1.  create directory in bitbucket /mit/bitbucket/wade
2.  copy /mit/ti_data/admin/pips.web to bitbucket
3.  remove the rfc record from pips.web (because it's too long for awk)
4.  type: awk -F: '{print $6}' pips.web | \
	tr "ABCDEFGHIJKLMNOPQRSTUVWXYZ" "abcdefghijklmnopqrstuvwxyz" | sort | uniq > u2
    [ this produces a unique lists of provider sources and rolls any
      uppercase names to lowercaes ]
    *** remember to add the source "all" to the list to get a listing
    *** of all files read, not just by a specific source
5.  copy the three read log files from /mit/net_dev/techinfo.archive
    to bitbucket and then uncompress them.
    [ this represents the read.log files for the lat three months ]
6.  merge the three log files into one file, in order of date.
    e.g. Jul, Aug, Sep.  [ use cat >> to merge the files ]
7.  copy /mit/techinfodev/src/stats/create_qtr_stats to bitbucket
8.  This file needs to be edited. Look at the two comment lines in
    the file to determine the lines to edit.
9.  I usually run this program from buddy-guy - a DS5000
    log into buddy-guy and cd into /var/ti/util
    ( so it can find the readlog binary )
    get tickets and attach bitbucket.
10. run the program by typing:
    /mit/bitbucket/wade/create_qtr_stats
	/mit/bitbucket/wade/uniq.source > /mit/bitbucket/wade/stats
11. to review the progress, type: tail -f /mit/bitbucket/wade/stats