On 6/14/06, Delirium <delirium(a)hackish.org> wrote:
(Mainly concerning wikipedia, but cross-posting to
foundation-l because
of some discussion of committees; see the end.)
We've discussed on and off that it'd be nice to vet specific revisions
of Wikipedia articles so readers can either choose to read only quality
articles, or at least have an indication of how good an article is.
This is an obvious prerequisite for a Wikipedia 1.0 print edition, and
would be nice on the website as well.
I'm working on a set of consolidated specifications for this, please
give me a few days until I finalize it (it's a fairly large document).
Allow me to quote from the set of requirements I start with:
* revision-specific tagging rather than article-specific tagging
* changeset-oriented: when possible, only review the changes from the
last reviewed revision to the next unreviewed one
* scalability in the overall quantity of articles reviewed per time unit
* distinction between different types of review, such as vandalism,
accuracy, neutrality, copyright status, etc.
** acknowledge that different people have different abilities to
review these different aspects of an article
* systematically involving editors who are self-selected as being
qualified in the particular disciplines
* universal applicability to Wikipedia, Wikinews, Wikisource,
Wikiquote, Wiktionary, and any other open wiki project using MediaWiki
* facilitate fixing problems over tagging them
* discoverability. All aspects of the user interface should be
discovered by the average wiki editor during normal use of the system.
* fun. Any review process must be addictive and simple in order to
motivate long term interest.
I've been mulling over this problem since 2002 and I think that all
proposed processes fail in several of these criteria. That is not to
say that I believe I have the "one answer", but I hope the specs I'm
working on will become a basis for further renewed discussion, and
lead directly towards a realistic implementation path.
Erik