Posted to the ba-unrev-talk list:

---------- Forwarded message ----------
From: Chris Dent <cdent@burningchrome.com>
Date: Wed, 31 Jul 2002 17:30:29 -0500 (EST)
Subject: [ba-unrev-talk] testing requested of unrev-ii archive database
    interface


Related to the database based archive of the unrev-ii mailing list
described in various messages on this mailing list, including the most
recent:

   http://www.bootstrap.org/lists/ba-unrev-talk/0207/msg00118.html

New features have been added to the CGI interface to the database that
require testing. These features support the following:

- Searches of the message collection by body keyword, subject keyword,
  or from address can now be limited by the averaged score of
  evaluations of the messages. Evaluations are provided by users of
  the archvie.

In a couple of weeks Kathryn and I will be announcing the official
opening of the archive for real evaluations from members of this
community, but before we do that the system needs to tested for bugs
and confusion in the interface. 

There should be plenty of both. I'm passing around a lot state between
instances of the CGI, and interface design has never been a strong
point for me. Suggestions, comments, bug reports, questions are all
very welcome.

To help out point your browser at

   http://www.burningchrome.com/~cdent/uviz/cgi/index.cgi

(for some context see http://www.burningchrome.com/~cdent/uviz/ )

and perform some searches. If you find messages that are especially
valuable, score them up using the provided tools. At this time any
message without a score has a score of zero. (This may prove to be a
problem in the future, as it makes it difficult to downgrade a message
in relation to messages not yet read. It may be better to give all
messages a base score of 5 or have a range of -5 to 5. We'll see.)

When we have things working well, all scores will be wiped and we'll
start over from scratch.

The overall goal for this project is to generate a multi-faceted
interface to the content of the archive using community evaluation and
annotation, faceted classification, semantic analysis techniques and
simple indexing. 

Your participation is appreciated.