Collaborating with Family Tree Maker

Others may have a better solution to this issue. I welcome suggestions. Meanwhile, here is my plan.

I have a Family Tree Maker database. It’s inside my copy of Family Tree Maker for Windows 2014. It’s also linked to an online tree at Ancestry.com.

When I get updates from other people, the normal procedure is to re-type the information. Merging information is ALWAYS a bad thing. However, I’m now faced with a different situation: Transcribing our paper and manuscript information into Family Tree Maker.

My intention is that the Family Tree Maker database be the official SFAA Historian research database moving through the 21st Century. Everything else can derive FROM this database.

Here’s what we have at the moment: The entire two-volume “Dwight’s History of the Strong Family” published in 1871 has been transcribed into a single Family Tree Maker database.

My next step is to split that database into separate descendant lines. To be sure, cousins marry cousins, and there are interconnections between the “separate” lines. We simply handle those connection points like we do in the printed books: “For their descendants, go to the (spouse’s) data over there.”

My intention, therefore, is that as we generate books or other materials, we do them based on the separate descendant line. If we want to combine things into one massive database for some reason (such as an online family tree), we can combine everything into one massive database. This massive all-in-one is a DERIVED database, that is, it is read-only.

So how do we get help transcribing things, when we only have one master database?

One possibility is to allow the helper Edit rights to the online tree. That means the transcriber has to work with the online tree. If you’re a fast transcriber (else what is the point?), that’s a major hassle and slow-down. It’s a non-starter for serious transcribing of dozens or hundreds of manuscript pages.

The other possibility is to “break off” a chunk of the tree and hand that chunk to the transcriber. Let me give a concrete example.

Sarah Strong and Joseph Barnard had Joseph Jr., who had Francis, who had Francis Jr., who had George Barnard. All of this took place in the Deerfield-Hadley-Hartford area of Connecticut and Massachusetts.

There were plenty of intermarriages. George Barnard, for example married Sally Higley, who descends from both Sarah Strong’s sister and Sarah Strong’s brother. One is via Sally’s father and one is via Sally’s mother.

Furthermore, Sally’s father and mother are both Higleys. John Higley’s first wife died and John remarried. Sally’s father descends from John Higley and one wife, and Sally’s mother descends from John and the other wife.

This all gets confusing and intertwined, which is why we need something like Family Tree Maker to keep things right!

So, George and Sarah/Sally Barnard had a son George Allen Barnard. He left town and moved to western Pennsylvania. He married a widow after moving to Pennsylvania, Rosannah.

My point here is that George Allen Barnard is a good “break” point. He moved across the country, and therefore we have fewer intermarriages with his left-behind cousins.

Here’s where the collaboration comes in.

In my Family Tree Maker database, I can “detach” George Allen Barnard from his parents. I can also make a note in the “research notes” section of George Allen Barnard, and in the “research notes” sections of his parents (George and Sally Barnard).

It’s just like checking out a real book from a library. There is an empty spot on the shelf where it was, and a library record saying it has been checked out, and by whom.

How do we handle the “check out” from a procedural standpoint?

  1. Make research notes showing the George Allen Barnard and descendants section is being detached.
  2. Detach George Allen Barnard from his parents.
  3. Export George Allen Barnard and descendants to a new file.
  4. This new file is now a new master database in its own right.
  5. Back up everything.
  6. Delete George Allen Barnard and descendants from the original database.

Here’s the problem. George Allen Barnard can never go back. He can’t be re-connected to his parents.

Why not? It’s all of the side connections. Suppose we do research on Rosannah Johnston’s ancestors (George Allen Barnard’s wife). It may be hard to get things connected back up right. Can we safely import 5,000 families? We don’t know!

In other words, once a piece of the database gets split off, IT becomes its own master database. Whoever “owns” it gets to sync to its own online family tree.

What about generating a book that includes both George and Sally Barnard, and their son George Allen Barnard? We’ll need to combine COPIES of the two databases into a single database, from which we generate the book.

So what about returning this database, the same as returning the book to the library? The problem remains that there can only be one master database. Between the two people, whoever is doing the “massive” changes should be the owner, and the “minor” changes can be done by the other person via the online family tree.

What if the “book” gets “checked out” half a dozen times? Each check-out means the other non-master copy gets thrown away and replaced.

Call this the George Allen Barnard database. It’s getting exchanged (“checked out”) between Ed and Jakob. Let’s see how this works.

  1. Ed splits off the database into a separate database.
  2. Ed backs everything up and then (once Jakob confirms successful import) deletes those persons out of the original tree.
  3. Jakob imports the database. This becomes a new database, and “his” database. He creates/syncs a new online Ancestry tree.
  4. Jakob decides to “return” the database to Ed.
  5. Jakob “disconnects” the database from the online tree. The online tree, in effect becomes abandoned and read-only.
  6. Ed imports the returned George Allen Barnard database. This becomes a new database and a new online family tree. I’m assuming there is no limit to the number of online family trees for a given account.

So long as both people have Edit access to the online family tree, both people can do edits to the database. The important point is to understand who “owns” the database at any given moment. Whoever “owns” it, that’s the master database. That’s where we do edits. Any other copies are read-only copies and potentially should just be deleted to remove confusion.

So how does this fit into the book project? It’s likely that we did our transcribing into a bunch of little separate databases. All those “little” databases are each the master database for that piece of the puzzle.

To create the “huge” database from which we build the book, we take READ ONLY copies of all of those databases and pull them together into one huge database. This huge database is itself read-only. It will be thrown away, because it’s not the master database.

How does this fit in with several book transcribers?

Suppose the transcriber is just that. Not a researcher, a transcriber. He or she enters the material ONCE and then turns it in. The piece being transcribed is the “master copy” while that person is working on it. Then, we CAREFULLY import it into the master database upon completion. The transcriber throws away their copy of the transcribed database.

How do we deal with active collaboration?

The best answer MIGHT be to split off a section to become a separate master database. Assuming the two people aren’t actively working on the exact same family line, there’s less in the way of data conflict. Meanwhile, the non-owner (the “other” person of the two) has edit rights to the online tree, so CAN enter information as it comes up.

Basically what I have in mind is one owner of “this” chunk, and one owner of “that” chunk. It “feels” better to have one huge chunk so it’s all tied together, but with collaboration that raises some major technical issues. Smaller chunks without conflicts are likely to be the better way to go.