Which should be the primary key?

[Travis]

I have been importing records of organizations and people that is aimed to creating a contacts database. I have several tables with different kind of contacts (such as churches, schools, businesses, etc.) Some of the contacts are actual people with out an organization attached to them, and some of the contacts are just the organization without contact names attached to them. Some have addresses, phone numbers, & emails. Some contacts have all or no addresses, phone numbers, or emails attached. I can't find any field to make the primary key so I can tie up the relationships between the tables. Should I create a "contact number" to identify, if so, how do I do that now that all of the records are already imported to the database.

[ESpringer]

You should certainly set up an ID that is not directly tied to any of the data. THis is a good idea anyway, since any given piece of data (name, phone, location) may change while the identity of the entity remains. Your contactID numbers don't have to "mean anything" or even be visible to end user.

Search in Define Fields forum for ideas about record IDs, Unique ID, serial numbers, etc. There are all sorts of tips!

It doesn't matter that you already have your data; you'll want to make a new field and run a script to loop through and populate it with your ID sequence, however you choose to construct it.

By the way, why do you have several tables of contacts? How are they distinguished? Why not one table where the different kinds of individuals/institutions are tagged as such in some field? Note: you can have one contact "point to" other contacts as "belonging to" it using a join table. (The advantage of one main table is that when you make a layout (such as an address labels layout, say), it becomes useful across the board for any batch of contacts.

You might want to think carefully through your table organization before carving too much in stone. (A lesson not entirely learned by yours truly.)

[Travis]

Thanks for the response.

I have seperate tables because as I import tables I have different fields for the data and want to enter data as the records are imported. It seems easy for me to do it like that. But, will I have to put them all into the same table to find similar records (such as location), to create mailing lists? I also need to know how to put new ID numbers in to the records that I already have.

[dkemme]

Once you get the data in FMP, you would be best organizing it a bit more. I like to put all phone numbers in one table, all addresses in another and so on... This allows you to have differing numbers of phone/addresses for each person. Some contacts have none, some have many. When creating the tables for phone/addresses, create a field for the contactID that the phone/address belongs to.