On Tue, Aug 25, 2009 at 7:46 PM, Thomas Dalton<thomas.dalton(a)gmail.com> wrote: I wonder if Jeremy Paxman or his researchers read this list? :-) Carcharoth

2009/8/25 Sage Ross <ragesoss+wikipedia(a)gmail.com>om>: Yep, that page is the closest thing there is to a canon summary of the proposed configuration for en.wp. Generally next steps are: 1) Finalize setup of http://flaggedrevs.labs.wikimedia.org/wiki/Main_Page with article import from en.wp; 2) Test proposed configuration (UI and workflows) and make necessary fixes; 3) Rollout on en.wp. 3) should ideally happen within 8 weeks or so, but that depends in part on people's experience during the testing period, and the extent to which we have to make further revisions to the extension. The FlaggedRevs extension has been used in many of our wikis, including the second-largest Wikipedia, for more than a year. However, contrary to what's been reported in some media, the community has had very thoughtful conversations about the ideal setup for en.wp, which isn't going to be the same as the one used for de.wp, and we want to make sure that the software properly supports it without causing confusion, especially in light of our general efforts to make Wikipedia easier to use and easier to contribute to. WiFi in Buenos Aires willing, I'll try to whip up a summary for the Wikimedia blog at blog.wikimedia.org tonight to help reduce some of the confusion. -- Erik Möller Deputy Director, Wikimedia Foundation Support Free Knowledge: http://wikimediafoundation.org/wiki/Donate

2009/8/25 Joseph Reagle <reagle(a)mit.edu>du>: IME the problem is the use of the word "editor". We use it as in "tens of thousands of volunteers", everyone else assumes we mean as in "the boss who decides what goes in." It's a jargon versus English problem. Starting with all the admins. Presumably any active editor who isn't a problem child will get it too, similar to rollback. - d.

Try this as a general approach: - This is not a new idea. - We've got vandalism down to under one article in 200, and a variety of advanced programs and patrols of hundreds of users who get the usual fixing time down to seconds or minutes when it does happen. - But obviously we want to do even better. - A lot of people gauge Wikipedia in terms of quantity of edits. The last 2 years the focus has been on improving quality of edits, and especially, finding even better ways to prevent deliberately harmful edits such as vandalism. - Our historical answer is "protection" - everyone is prevented from editing a page if it is being badly mis-edited. That's highly disruptive and frustrates many edits since one bad apple can hold up the process. - A more recent addition is the Abuse filter, a system that allows flagrantly bad edits to be prevented but lets through good ones. It's a program though so it can't differentiate apparently good posts that are really not good. - Our newest answer is therefore this thing called "flagged revisions" - the requirement that when an edit is made to a sensitive article, someone who's been round a while, which is one of thousands of users, checks to say it's okay, before letting it go "live". - Our test bed has been the german wikipedia, the second largest language to English in the Wikipedia websites. - Our main target and test bed is articles about people, because those are seen to be more sensitive and of special importance to get right. The wider reported vandalism cases usually relate to these articles, just because articles about people are so visible. So it makes sense to apply possible solutions to these and see what effect it has on editing quantity and quality. That's how I'd explain it (condensed and simplified as needed for the media concerned). FT2 On Wed, Aug 26, 2009 at 12:26 AM, Thomas Dalton <thomas.dalton(a)gmail.com>wrote;wrote:

2009/8/25 Sage Ross <ragesoss+wikipedia(a)gmail.com>om>: Yep - flagging as and when needed, in place of protection. If we have days-old unapproved revisions at the end of the two months, that'll be a failure because editors wouldn't stand for it. But, speaking as a BIG FAN of flagged revs for BLPs, I'll be trying to do my part to make this work! - d.