Risker,
This is a rather belated response to some points you raised earlier about pending changes.
On Wed, Apr 18, 2012 at 5:47 PM, Risker risker.wp@gmail.com wrote:
Having been very involved in the trial, I would not re-enable the use of Pending Changes until significant changes to the proposed policy are made. Most of the problems that were encountered in the trial are left completely unaddressed. There should be a prohibition on it being used for articles larger than 55K - after that point, too many people crashed when trying to review.
That's never happened to me in de:WP, so I think it's a software problem that is fixable (and seems to have been fixed long ago in de:WP, if they ever had it).
There should be a prohibition on its use for articles that are moving rapidly; contrary to what some thought, pending changes was not really effective for current events articles, because the proposed edits were being overwritten before anyone even reviewed them; and because there is no way to review a single pending change at a time (instead of ALL pending changes), it is inevitable that either bad edits will be accepted or good edits rejected.
It could be a problem for very fast-moving articles - like an edit a minute, in response to some news event. But I know that the Germans manage, and I have never seen it raised as a problem there. The worst thing that could happen is that IPs make changes which never see the light of day, whereas in en:WP they would have been visible to the public briefly before being overwritten. In either case the solution is to slow down.
I haven't found reviewing several unsighted edits a huge problem in de:WP – yes, it can be a pain if the 1st, 3rd and 5th edits were good, and the 2nd and 4th weren't, but that situation is relatively rare. On the few occasions where it has happened to me, I opened a second window with the last sighted version and manually transferred the good changes. It's doable.
I'd keep pending changes off of biographical articles that have a history of attracting vandalism or excessive vitriol or fandom. Using pending changes for these articles effectively enshrines the otherwise-never-existing vandalism into the history of the article. We saw this in quite a few highly visible biographies.
It's perfectly possible to have semi-protection in addition to pending changes. The Germans have pending changes as default on all articles, but still use semi-protection or full protection alongside whenever there is IP vandalism, or an edit war.
Everyone needs to be clear what exactly the role of the reviewer is; this created a considerable amount of strife during the trial. I have been given various interpretations of the manner in which flagged revisions is used on German Wikipedia, so do not want to characterize their policies and practices; however, in the absence of good quality, confirmed information on their processes, it's not appropriate to say "let's do it like they do".
The German Wikipedia has passive and active reviewers. The main rules given at http://de.wikipedia.org/wiki/Wikipedia:Sichten are as follows:
Passive reviewers autoreview their own edits, but can't review others'. Passive reviewing rights are automatically given to users who have been registered for at least 30 days and have made at least 150 article edits (or 50 article edits subsequently approved by a reviewer).
Active reviewer status (i.e. the right to approve others' edits) is automatically conferred on users who have been registered for 60 days and have made 300 article edits (or 200 article edits subsequently approved by a reviewer).
There are some additional details (no blocks, use of edit summaries for at least, work spread out over a number of different articles, etc.), but these are secondary.
The system works and keeps out a lot of nonsense. The only thing I would change is that I would set a higher standard for users wanting to approve BLP changes.
Cheers, Andreas
Until it's clear what the role of the reviewer is, editors have no way to know whether or not they are performing in the manner that the community expects. Further, there is no guarantee that reviewer permissions won't be removed for reasons that have nothing to do with the act of reviewing.
The proposed policy essentially says " you can use this instead of semi-protection", but it does not change the criteria for protection in any way. Therefore, the articles you propose to be covered by pending changes aren't eligible. What if you think something should be under PC, and another admin comes along and says "hold on, doesn't meet the policy, off it comes"? Right now, decisions about protections are rarely the subject of inter-admin disagreement. Is that going to change? If so, who wins?
The RFC started from the wrong place. It should have been focused on what kind of PC policy we would want to have if we wanted to have one. I do see potential uses for pending changes, but I do not support the policy that is being put forward. https://lists.wikimedia.org/mailman/listinfo/wikien-l