We quickly recognized, after becoming interested in genealogy, the need to organize the family information we had begun to collect. We had owned and used a personal computer for some time, so it seemed logical to seek software designed specifically for genealogy. Our investigations indicated that the Mormon PERSONAL ANCESTOR FILE package was a competent and (at a delivered price of $35.00) very economical starting point. About two and a half years ago we bought version 2.0, and upgraded it for $25.00 to version 2.1 in mid-1988.
We have used PAF extensively to record ancestors and relatives, to exchange information with other hobbyists and “cousins,” and to document our research. We regard the package as an excellent, honest set of tools — easy to learn and use with or without previous computer experience.
Despite our firm approval of PAF, we have encountered some limitations and frustrations while handling actual genealogical situations. We share these experiences, not to criticize the LDS software, but to advocate continued improvement.
Vital Features: Data Transfer, Storage Capacity, Resource Use
Before buying any genealogical software package, realize that it will contain only the information you type or transfer into it. PAF, the best-selling genealogical package, accepts transfer of information from other PAF users and any program with the capability to create a GEDCOM file. There are many thousands of such users, but if your particular collaborative source has information stored in some other software, you must create a GEDCOM file to be used by PAF. Most competing suppliers recognize PAF’s predominance and provide conversion capabilities from PAF information to their products.
The software and its database(s) must have a large storage capacity. While many computer users have floppy disk systems, others have far more spacious hard disk drives. We never expected in our early research to find a large number of ancestors; in fact we long ago exceeded the 4,096 person ceiling of the Roots II package (significantly increased in Roots III), and are nearing the 9,999 limit of Genea-Link, but are unlikely to approach PAF’s 65,000 person limit.
While such family crowds are well beyond the capacities of floppy disk systems (our best estimate is that such a system can handle information for 1000-1200 people), they can be easily managed if your system has a hard disk. To be fairly safe, estimate the likely size of your genealogical population, then increase your best guess by a factor of three. This number should allow for the presently-unknown relatives and ancestors who will inevitably appear.
Similarly, any program should fully exploit your computer’s memory. PAF requires about 338K of memory to run its largest program. If the user has more memory (few systems are now sold with less than 512K) PAF ignores the additional horsepower. Some problems we mention could be minimized or eliminated if all available resources were applied.
From the earliest days of the IBM Personal Computer, PCs have had at least 10 function keys allowing frequent operations to be executed with one keystroke. Most software packages use all ten; many use them with the Shift, Control or Alternate keys to simplify 40 different operations. The LDS staff has so far used only four function keys in PAF. Thus you have to learn other ways (which do exist) of performing repeated procedures.
The latest version of PAF provides increased capacity (roughly 32 full screens or 15 typed pages) for text information about a person. Most of us have at least a few ancestors whose deeds or misdeeds require greater space. We can shift to our word processor and store such text there. But doing so isolates information and prevents direct retrieval, updating, analysis, or automatic inclusion in reports printed by the PAP program.
PAFUTIL promises to print a full cascading pedigree (all ancestors of one person, printed 16 people per page), and an index for over 10,000 ancestors. Although we jointly haven’t reached that number, neither of us has been able for some months to produce complete pedigrees. The actual limit now is 750 pages rather than a specific number of people. (The LDS staff indicates this limit will be increased.)
The cascading pedigree index is limited when confronted by multiple marriages or complex family patterm. We are both descended, for instance, from five daughters of a single 13th-century couple. The index notes only the first page where a person appears, requiring a tedious search for every subsequent citation. A sound pedigree program should cope with and index whatever ancestral patterns it encounters, no matter how intricate.
Convenience Features: Entering/Editing Text
One of PAFUTIL’s strengths is its on-screen pedigree and accompanying editing capabilities. Using this perspective, you can see and revise three generations at once. We find it most valuable to view generational relationships when making updates or corrections. Lacking, however, is the ability to add or delete persons or generations without kangaroo-hopping to other sections of the program.
