scoring problems

I'm still running into problems with CRM 114 scores clustering around the red
limit of my user interface for training.  The end result is that over time, the
user interface gets filled with a whole bunch of clearly bad messages and the
user experience goes to hell in a handbasket.  This problem is bad enough, that
it alone has driven away users and they are left with a very bad impression of
CRM 114.

It's possible that the problem is of my own doing but, I've had a few people
look at the code and, so far everything looks right on my end.  I could use some
help building a test case for Bill on this.  If someone has some spare cycles,
is familiar with Python and SQLite I would appreciate the help in building some
code to help diagnose the problem.


---eric

------------------------------------------------------------------------------
Stay on top of everything new and different, both inside and 
around Java (TM) technology - register by April 22, and save
$200 on the JavaOne (SM) conference, June 2-5, 2009, San Francisco.
300 plus technical and hands-on sessions. Register today. 
Use priority code J9JMT32. http://p.sf.net/sfu/p