Wikipedia:PC2012/RfC 2

Pending changes discussions

The following discussion is closed. Please do not modify it. Subsequent comments should be made in a new section.

Some sections are far more obvious than others, so we'll start with the easy ones. There is strong consensus to use pending changes in response to vandalism and non-preemptively for violations of the biographies of living persons policy (though for very high-traffic articles such as George W. Bush, semiprotection has proven to be the better solution), and similarly strong consensus not to automatically accept revisions after some length of time. Also, the idea that anti-vandal bots should be able to revert to the latest accepted revision without requiring a reviewer to then make a null accept didn't meet with any opposition.

To the others, now. Beyond use for vandalism and BLP violations, there is no consensus here to use it for the other scenarios listed below. As to when reviewers should reject edits, everyone agreed obvious vandalism, BLP violations, and copyright violations should always be rejected, but beyond that it was less clear. The proposal getting the most support was WhatamIdoing's five point proposal, but there was a significant minority which felt that was a bit too complicated. Therefore, it appears that the best way forward would be to use this as a starting point and make any modifications/simplifications to it once we see Pending Changes in action.

While there seemed to be no problem with automatically accepting anti-vandal bot reverts, there did not appear to be consensus to automatically accept reversions from non-reviewer editors; this is again something we should monitor upon implementation to see if it would be the best solution. Finally, there was a proposal to ensure we slowly ramp up the use of PC and watch the backlog. There was consensus this was a good idea, but there was disagreement over what the threshold at which point admins should consider reducing the number of pages under PC was. The proposal stated an hour, but several editors argued it should be longer; the most common was four hours, so this appeared to be what had the most consensus.

Further discussion about those issues which require analysis upon implementation should be monitored and discussed whenever we have sufficient data. Obviously there is no way to definitively state when that will be, but it seems likely that within a month we should be able to spot and fix glaring problems. Within several months we should be able to work out other, more nuanced issues, including those mentioned above and anything else that comes up. The Blade of the Northern Lights (話して下さい) 18:15, 22 October 2012 (UTC)[reply]

This is the second informal Request for Comment on various ideas for changes to the provisional policy on Pending Changes (PC).

It includes questions on when to apply Pending Changes, criteria for rejecting edits, whether to automatically accept old changes or reverts, and whether to specify in the policy how frequently it should be used in the early days. 12:00, 01 October 2012 (UTC)

Procedural note: Because several questions beyond the scope of this RfC remain to be settled (preferably before PC use formally begins on December 1st), an additional RfC may be required. After a minimum discussion period of one week, this RfC may closed. The mechanism for deciding when to close the RfC is a simple majority vote.

"...it is clear that though the community supports the adoption of Pending Changes, the community is also well aware that both the Pending Changes Trial of 2010 and the draft policy as presented in this RfC suffer(ed) from weaknesses. Therefore...the community should dedicate itself to determining the implementation of Pending Changes that it wishes to be turned on. Many opinions have been presented in the various areas of this RfC as to what works and doesn’t work in Pending Changes as implemented/proposed; the community must now focus its energy on optimizing the implementation of Pending Changes that it wishes to see go live."

— Closing admins of the 2012 Pending changes RfC

Current provisional policy

When a page with Pending Changes (PC) protection is edited by an unregistered user or new user, that edit and all following edits by any user are not included in the article displayed to the general public (that is, for readers who are not logged in), until the edits are approved by someone with the "reviewer" user right. PC protection is intended to quash inappropriate editing while allowing good faith users to submit their edits for review.

Reviewers are users with a similar level of trust to rollbackers (including all administrators) and the right can be granted and removed by any administrator. Reviewer rights are granted upon request at Wikipedia:Requests for permissions. Potential reviewers should recognize vandalism, be familiar with basic content policies such as the policy on living people, and have a reasonable level of experience editing Wikipedia. Reading the reviewing guideline, where the reviewing process and expectations for a reviewer are detailed, is recommended.

Pending changes should be used on pages where the disruption to good-faith editing caused by existing protection tools would be disproportionate to the problem the protection seeks to resolve. Suitable issues for pending changes protection include persistent:

Vandalism;

Re-insertion of rumor, error, and NPOV/V/OR violations;

Edit warring by large groups of unregistered users;

Disruption by users on highly variable IPs.

These standards are to be interpreted more liberally on biographies of living persons, or in any situation involving content related to living persons. As with other forms of protection, PC protection should not be used preemptively.

As with other forms of protection, the time frame of the protection should be proportional to the problem. Indefinite PC protection should only be used in cases of severe long-term disruption.

Like semi-protection, PC protection should never be used in genuine content disputes, where there is a risk of placing a particular group of editors at a disadvantage.