Single fields can be duplicated (surname, town, etc.) from one record to the next. With many families and their children born, marrying, dying, or buried in the same town, a broader duplication feature, turned on when appropriate, would speed input. Also available is a (very awkward) feature that repeats previously-typed names. By the time you’ve hunted through your last name listing, you ignore this feature and just retype Nottinghamshire, Lake Memphremagog, etc.
PAF creates, maintains and checks every name you type against a list of those you have used earlier. Any name you’ve typed identically twice in succession is added to the list, and the program beeps whenever confronted by a new name. This beep is valuable in encouraging consistency; there is, however, no way to remove names that are misspelled (yes, we do make mistakes!) or superfluous. Once the wrong name has been used, there is no easy way to find and replace it consistently through the whole database. The sound objective of helping you avoid errors eventually becomes a barrier to final correction.
PAF contains a field to indicate divorce, but no provision for its date, leaving broken marriages only partially defined. No facility exists for such other realities as adoption or illegitimacy.
PAF currently does not allow dates prior to 100 A.D. Conjectural ancestors before that date must be omitted, and no provision exists to record or print accents and other symbols integral to many non-English names.
PAF’s logic flow makes you answer some inefficient, annoying questions. For instance, we have text notes for about 21 percent of our ancestral records. Yet we must tell the program that we don’t want to create them for each of the other 79 percent. More flexible options or better use of function keys would make the activity more convenient and less irksome.
Aids for Further Research
Unless other genealogists are far wiser or more intuitive than we, they record during research many possible ancestors or relatives. Over time, some “possibilities” are confirmed, many others disproved. It should be convenient to identify and remove deadwood that otherwise clutters the database. There should also be a simple way to tag conjectural, unproved records as worthy of further research.
PAFUTIL lists persons without parents, but it has no provision to list those without children — one way of spotting “possibles.” Such a facility would be even more versatile if it focussed on a specific locale or time period.
Sorting and Manipulating Data
PAF incorporates several built-in sorting function. The user has no way of changing them, and they are slow when working through a data population of our size.
There is no shortcut way to look at a person or a marriage record; you tend to use the same keystrokes that you used to create or correct it. PAP then thinks that changes have been made and repeats its sorting. If your printer jams while doing a sorted list, both the sort and the printing must be repeated. While we appreciate the convenience of sorted lists, we grow frustrated by their slowness and repetition.
Our research has suggested some intriguing patterns (consistent marriage ages, frequency of births, etc.). It would be most interesting to be able to define (say) a geographic area or a time period and to search for people falling within that definition. PAFUTIL at present does nothing beyond a bar graph of all recorded birth dates.
PAFUTIL has a useful date calculator. If you know, for instance, a death date and a person’s age, the calculator quickly figures a birthdate. It is not comfortable switching between old and new calendars. This flexibility would enhance the program.
PAFUTIL will also calculate relationships; in fact it undertakes to list all possible blood kinships between two people. The process, however, is very slow; it takes many minutes to determine our relationship to our parents. When the relationship calculator finally decides that you are in three cases someone’s sixth cousin twice removed, it does not identify common ancestors. The concept is valuable, but there is room for improvement.
PAF documentation is clear and informative in all the areas it covers. Understandable training situations and examples are included, with well-executed illustrations and explanation. Lacking, however, are any strategic suggestions on how most easily to tackle the job of putting existing information into PAP. Another serious omission is any description of FRCHK, an included program that examines all your data and reports any computer errors. When you’ve spent the time recording your research, the last thing you need is to lose some or all of it due to a database error.
Purchase and use of a product warrants ongoing support from the supplier. For PAF’s modest cost the Ancestral File Operations Unit in Salt Lake City provides exceptional support, answering letters, responding to phoned inquiries, and sending software corrections or improvements. You must, however, return the registration card included with your software and take the initiative in requesting support.
These pages cover our own concerns, wishes and suggestions. Another user could produce a different, equally valid list. Our aim is to stimulate discussion and to encourage development of even better software than exists today.
Editor’s Note: Improvements in software packages usually mean a higher price. Correspondence is welcome on which of these suggestions by the Wattses, or other added features, PAF users would most like. Readers may contact the Wattses at 18 Byard Lane, Westborough, MA 01581.