What's what:

  build.pl   
    Perl script that makes DBM databases out of genea.txt
  child.*, marriage.*, parent.*, person.*
    DBM databases that represent the various interpersonal relations
    needed to draw family trees
  genea.pl
    older, uglier version of genea2.pl
  genea2.pl
    CGI script that extracts info out of DBM databases and displays
    family trees using HTML tables
  go
    shell script for lazy typists: "build.pl genea.txt"

Notes

  genea2.pl is written to work with the DOoM servers, and has not been
  tested with other HTTP servers.  In particular, it checks ARGV[0] to
  see whether the penultimate path component is 'private' to see whether
  or not it is authorised to display information about living persons.
  (This is admittedly ugly.)  It also returns HTTP result codes in
  a manner that should be identical to that of 'no header' CGI scripts,
  though this (as I said) hasn't been tested.

  genea.txt consists of two types of records, person and marriage,
  samples of each of which are given in comments at the head of the
  file.   See individual records for examples of things like non-ASCII
  characters (use HTML entities), reserved words ('alive' as a death
  date for living persons, 'not' as a divorce date for married persons
  known to still be so).  By convention, disjoint trees are allotted
  number spaces in separate blocks of 10,000's: jjc is 1xxxx, European
  royalty is 2xxxx, jbb is 3xxxx, agnes is 4xxxx, dekw is 5xxxx and
  Kristen is 6xxxx.  Please e-mail me to request a block.  Also by
  convention, men have odd numbers and women have even numbers, and
  some scripts may depend on this.  Siblings are listed in order of
  their IDs; this means that they ought to be entered in the same order
  in which they were born.  IDs should in general be entered in 
  increasing order, though they don't have to be consecutive.
 
  I am happy to maintain genealogical data for others, especially if
  you can provide me with clean data files in genea.txt format.