Wikipedia's next five years
The wiki medium, Socialtext's Ross Mayfield has said, "denatures personality." Whereas open source developers are often motivated by a desire to build strong individual reputations, Wikipedians care more about inclusiveness and consensus. Enforcing better accountability without eroding these core values will require tricky social engineering.
Wikipedia is subject to a lot of armchair quarterbacking these days, so the following recommendations should be taken with the requisite grain of salt. I'm an interested observer and sometime reader of Wikipedia, but rarely a contributor, opting instead to use my limited supply of keystrokes here. That said, outside perspectives may help clarify ways to ease Wikipedia's transition into the mainstream of society. So here goes.
Judging the quality of article entries is an even trickier matter. Like open source projects, Wikipedia is a meritocracy that rewards hard work and excellent results. Neither culture values the Ph.D. for its own sake. But where open source operates in a single domain of expertise, Wikipedia spans myriad domains. While it can't hire expertise à la Digital Universe, a program to attract qualified volunteer reviewers could gain traction. [Full story at InfoWorld.com]
RSS feeds. As Tim Bray and doubtless many others have noticed, Wikipedia (unlike many wikis) does not offer feeds. That's a non-issue for folks who are closely involved with the process. But for those who want to be more loosely coupled to it, syndicated feeds of watchlists, page edits, and statistics would make staying in touch a whole lot easier.
There was excellent progress on this front last summer, as I documented in a screencast. However, the proof-of-concept hasn't advanced since then. It's remarkable that third-party scripts can animate the flow of changes, but the support for such visualization properly belongs in the MediaWiki engine.
Although Wikipedia's change history does differentiate between minor and major edits, there's nothing corresponding to stable versions in open source software projects. In the early life of most articles that would be overkill. But for more mature articles, and especially active ones, version landmarks might be a useful organizational tool. Of course it's an open question as to how exactly a version could be declared stable. When the current crop of contributors agree to put a stake in the ground? When invited experts agree? A joint consensus? Whatever the convention might be, Wikipedia seems to be very adept at inventing and following conventions. It's interesting to imagine a view latest stable version option with, of course, a transparent view into whatever "stable" is defined to be.
Savvy users of Wikipedia know that it's a jumping-off point for research, not a final destination. Whatever you may think about the quality of an entry on a given topic, it is often a great source of relevant external links. But there's no aggregation of inbound links. I use the term trackback metaphorically here, because I don't think a literal implementation of the trackback mechanism would be appropriate. However, the reaction to a page -- as expressed in bookmarking services, on blogs, and elsewhere -- can be assembled and displayed. This stuff need not, and arguably should not, be included directly in Wikipedia. It might more properly be done using a third-party service in cahoots with a local Greasemonkey-style transformation.
Wikipedia's fifth birthday is right around the corner. It's been an amazing half-decade, and I'm really looking forward to the sequel.
Former URL: http://weblog.infoworld.com/udell/2006/01/04.html#a1363