New branches outside the scope of research

[elevator 0]

Good evening,

I have a question for you masters out there: I have regular contact with a few family members that I am VERY distantly related to. We're talking 7 and 8 cousins here. I am very hesitant to include these branches in my regular family tree because I fear they will only clutter things up. Is there a way to add these branches into the family tree but still maintain the integrity of the main database?

 

I was thinking something like adding a new dataset for each of these family branches. If I do that, is it then possible to hide that dataset or delete it entirely in the future should I decide I do not need it anymore? What will happen to the parent/child links if for example the parent is in one dataset and the child is in another?

 

Just wondering if someone else have done this already and if so how they did it.

 

Thanks,

Ken V. Nordberg

[Terry Reigel 0]

I am very hesitant to include these branches in my regular family tree because I fear they will only clutter things up.

I'm hard pressed to understand how adding a few dozen people will "clutter up" the main data set.

Is there a way to add these branches into the family tree but still maintain the integrity of the main database?

Yes, you could place them in a separate data set in your main project, or in a separate project. That would keep them entirely separate.

I was thinking something like adding a new dataset for each of these family branches. If I do that, is it then possible to hide that dataset or delete it entirely in the future should I decide I do not need it anymore?

Yes, you can disable one or more data sets, which effectively hides them.

What will happen to the parent/child links if for example the parent is in one dataset and the child is in another?

There won't be any - the whole idea of separate data sets is to keep the data entirely separate. So you can't link people in one data set to those in another.

Just wondering if someone else have done this already and if so how they did it.

While you could put these lines in a separate data set, I'd strongly urge against it. I see little advantage, and many disadvantages. People in different data sets cannot be linked. Things like sources, repositories, customized tag types, and customized source types all have to be duplicated, and there is no easy way to copy them between data sets. I put everyone in a single data set. I'd only consider separating them if your current project is approaching the limits of your hardware (a consideration at around 50,000 people), and then you need to use separate projects, not datasets, which is even more of a bother. For more on projects and data sets, see my article on that topic on my website - link below.

[elevator 0]

Thanks for your reply Terry. Some of these lines represent the research from other family members on a particular line and may in some cases contain thousands of individuals, but as my own family tree only contains around 5000 people, maybe you're right, that it wont be too much of a problem. Thanks for clarifying the downsides about datasets. I didn't know that you couldn't link between datasets, which makes them really unattractive. I guess you learn something new every day! I was really looking for a way to link branches to the main tree without having to include it in the main dataset, but after reading your help I realize it may be impossible without including the data in the main tree.

 

Thanks Terry.

 

Ken V. Nordberg.

 

 

I'm hard pressed to understand how adding a few dozen people will "clutter up" the main data set. 

 

Yes, you could place them in a separate data set in your main project, or in a separate project. That would keep them entirely separate.

 

Yes, you can disable one or more data sets, which effectively hides them.

 

There won't be any - the whole idea of separate data sets is to keep the data entirely separate. So you can't link people in one data set to those in another.

 

While you could put these lines in a separate data set, I'd strongly urge against it. I see little advantage, and many disadvantages. People in different data sets cannot be linked. Things like sources, repositories, customized tag types, and customized source types all have to be duplicated, and there is no easy way to copy them between data sets. I put everyone in a single data set. I'd only consider separating them if your current project is approaching the limits of your hardware (a consideration at around 50,000 people), and then you need to use separate projects, not datasets, which is even more of a bother. For more on projects and data sets, see my article on that topic on my website - link below.

<{POST_SNAPBACK}>

[elevator 0]

Just as a side note, another reason for keeping the branches outside of the main dataset is of course that the research is done by someone else and the sources need to be verified. I would be really hesistant about adding unproven information to the main tree. Of course I guess I could just make a flag indicating that this particular individual needs their sources verified....

 

Ken.

[Terry Reigel 0]

Ken,

 

I agree that the source issue is an important one, but not one, in my mind, that requires keeping the data separate.

 

You could record the sources as provided to you and mark the individuals with a flag, but that would not be my recommendation. One issue with this approach is that if you should include some of these people in a report you create, it would appear that you had examined the sources and agree with the interpretation shown. I don't think that's what you would want.

 

There are at least a couple of alternatives. In my view, the "purest" is to cite only the file or other media which you received as the source. Then, in the CD for that citation, include a note showing that your source cited whatever you received. I generally do this by putting something like "..., citing Pawlet town records, vol 1" or whatever the source you are given says.

 

There are a couple of problems with this method. One is that it's a lot of work to do, especially if you are importing a well-sourced file from another researcher. The other is that your source notes can get pretty long and repetitious, especially if the descriptions of your received sources is detailed - you can't very well get a "full footnote" for you indirect source the first time and short one thereafter.

 

There is another approach that I'd think would work especially well if you are importing a file with sources attached. That is to modify each of those sources by adding some sort of disclaimer... maybe something like "cited my Cousin Bob, not seen by author." You could do this after the import, but before merging the imported data with your own. You could put the disclaimer in the Comments field on the Source Definition, and make sure the [COMMENTS] source element is included in each source type you use.

[Matt_G 0]

For more on projects and data sets, see my article on that topic on my website - link below.

<{POST_SNAPBACK}>

Off topic, but I just wanted to post and says thanks to Terry for putting together such a great web site. I'm a newbie to TMG and it's been a tremendous help to me.

Thanks Terry!

[GenerationGoneBy 0]

I throw them all in there. I figure you might need that research to solve another puzzle at some point in time. But then I have 24,000+ so a few thousand more ain't that big a deal.

[Karla Huebner 0]

I agree you could do a flag... I know Terry didn't think it was such a great idea (and I think his ideas are good) but with a flag, you could then easily exclude any of these people from your output, say for instance WorldConnect or Second Site. So you might want to combine use of a flag and Terry's suggestions.

 

Karla

 

Just as a side note, another reason for keeping the branches outside of the main dataset is of course that the research is done by someone else and the sources need to be verified. I would be really hesistant about adding unproven information to the main tree. Of course I guess I could just make a flag indicating that this particular individual needs their sources verified....

 

Ken.

<{POST_SNAPBACK}>