Why PAWriter Was Developed
3 May 2006

Many years ago, the limitations of the old Mac PAF drove me (the author of PAWriter) to create a similar program without some of those limitations, and to rescue years of work already committed to PAF so it could be processed in today's MacOS. Here's a few of the objectives I wanted to and did attain.

The number of people on file and the size of notes for any person had to be virtually unlimited. The length of person and place name pieces had to be doubled. Some additional GEDCOM tags had to be accepted. Book writing (including automatic generation of an index) and HTML capabilities had to be added. PAWriter had to run equivalently under MacOS 9 and MacOS X.

For writing books, the notes wouuld have to carry the load of most of the book text. They would contain all of the evidence, substantiation, documentation, background information, clarification, interpretation and/or other relevant commentary pertinant to a person on file.

On the other hand, the entries for a person's vital statistics (birth, marriage, death, etc.) would represent the researcher's best current conclusions as to the true facts, given the information in the notes.

For the notes to fulfill this purpose, footnotes within the notes were required, so that any statement in the notes could be sourced by a footnote following that statement, without any limitations, as one typically finds in genealogical books. Also, boldfacing, italicizing and a few other features within the notes and the footnotes needed to be added.

These features within the notes were accomplished with a simple markup language. The markups allowed PAWriter to generate equivalent HTML, RTF and MML files. (MML was required for publishing using FrameMaker, at the time and still the best book production system, although now unsupported.) Citations of source documents were centralized so they could be easily referenced from within the notes in multiple formats (short form and long form in footnotes and long form in bibilographies). Standardized entry of census citations was added.

Along the way, I added additional features to PAWriter to assist in my research: basic pedigree charts, family group sheets, possible problem reports, self-checking of file consistency, etc. Ultimately the inclusion of pictures was added so that images of census pages, other source documents and family memorabilia could be displayed.

To navigate through very large numbers of people on file, a powerful "tagging" system was included. It provides flexible selection of subsets of people in a file by marking them as members of the subset using a host of selection criteria. Selections include inclusion, exclusion and intersection of the set of people meeting the criteria with those already tagged.

Some things PAWriter is not meant to do: Anything more than minimal LDS ordinances (defined as LDS baptisms, endowments, sealings to parents and sealings of spouses containing only status, date, temple code and place if living); complete GEDCOM compatibility as defined in the version 5.5 standand (which is handled nicely by John Nairn's GEDitCOM); and fancy charting and other bells and whistles that are included in other genealogical programs (no disrespect intended) such as Reunion. The emphasis is on secular research (including intercommunication with other researchers) and publishing. Also note that PAWriter uses the English calendar (pertinent for the Julian to Gregorian transition in 1752).

In my opinion, PAWriter has accomplished these goals as witnessed by the publiication so far of three large tomes, the last by Heritage Books, Inc., wherein the entire register portions of the books (most of the contents of the books) was produced by PAWriter. On my 1.4GHz G4, a 3,000-page register in final format can be generated from my current PAWriter data file for FrameMaker, and converted from there to PDF and burned on CD, in less than an hour, thereby facilitating any number of last-minute changes in the copy sent (and re-sent) to the publisher.

One more practical point as a researcher: Over the years I have generally found that importing other people's GEDCOM files into my existing data file to be impractical. I've yet to find a GEDCOM that doesn't require a large amount of work "cleaning" it and verifying each item within it. I've determined that the best that can be done in most cases is to print it out and manually enter each piece of information into my data file after the information is cleaned and verified.

This is not to say those GEDCOM's aren't useful. They just require a lot of hard work before the information can be used (frequently as just clues) in your own research file.

So to some extent I question how useful features are which import and automatically merge other people's GEDCOM files, even with the interactive comparision of old and new information that some programs provide. Thus I haven't expended the necessary (large) effort to be 100% compatible with the version 5.5 standard for GEDCOM tags in my PAWriter import.