Summary of new features in V1.7
-------------------------------
1. Various internal tidying and reorganisation of code
2. Bugs and potential crashes fixed:
   a) When setting zones from a PIR alignment file, used to fail if there
      were deletions in both sequence.
   b) STATUS command didn't work properly if atom names contained 
      spaces.
   c) Used to crash if you tried to use ALIGN on a file which contained
      no amino acids.
   d) Used to crash if you specified atom names of more than 7 characters
      or more than 50 atom names
3. Error reporting improved: no atoms read from PDB files distinguished 
   from no memory.
4. New features:
   a) BVALUE command: Allows a cutoff to be applied to B-values; atom
      pairs where the atoms from either structure has a B-value above
      this value are ignored
   b) BWEIGHT command: Weights the fitting by the inverse of the
      B-value such that mobile atoms are less heavily wighted.
   c) Atom name specifications may now include wildcards

Summary of new features in V1.8
-------------------------------
1. New features:
   a) The BVALUE command now takes a negative value to reverse the
      sense of the cutoff and can also be restructed to reference or 
      mobile structures rather than using an average
   b) New option to ignore missing atoms
   c) Added NFITTED command to show number of fitted atoms
   d) RESIDUE and WRITE commands can now output to a pipe
2. Bug fixes:
   a) Completely blank lines in a PDB file caused the previous line
      to be repeated
   b) Multiple-occupancy PDB files where the 'same' atoms didn't 
      appear next to eachother in the file caused a core dump
   c) ILE C-delta correctly written as CD1 rather than CD
   d) Zone setting from sequence alignment had a bug with single
      residue zones
3. Internal changes:
   a) Complete rewrite of the zone setting based on sequence
      alignment. 

Summary of new features in V2.0-V2.2
------------------------------------
   (V2.0 and V2.1 released as test versions only)
1. New features:
   a) Now supports fitting of multiple structures. Each structure is
      fitted to the reference and the average coordinates are
      calculated. All structures are fitted to the average and the
      process iterates to convergence.
   b) Now supports iterative zone updating. After an initial fit, atom
      pairs more distant than a cutoff are discarded and atom pairs
      closer than the cutoff are pulled in. The fitting iterates to
      convergence
   c) Coordinates can be written with centre of geometry of fitted
      region at the origin instead of translated to the reference set
   d) Added QUIET option to switch off warning and informational 
      messages
   e) Added LIMIT command to restrict the region read from an alignment
      to create fitting zones
2. Internal changes:
   a) Code now writes the atom names exactly as they were read instead 
      of cleaning them up.
   b) Some fairly major rewrites to support the multiple structure
      fitting.