Veropedia

» see veropedia.com as it lived, on the wayback machine

~ the goodbye ~

Veropedia sat inside an argument I was reading rather than one I was in. It started in October 2007, run by Danny Wool, who had been a coordinator at the Wikimedia Foundation until he left earlier that year, along with a group of editors who had done a lot of work on the English Wikipedia and had opinions about which parts of it held up.

The mechanism was simple enough to explain in a sentence. Editors went through Wikipedia looking for articles that already met a reliability bar, checked them, and posted a copy to Veropedia that would then never change. No fair use images came across, because the site carried advertising and could not defend them. Contribution was by invitation. The result was a stable, citable, commercially clean subset of an encyclopedia somebody else was writing.

It grew at roughly three hundred articles a month through the end of 2007, and by October 2008 the database held about 5,800 of them. Then in December 2008 the content came down. What replaced it was a notice saying the team was building a new version with better article handling and a better interface. Nothing came back. The site was gone through 2009 and the domain has been dead ever since.

The problem it was aimed at was real and I was on the receiving end of it. In 2007 I was finishing an engineering degree, and Wikipedia was already the reference layer for everything, while also being the thing you were not allowed to cite. Veropedia's answer was to hand you a version of the same paragraph with a promise attached. That answer never worked, because the promise was the only new thing in the box.

~ the chronicle ~

~ the argument it was making ~

Wikipedia's open editing was the source of both its coverage and its credibility problem, and in 2007 half the people writing about the internet believed the second would eventually kill the first. Larry Sanger's Citizendium answered by requiring real names and expert oversight on original writing. Veropedia answered by leaving the writing where it was and taking a photograph of it at a good moment.

The photograph idea was not stupid. Wikipedia itself had been arguing internally about flagged revisions and stable versions for years, and the German edition went on to actually ship a version of it. Veropedia's bet was that it could do that from outside, faster, and sell advertising against the result.

~ the same text with a different logo ~

What was actually different was almost nothing, and the critics said so immediately. Most Veropedia articles were identical to their Wikipedia counterparts, because that was the design. The vetting selected articles rather than improving them. So a reader comparing the two pages saw the same paragraphs, minus some images, plus ads.

It also skipped the hard cases by construction. An article stable enough to freeze is an article nobody is fighting over, which meant the contested topics where a reader most wants to know whether the page can be trusted were the exact topics Veropedia had no entry for. Nicholas Carr wrote it up at the time as mining Wikipedia rather than building on it, which was unkind and roughly correct.

So it died by announcing an improvement, which is a particular kind of dead. There was no acquisition and no lawsuit and no bandwidth bill anybody has ever mentioned. The people funding it were mostly the people running it, and at some point in December 2008 they took the pages down and wrote that a better version was coming. Knol did the same thing on a much larger budget two years later and ended the same way. A reference work is not a quality process bolted onto somebody else's text. It is the text, and if you did not write it you have nothing to keep.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »