I wish Larry would stay on the soapbox a little longer. I'd be happy to bring him
coffee and donuts...
I see two parallel and complementary ideas regarding certification being discussed here:
1. Improving signal to noise ratio
2. Verifying quality of scholarship
Simple, mechanical schemes like "some signed-in user read or wrote this" can
help distinguish between sheer vandalism (noise) and someone's sincere attempt to
write useful stuff. This is like the squelch control on a radio.
On a higher level, we need to verify content on important matters that only expert
reviewers are competent to judge. This is roughly analogous to what the military calls
"authentication" (like a PGP key).
There's no way I would want Wikipedia degenerate into a back-patting, incestuous,
self-congratulatory mutual admiration society. Like a newsgroup where everyone agrees with
each other. That way madness lies.
Nor do I want to see Wikipedia become the victim of its own success. With 100,000
signed-in users and 10,000,000 hits per day, how will we be able to stop the graffiti
artists and script kiddies? Once the goat man gets a following...
I still think we should offer our "customers" advanced filtering features.
Libraries and moms will want a no-porn guarantee. A CD-ROM publisher donating to the third
world (maybe a Unificationist organization like Professors for World Peace) might want to
focus more on scholarly topics, and not care about pop music reviews.
We can make the filters as sophisticated as we want. I haven't volunteered (yet) to be
a developer, but I'm actually quite good with pattern matching and SQL joins. What I
mean is, I'm willing to write code. But it's more a problem of deciding on
features than figuring out how to implement them.
So let's take some more about these (and other) kinds of certification. What do we
think the public wants and needs?
Ed Poor