To Merge or Not. Or What!??
I remember 1986, the year we got our first IBM compatible computer. At last I could use Personal Ancestral File software. I spent every spare minute for three years hand-entering data on my ancestors, finally emerging with a database of over 5,000 names!
About that same time I decided to write a family history about my great-grandfather. The first thing I did was to make a GEDCOM (transfer file) of Great-Grandpa’s families and put it into a new computer program called RootsIII. Then over the next three years I collected Great-Grandpa’s descendants–more than 3,000–and added those to RootsIII also. Well, the book was wonderful; and the family loved it; but now I had two databases. What I really wanted was all my family in one file, so I could share it with my kin and submit it to Ancestral File.
Attempting to combine these two databases has been a real adventure. I went back to my friend GEDCOM and transferred all those descendants from RootsIII back to PAF. All went well except for one tiny problem–the notes. My beautiful documentation did not survive the transfer (but that’s another project). I have always considered myself fairly adept with computers, that is until I met my match. . .er, ah, merge.
My first mistake was trying to fix my place-name abbreviations using Don Snow’s WordPerfect search and replace technique. Just load your GEDCOM file into WordPerfect as a DOS file and replace all the two-letter state abbreviations with the full state name. I found it takes a fair amount of time to do a replace in a GEDCOM of 4,000 plus records, however. Add to that the fact that a separate pass through the file is needed for each change, and you can count on at least a couple of hours. If you have relatives in several states you’ll see what I mean. You might want to go get a snack or keep something handy to read while you are waiting for WordPerfect to do its work. (In retrospect it’s probably wise to work with smaller files.)
Finished at last. Now to put this corrected file back into PAF. All went well at first, then PAF stubbornly stopped on record 132. Back to take a peek at GEDCOM. Don’t see anything wrong. Back to PAF. Stopped again on record 132. Maybe I should go back and take a look at my notes from Don Snow’s class…. Of course! Make sure you set your margins to zero before loading your GEDCOM into WordPerfect. Back to GEDCOM. Sure enough, record 132 had a long note which had wrapped.
Once the margins were fixed, everything went smoothly.
I was feeling so successful at this point I thought, “While I’m at it, why not merge this file with the one a distant cousin gave me last summer?” He had downloaded everything there was on this line from Ancestral File and combined it with the data he had from his branch of the family. Although this would mean another evening of searching and replacing in his file, which was nearly as large as mine, it would be worth a little extra time, I thought, to have everything I know about this family in one place.
Now the fun begins–matching and merging those 8,000 records. It wasn’t long until I realized that my searching and replacing was coming back to haunt me. I began to see some rather unusual names in my data base, names like TALABAMACOTT and MCBRIDAHOE. Hmmm….
Of course, why didn’t I think of that? Replacing AL with ALABAMA replaces AL inside names too. I should have replaced AL space with ALABAMA. When WordPerfect asked me if I wanted to confirm all those changes, I had a feeling I should answer ‘Yes,’ but I didn’t. Do you know how long it would have taken to press ‘Y’ 8,000 times? Now what do I do? I already have considerable time invested in this project. I can’t stop now.
With renewed determination I set about changing names–GREGOREGONY to Gregory, WILLINOISLETT TO Willett, BRIDAHOGUM to Bridgum, ALABAMAEXANDER to Alexander, and CROOKLAHOMAER to Crooker. What a mess! I’m not a quitter, though. Week after week I dedicated Sunday afternoons to correcting names and merging records. You get a big family like the BENEDICONNECTICUTs extending over several generations, and the maze never seems to end. The fact that every family had 10 or 12 children didn’t make it any easier either. I was sure I had all of them corrected, yet BENEDICONNECTICUTs kept popping up over and over and over again.
Maybe I should forget the BENEDICONNECTICUTs and concentrate on merging. That might eliminate some of them. I had been told that Ancestral File numbers were a pretty safe match; but much to my dismay, I found that even those numbers are not sacred. They change from one edition of Ancestral File to another! Okay, how about merging individuals who have spouses with the same RIN? That worked. It also helped to look at the parents or the children of the principal, but most of the time there simply wasn’t enough information to make an intelligent decision.
I thought I’d be able to tell which records came from my cousin’s file by looking at the RIN numbers. It seemed reasonable to assume that if I had 4,000 records and he had 4,000 records, and I added his file to mine, RINs above 4,000 would be his and those less than 4,000 would be mine. Not so. When adding records to a file, PAF uses up all the deleted records first. I must have had a bunch of them, because the end the result was the same as if I had dumped all the records from both files in a bushel basket and stirred them up.
With each new endeavor, I seemed to be getting in deeper and deeper. Would I ever be able to clean this database up to the point where it would be “worthy of all acceptation?” The prospects looked pretty dim. Maybe a good night’s sleep would help.
I hated to admit it, but I was licked. I had to give up. There is more to merging than pressing ‘Y.’
Thank goodness for tomorrow–and backups!