2005-11-04
  Extra ' ' printed after key.

2005-11-04
  Things to Debug
    (x) total 1 n <file>
    (x) total -s1r 1 6,7 <file>
    (x) ./total -m 1 6s1 1s1 <file> works but not
        ./total -m 1 6+1 1+1 <file>
          * Had to fix convert_data_inc().
  Add
    (x) -m option to low level control.
    (x) when doing manual, make sure that out-> does not read column
        beyond what inc-> stored.
    (x) i and I options to key (for print internet and padded zero internet)
	 (x) Add multiple input files, have total do gzip or not as needed.
    (/) -g,-p for gzip and plain text input.
          Using multiple files so we can't have a single -g or -p option
    (x) Autodetection of .gz suffix for gzip.
    (x) See by argument '-m 6+1,7+1 1+1' doesn't work 
            * It does work.
    (x) Write single open,close,gets function which work with plain,
        stdin and gzip.
    (x) Add buffering gzip to speed it up.
    (x) Add comma formatting.
    (x) Copy original code from make_key() to make current make_key()
        faster.  (Makes total about 10% faster).
    (x) Add -i option to use atof_quick() for reading integer fields.
        Improves speed up to 10%.
    (/) Make separate convert_data_inc for firstime and other times,
          check speed improvement.  
         (original total is faster then new total when new total is
          not using gzip, -i or I.  Either because of make_key or
          convert_data_inc).
        2005-11-13.
    (_) Add 1+2 automatic rule?
          i[-+]j[-+]k[ads..],

  makelocal -
    (x) Add ability to read gzip files.
    (x) Add ability write summary to file while writing convert to
        stdout.
     DEBUGING
       This works
          ./makelocal -s -o 1.out 68 work/total.in.gz
       but this doesn't
          ./makelocal -s 68 work/total.in.gz
     


2005-11-11
* report.sh takes about 2m18s with 30min of real data
  (/tmp/test.dat.gz).  Has all 30 minute report calculations
  except first nine header fields.
    Connection, Packets, Byte, Incoming, Outgoing, Total,
    Internal, External, Other.
