloc § 01:45, 3 November 2007 (UTC)

I was refering to "discouraged", as I mentioned at the start. Never did I bring up "can". Backsword 07:13, 4 November 2007 (UTC)

Yes, the point is that the "can" and the "strongly discouraged" are in the same clause and directly oppose each other, and the "can" is an absolute while the "discouraged" is not, thus the logical conclusion is that the "can" has precedence. (Aiiane - talk - contribs) 11:02, 4 November 2007 (UTC)

We're not dealing with formal logic here.Backsword 11:34, 4 November 2007 (UTC)

What are we dealing with then? "I'll just randomly pick out words and attempt to construct meaning with them"? How does "strongly" turn the word "discouraged" from a suggestion to an absolute? Especially when there's already an absolute statement in the same sentence? Yes, "strongly discouraged" is meant to attempt to keep people from doing it, that's why you discourage things, because we don't want people to do that if at all possible, but we also leave it open because we can't really prove that X is a friend of Y, and thus we don't want to start witchhunts. (Aiiane - talk - contribs) 21:47, 4 November 2007 (UTC)

This seems to be getting out of hand quite quickly. How about we just drop the fighting and get back to discussing the concensus without being so tense.--§ Eloc § 01:49, 5 November 2007 (UTC)

Eloc, we are discussing the consensus, specifically, what holdout issues people might have that would prevent them from reaching a consensus on the current draft, if any. (Aiiane - talk - contribs) 10:44, 5 November 2007 (UTC)

Ok, sorry. Well, I'm giving this sort of a Neutral statement as I don't see many revert wars happening & they don't seem to disrupt the Wiki. Most people will normally just give up on reverting after the 2nd time &/or go to the Admin Noticeboard.--§ Eloc § 15:00, 5 November 2007 (UTC)

(Reset indent) I've made a slight change to the wording of that section, which I hope will resolve Backsword's issue with the previous version. Are there any other direct issues that would stand in the way of this becoming policy? (Aiiane - talk - contribs) 00:55, 7 November 2007 (UTC)

Backsword had a problem with the word "strongly" was it?--§ Eloc § 01:22, 7 November 2007 (UTC)

All of the conversation is on this page, Eloc, if you're still unsure about it feel free to ask him. (Aiiane - talk - contribs) 02:02, 7 November 2007 (UTC)

Ya, but just kind of confused while he's against the word strongly.--§ Eloc § 02:19, 7 November 2007 (UTC)

The fundamental problem is not specific words. When you make unqualified statements that boil down to "Don't do X" in a policy document, it is to be expected that someone will understand that as actionable. But, as I mentioned above, explicitly stating that it is allowed makes it too much of a recomendation; we don't want the effect of the document to be that someone gets an idea for disruptive behavious they didn't know about beforehand.

But you wanted a suggestion. How about we take inspiration from what Reyzk did with the blocking policy and split this into one policy and one guideline. This was originally written before guidelines existed. This would both serve towards a clear and concise policy as well as allowing us to use strong and direct language in the guideline; the fact that it is in a guideline, marked as such and linked to an explanation of guidelines will offset anything misleading as long as we stay within the guideline policy. Backsword 10:19, 8 November 2007 (UTC)

I mentioned it before, there's really not enough here that's not part of policy to make for a useful guideline - the blocking policy had a lot more content and there were two clear components to it, one of which was a hard limit and the other a recommendation. Here, the contents are much more linked and I don't think it would make it easier to understand to split it - having it together in one place makes it easier to read (IMO) than having to reference two separate pages.

"When you make unqualified statements" - last I checked, the portions of the policy you were referencing were not "unqualified". (Aiiane - talk - contribs) 10:31, 8 November 2007 (UTC)

Another issue

Do we need any transition rules? I could easily see how conflict would arise on wheter an edit or revert made before the policy went into effect should count or not. Backsword 10:21, 8 November 2007 (UTC)

This policy governs actions, not results, thus, it would not apply to any state of affairs that existed before it went into effect. In the same sense, if a user had already done one or more edits to a given part of a page prior to the adoption of this policy, performing another before consensus is reached would then be an action that would violate the policy. (Aiiane - talk - contribs) 10:31, 8 November 2007 (UTC)

To be honest, I can't even tell what your position is here. Do any edits and or reverts done before implementation count? Backsword 10:57, 10 November 2007 (UTC)

Reverts done before the implementation cannot be used as grounds to block a user who does not further revert the same section(s). However, if a user reverted a section before implementation (and there was not a subsequent consensus reached), then if they reverted it again after implementation, they would be in violation of policy (as they had an existing revert already performed, and they then performed a second without consensus).

Essentially, no one can be in violation of this policy without having performed at least one revert after the implementation, however they can be in violation of it if they perform one revert after its implementation, but only if they had already performed at least one revert prior to its implementation. (Aiiane - talk - contribs) 13:27, 10 November 2007 (UTC)

So, reverts before implementation counts for the purpose of 1st rv, but not for the purpose of 2nd rv. That would work OK, and I guess it is what would be easiest on SysOps, but it does bring up (again) the standing of talk pages. If someone is warned or blocked for violation of this policy, and then claims they should not be as their first revert predates it, can we just point to this talk page section? What if they say 'that's a talk page, not a policy, so it's invalid'? 11:40, 11 November 2007 (UTC)

Policies are only as valid as the consensus behind them, and the consensus is what is arrived at on the talk page. Consensuses on talk pages are actually more valid than policies, because the latter change to conform to the former, but the former will never change to conform to the latter. That's the core of the wiki model. (Aiiane - talk - contribs) 11:43, 11 November 2007 (UTC)

While that is a good point, one that can be applied on more cases that this, the issue would be determining consensus. In a way, a policy is simply the formal way to note what consensus has been reached. But without such a note it'd be easy to argue that no consensus was reached. Even in this specific issue, despite there being no opposition, which is what it formally takes to break consensus. Backsword 11:53, 11 November 2007 (UTC)

It's not something that's worth having as a permanent part of the policy (IMO), thus I think discussing it here, and allowing a reasonable amount of time for people to voice their comments on it, is the best option. (Aiiane - talk - contribs) 12:27, 11 November 2007 (UTC)

I generally agree that permanent transition rules look ugly, so hopefully this section will be enough to prevent it becoming an issue. Backsword 20:07, 11 November 2007 (UTC)

A block isn't mandatory. Generally, I don't think we we should block people for violating this policy unless they've already been warned or are clearly aware of the policy. -- Gordon Ecker 11:27, 10 November 2007 (UTC)

Proposal -> Policy

Since all significant pending issues seem to have been addressed, and there have been no objections stated to the transition discussion above, I'm going to go ahead and move this to policy status. (Aiiane - talk - contribs) 19:12, 14 November 2007 (UTC)

We finally have a revert policy, lol. Yay! -- ab.er.rant 02:33, 15 November 2007 (UTC)