New in QM 1.31
  
 - QM processes up to 2048 echomail areas.
 - Messages as large as 30k are processed.
 - QM can swap itself to EMS or XMS or disk during toss and
   pack shell operations.
 - The QM duplicate message tracking database is stored as
   one file, eliminating the need to maintain directories
   for passthru echomail areas.
 - Circular Path Detection (CPD) can be used to detect and
   stop echomail dupe loops.
 - Qm_route.Cfg "Dos" directives are executed every time
   MaxMsgs packing is done rather than only after all mail
   has been tossed/scanned.
 - All shell operations (unpack/pack/dos) can execute BATch
   files for virtually unlimited external control.
 - LHA and ARJ compressed mail is recognized and processed.
 - Missing Outbound.### directories are created as needed.
 - QM can generate outbound mail using aka address matching
   by zone.
 - RouteThru echomail is available for pre-arranged
   destinations.
 - The scan buffer size can be set to 32k or 64k for thruput
   improvement. The buffer was only 20k in v1.00.
 - Compressed mail can be deleted between unpack and toss for
   better disk management and less complicated recovery.
 - A message reply link feature is built in.
  
Fixed since QM 1.00
  
 - Areaname lookup does not crosslink areas if CRCs match.
 - Echotoss.Log can be generated without Qm.Log.
 - Message header dates are not grunged during toss.
 - Maxmsgs exit errorlevel is 5 as documented.
 - Maxmsgs exit now flushes statistics to Qm.Log.
 - Areas.Bbs files larger than 64k can be used.
 - Bundle filenames beginning with - are ignored.
 - Duplicate detection no longer uses EID kludge lines or
   information from the message body in order to reduce false
   dupe conditions.