This is the long awaited summary for the UIDE paper class
discussion, written by John Canning



The presentation of the UIDE paper was never completed due to the
general conversation that started.  Here are some of the issues
discussed.



1.  The importance of the program giving the user a choice of
whether to accept changes in the format of the program

2.  There was also some question as to how the user model was
created.  It was argued that the statistical analysis should take
into account mistakes.

3. Actual systems that implemented the UIDE system such as the
train scheduling system were discussed.

4.  The portability of the UIDE into other systems was explored
as well as the ability to rebuild user models.

5.  We also discovered that many of the functions described in
the paper are theoretical applications that art yet to be
implemented. 


John Canning