Every personal computer on the market today is supported by “software” to manage a database. Software is the program which the computer uses to control the access to the database, for selecting and/or sorting the information in chosen fields, to design reports in various formats, to store the informtion, and finally, to print the information in the database in the desired format.
In establishing a computerized database, several constraints are in operation and their consideration is paramount in choosing a software program and in designing the format of the record with its fields. A few programs are “memory-resident” and the entire database is active in the computer’s memory at all times. This system is usually much faster at finding and sorting individual records, but a price must be paid. Most personal computers (especially those in the price range the normal hobbyist genealogist will use) have finite memories, and any sizable database will quickly exceed the memory capacity of the computer.
Most database software programs are “disk interactive”; that is to say, the information is stored external to the computer’s memory banks and each individual record is called into memory as needed. The constraint is in the amount of information that can be accommodated by the storage device, usually disk drive which uses either the 5.25-inch, so-called “floppy” diskettes or the newer 3.5-inch disks. The 5.25-inch disks will accommodate about 350K (350,000) bytes of information, while the 3.5-inch disks will accommodate around 800K bytes. Then, there are the “hard disk” drives which have storage limits up to the extent of one’s pocketbook but these also are finite, whether the limit be 10, 20, or 40 million (or more) bytes.
When designing a record, it is necessary to choose a format which will accommodate the desired information but NOT MORE. Remember that each letter, number, symbol, space, etc., will use one byte of information. Once the desired fields have been chosen,  simple arithmetic (the total storage capacity of the device divided by the size [length] of each record) will tell approximately how many records any one storage device will accommodate. It must be remembered that operations on the database such as sorting will also require space on the disk. If such space is not allowed for it could prove impossible to work with the database. Personal experience has shown that at least 10 to 15K should be left free on each disk for these operatiom.
The computer has proved to be an extremely useful tool for the genealogist. With a good word processing program, gone forever are the days of filling out family records by hand or on a typewriter, and computers can easily handle family trees. All that is necessary is to devise your desired format and store it on a disk as a blank form. It can then be called into memory and the blanks filled in and, presto, it is ready for printing. Letters in search of information become a snap and storage on a disk provides for instant recall at any time. Even without a genealogy-specific program, computers can be turned to many tasks in your genealogical research program.
Several good genealogy-specific programs are available for almost every computer on the market today, and most serve their purposes admirably. However, at some time in any genealogical research program, a stone wall is sure to be encountered. A family line has been proved to a certain point and can be carried no further even though it is manifest that other ancestors do exist and need only be found. At this point, the researcher turns to other occurrences of the name in public records and attempts to conned them with his own line. It is at this point that the computerized database becomes most valuable. The genealogy-specific programs are fine for entering known family lines but offer little help when unidentified ancestors must be considered.
Such an impasse was reached early in my own genealogical research. I had traced my family line to Tennessee in the early 1800s where, unlike the New England area, records are few and extremely difficult to find. No city or county vital records such as are found in the early New England colonies, and very few pertinent family histories are available. Being blessed with a somewhat unique name, I felt that any bearer of the name might be a relative and I began my search, scouring all records that could be found in any part of the country. Censuses, court records, and a few marriage records yielded a large number of references to the name. Correspondence with other persons interested in the name yielded more information. All of this soon became too voluminous to manage and I turned to the computerized database, reasoning that if the surname database maintained by Everton Publishers was useful to so many researchers, the same approach might be helpful to me.
Mine is a one-name database -- BRYMER -- in all its varied spellings. I devised my database record to include fields for name, event, date, location, and a unique identifying (ID) number, and provided for 80 spaces to enter remarks as to where the information was found and any amplifying information that can be included within the space limitation All of these fields yielded a record length of 180 bytes which allows for about 1500 records on each 5.25-inch disk and over 4000 records on each 3.5-inch disk with room left on the disk for manipulating the records. When the disk capacity has been reached a second disk will be started, which will introduce some problems, since it will no longer be possible to work with the entire database as an entity.
For all of the above steps, I used the Commodore C128 system with the Superbase 128 database program from Precision Software of England, a system which is usually scorned by those using an MS-DOS (IBM standard) system. I am not trying to sell Commodore computers and I will probably graduate (with some regrets) to an MS-DOS system in the future. I merely wish to point out that an adequate genealogy research program, including word processor and database program, can be implemented using any computer.
As my particular database grew to respectable size it began to yield some impressive results. By manipulating the database it was possible to arrange events in a chronological order which resulted in a vivid picture of the migration of the name. Further manipulation by time and location showed, almost at a glance, relations between the various people in the database. In this regard, bear in mind that almost any census listing from 1850 on will result in a single family listing. Succeeding or preceding census listings may show further family lines. Identification of these is greatly facilitated by the use of the computerized database.
At this point, the unique ID number included as a field in the database comes into play and is used to identify known family lines. In my case, I have devised my own numbering system which begins with the earliest identified progenitor of a line; a subsequent combination of numbers and letters identifies each specific individual of each subsequent generation. To illustrate, one progenitor has arbitrarily been designated "w1” for no other reason than that he was the first identified member of a line and his name happened to be William. His first child was numbered “w1a” and so forth. Other family lines are designated in a similar manner. As an example, the ID number “w1c3d2a4” tells the observer that this individual is the fourth child of the first child of the second child of the fourth child of the third child of the third child of w1, William. Seven generations are identified at one glance. To my mind, this combination of numbers and letters prevents confusion between the generations and provides the most efficient use of computer memory in the database. Twelve spaces should be sufficient for most users, to provide for about 11 generations.
As a matter of general interest, my one-name database has now resulted in the identification of eight generations of descendants from my William1,  naming some 350 individuals, and the list is still growing -- all as a result of trying to prove my descent from my great-great-great-grandfather. A simple database search based on the ID number readily reveals all known descendants, while the absence of such a number indicates that more work is to be done.
I cannot recommend too strongly the use of a database in genealogical research, although I recognize that a system such as mine works best with a rather unique surname. Readers can devise their own database programs to help in their research, and I will be happy to answer any questiom. Any additions to the BRYMER database will be gratefully accepted.
Editor’s Note: This article was inspired by Virginia Wood’s “The Family of Rev. Richard Mather: A Genealogical Database” in NEXUS 5(February 1988)16-17. Address any inquiries to Col. Raymond C. Brymer, USAF (Ret.), P. O. Box 2868, Harbor, OR 97415-0503. Col. Brymer’s numbering system is similar to time Henry system often used in early 19th-century genealogies, which tells birth order and distance from the immigrant but can become unwieldy if time line is carried 13 generations or more. For an example of the Register-style nunmbering system, used also in most NEHGS-published genealogies, see time “The Benchley-Bensley Family,” by Roger D. Joslyn in the Register 142(1 988):3 -1 5; 177-95; 281-97, i71-85, to be continued.
If you have conmputerizcd your genealogical records and are willing to be a clearinghouse of information for other NEXUS readers, let us know. As usual, we are also interested in the various components of your database management system and your evaluation of its effectiveness.