Thursday, November 29, 2007

Metapedian, 2007-11-29

ArbCom elections underway (but you probably already knew that); self-nomination period ends tomorrow. Voting begins December 3rd.

Current adminship candidates: thedemonhog, Hbdragon88, Dlohcierekim.

New bureaucrat: WJBscribe.

A software bug lasting approximately 20 minutes caused many errors while saving edits to pages including tags such as , , and . More information, including cleanup effort, can be found at [[User:Splarka/UNIQ list]].

Seven cases currently before ArbCom. In the evidence phase: Anonimu, Episodes and characters, Winter Soldier 2. In the voting phase: Durova and Jehochman, Macedonia, Asgardian-Tenebrae. Current motion to close on Privatemusings.

Core Contest aims to encourage article improvement; offering $100 to authors of articles which see the most improvement before December 9th.

The Ombudsmen Committee, charged with investigating alleged violations of the privacy policy, has three new members; more information on foundation-l.

This doesn't quite feel like what I'm going for, but it's a first shot at it. May take me a bit to get into the swing of it. Feedback, comments? Feel free.

Tuesday, November 27, 2007

Scaling discussion: The Metapedian

Anybody who's spent even a small amount of time on Wikipedia could tell you that there's a huge number of discussions going on at any moment, between dozens of noticeboards, a hundred or more deletion debates per day, and a countless multitude of user talk pages and dispute resolution processes. It's easy to get lost, or have difficulty keeping track of some important discussions which have wider relevance across the project, or sometimes even across multiple projects. Over time, as the project grows, and more users become involved, this problem will only worsen.

People need to be able to find important discussions, so that they can offer their input, or just keep an eye on things. This, in turn, can lead to a second difficulty: scaling individual discussions to allow for more participants, so that the signal-to-noise ratio doesn't get too bad, and important comments are seen.

To put it more directly, there's two problems, here: people need to be able to find the most relevant discussions, and they need to be able to extract keystone comments from those discussions. The forum itself must scale, as must the discussions within that forum. That second problem is probably the harder one to deal with.

On Wikipedia, we've scaled the forum by splitting up discussions among multiple pages (more traditional forums do the same, organizing discussions into topical subforums), and while users can watch individual pages, that clearly doesn't solve the problem of sorting through thousands of discussions. We have some wonderful efforts at the Wikipedia Signpost, providing general project news on a weekly basis, or the centralized discussion linkbox, offering a quick peek at a dozen or so discussions at a time. But what about timely discussions which get some attention, without quite reaching that level of long-term, project-wide concern? At that sort of level, there's not really much more than keeping your own eyes and ears wide open, as far as I can tell.

Links passed around in mailing lists, IRC, and other areas can help us find discussions, but there's a bit much noise mixed in with the signal. Is there a more complete solution? I'd like to think a link aggregator of some sort might help. That way, at least, we're increasing the availability of these links to a wider audience, so that more people can see them, and reducing the amount of noise surrounding those links, so that they're easier to find.

So, yes, something vaguely akin to Fark or Slashdot, focused on Wikipedia. It seems to me like it could scale, if managed by the right people. Lacking the resources to do anything like that, at the moment, I suspect I'll be blogging about it. Call it “The Metapedian.” If people are interested and feedback is good, it could continue or grow, but we'll see what the future holds.