Updated versions of the RHESSI roll database routines have been posted
to atest.  These include changes that Andre made up to Sept 19. 

The roll database itself now covers February 2002 through August 2003
and has been put in dbase/aspect as a .fits file.  The RAS solution
for the July 23 event is integrated transparently so that the
long-standing kluge using interpolated PMTRAS data is no longer

Still to be done (in order of decreasing priority): 
* Breaking up the database file ( >7 Mbytes and growing) into monthly
* Development of automated updating scripts.
* Integration of RAS data for flare-relevant gaps.
* Further reduction in the number of gaps (now 5.3%).

Please let me know if you encounter any problems or if you find any
suspicious flare positions.