Photo: Flickr user lifeontheedge

Friday, February 01, 2008

The people who brought you the 10,000-year clock are excited about Wikipedia's potential as a canonical "Futurepedia".

Think about articles like 2020 -- "If expanded greatly the official future timeline might prove to be a useful document of what we expect."

I hate to be the one to break it to them, but there's no way that's going to happen within Wikipedia proper. It's not verifiable. It doesn't cite authoritative sources. It's not encyclopedic. Basically, it breaks all the rules, and it'll get slapped down before it gains any steam. Wikipedia's incredible realized potential lives side by side with a lot of squandered potential.

You're not going to win this one by fighting it out in the AfD trenches -- anyway, 1head-to-head combat isn't the wiki way. Wikipedia desperately needs to be nested inside some broader, more permissive architecture.

2 comments:

Jeff said...

Does there need to be a broader architecture, or can we simply count on the growing swarm of wikis to eventually subsume every potential purpose, including a Futurepedia?

Many of these, including a Futurpedia would, as you point out, violate some of the rules of Wikipedia; they would have to arrive at their own set of rules by trial and error and consensus, and in some cases might prove to be fundamentally unworkable. I started Wikipopuli.com as a home for articles that violate WP:BIO (non-notability), which is a transgression that Jimbo seems particularly to despise. It remains to be seen whether its goal of being a universal repository of biographical info is workable and stable or will come crashing down in a flood of spam and libel. In the world of wikis, it all comes down to the behavior of the crowd, which no one can predict.

Ben Yates said...

Crowds are hard to predict, but not impossible, and there's definitely stuff you can do to create the right atmosphere. In a few years people are going to be getting bachelors' degrees in online crowd-shepherding.

Does there need to be a broader architecture, or

The direction all comes down to what happens in the next couple years. The following technologies are all totally workable:

* A firefox addon or greasemonkey script that displays other content alongside articles whenever the user visits Wikipedia. (This "mashup engine" could also exist on a server somewhere instead of in the user's firefox folder. If I was a programmer I'd mash wikipedia and everything2, partly to liberate e2 from its godforsaken interface.)

* A wiki superset that functions like this: whenever you visit a wikipedia article page that's been deleted, you're shown the article anyway -- the article is pulled in from some other server and displayed seamlessly. (You've got to have the extension installed, of course.) Note how this sidesteps the on-wiki constraints -- you don't have to convince anyone to keep articles or change how the "this article doesn't exist" page looks; you just do it yourself and pipe it in.

There are a bunch of other ideas, but the key is implementation -- the approach that spreads will the the one that doesn't fuck it up -- and you have never seen fuckups like social software fuckups. (Interface fuckups, I mean. I'm not trying to insult programmers who are geniuses in other respects.)

Take Wikipedia's article-doesn't-exist screen. There are 2 pieces of information it's absolutely essential for the page to convey:

1. Whether this article existed at some point in the past. If not, it means you've mistyped a URL or followed a malformed link. If so, you'll probably want to know ...

2. Where you can find the article. Because you can always find it somewhere. No exceptions. If the article has been moved to another site, there should be a link. If not, the page should inform people of Wikipedia's best-kept secret -- no information is ever really gone, and you can always ask an admin to send you whatever article once existed. There should be a link to an admin who can help out.

The existing "no article here" screen flubs #1 (there's a tiny "deletion log" link that shows up even if there's never been an article, and is buried under a mountain of irellevent text if there has) and utterly fails at #2. 99% of deletion pissed-off-ness is because people think stuff was actually deleted!