User:Crazynas/thoughts on indefinite semi-protection

Protected pages

My current project is a backlog I seem to have discovered myself. Indefinitely Semi-protected pages. I find that they broadly fall into seven categories:

1. Obvious semi-protection: Such as obscenities very high profile BLP's and obviously controversial topics. This is essentially the category that you know without even looking into it why it's protected.
2. LTA or ArbCom: Pages to deal with some (or one) troublesome individual(s). These generally come in two varieties. They are either an abuse report (for persistent individual vandals) or an arbitration remedy (for extreme content disputes).
3. Redirects: Protected because forks or recreations are at high risk of being created.
4. BLP's: These fall into two broad groups: OTRS tickets, or not. Of those that are not OTRS, they are usually either persistent vandalism or some form of more complex attacks.
5. Low profile vandalism: Something that is less obvious then the obvious.
6. Edit wars: A small but notable number of pages have had protection for years over an anonymous edit war.
7. Other cases: Some non standard reason, or a request to contact before unprotection.

When auditing the logs, my primary question is: does this page have a reason to be protected now? 'the obvious', LTA/ArbCom, redirects, and BLPs protected because of an OTRS ticket I consider an uncontroversial application of the protection policy and pass over without question. Vandalism protections, edit wars and BLP's protected without a ticket I examine further to determine if they still warrant the protection.

When examining pages I look at the edit history and talk page. If there is somewhat recent auto-confirmed vandalism or POV pushing, I leave it. If it's a BLP I consider if it is still a high vandalism target today, or possibly whether it is a good candidate for pending changes. I almost always request unprotection of pages protected for edit warring . I made some mistakes starting out, and it took me a while to understand LTA. However, after I reviewed the active list of abuse cases I find I can recognize most of them.

There are slightly more then four thousand indefinitely semi-protected pages in article space >500 bytes (this excludes redirects). I feel that indefinite semi-protection should not be applied in cases of low profile vandalism or editing disputes, instead I feel our semi-protection policy needs to be amended so that administrators should apply some standard (such as one year, followed by two years) in these cases. Although it would increase the workload marginally for administrators to review the pages unprotected, the benefit of the pages that are protected for years without reason would be great. I currently stand at over 90% success rate of the articles I request unprotection on at WP:RFPP. At my current rate, working from the back of the log, there are possibly more than 500 pages that could be unprotected.

Unless the case is particularly unclear, or there is a specific note, in the protection log, I twinkle a RFPP with my recommendation as to the change I think needs to happen. Often there are borderline cases where I have a secondary recommendation (this is often pending changes protection, where I think the page should be unprotected, but is a suitable candidate for PC should the sysop determine that protection is still warranted). In cases where their is a request to contact before unprotection, if the administrator is still active I, of course, contact them.

Pending Changes

I don't think the benefits of pending changes have fully been realized, and in some cases, this would be a better solution. There is an argument that pending changes doesn't work on a large scale because of the backlog it would create. I would argue that if we relaxed PC1 slightly more, and allowed all autoconfirmed editors to accept revisions it would go a long way in allowing us to expand the scope of this basic measure of quality control. This would make PC1 and PC2 match up, insofar as they both have accept and edit privileges at the same threshold.

PC2 is a much more complicated issue. I feel the best way to make PC2 an accepted alternative (and a backdoor way to relieve sysop shortages) is to decide that there is no policy difference between PC2 and full protection. Revisions would have to be accepted by uninvolved editors, who would be required to determine that the edit was either uncontroversial or had prior consensus. It could also be used in non-mainspace areas (such as templates) where protection is required for security, but reviewers would be trusted to edit, not, 'mess things up' as it were.