At 8/15/2015 14:18, you wrote: >The project vs dataset question reminds me of a question of my own. > >My main project (which has several data sets) has my own family. However, >over the years I've also created a project for one of my friends and a >project that contains many of my friends, with their birthdays and >miscellaneous data. > >I've kept the friend projects separate because I've never intended to share >them with family, whereas I occasionally do send my brother a backup file >of the family project. > >Can I add the friend projects to my family project (simplifying sourcing) >but keep them out of backups of the family dataset(s)? Yes, you can merge project A with project B (or vice versa). This means that the merged project will have two data sets (or more depending on how many data sets were in the original projects). However, I don't believe that you understand that each data set is _completely_ separate from all other data sets in the project. For example, Source ID# 1 in data set 1 will not be the same as Source ID#1 in data set 2. Now Source ID#1:1 may happen to be a letter dated 2/23/1956 to Mary Smith and Source ID#2:1 may also happen to be the same --- but they are not the same Source insofar as TMG is concerned. Further, Source ID# 2:3 cannot be used in a Source Citation for an Tag in data set 1. Think of a project as a file cabinet with each drawer being a data set. You can insert a Family Group Sheet (FGS) into drawer 1, but it can't be placed in drawer 2 at the same time. Of course, you can photocopy that FGS and place the copy into drawer 2, but they are separate pieces of paper. In the same was for TMG, the same data may be in two or more data sets, but all are different pieces of data even if they happen to read the same. The advantage of having multiple data sets is the convenience of viewing them together if desired and the ability to copy or move data from one data set to another. I will often keep client data sets with my main data sets for the convenience that affords. But, there are times when I will move the client data set to a separate project. This is usually in preparation for extensive work with that data set and maybe others, preparation for sharing with another researcher/the client, etc. In such cases, the original data set is deleted from the original project and will likely never be seen there again as this is one of the last operations with that client's data. Lee Hoffman/KY