1 Reply Latest reply on Dec 16, 2010 9:49 AM by philmodjunk

    Keeping the database separate to the layouts



      Keeping the database separate to the layouts


      I'm pretty new to Filemaker, but I'm about to trial my first database at the local volunteer rescue station.

      As I prefer to work on the database layout at home on a development version of the database, is it possible to separate the layouts from the data? 

      As the station runs 24/7 I was hoping to make changes at home or in a development version, then copy them across to the live version without losing any data or records.

      I've searched the web but can't find an answer, so hopefully someone can help - thanks!

        • 1. Re: Keeping the database separate to the layouts

          Yes, this is called the "data separation model" and it has been discussed frequently here in this forum and others.

          Here's how to take a single file solution and split it into a data "back end" and an interface "front end":

          Make a copy of your file so that you have two identical copies in the same folder. (Same folder permits a relative path file reference from the interface file to the data file and this is the most trouble free way to set this up.)

          Open the Inteface copy and select Manage | Database | Relationships.

          Double click each table occurrence box, one at a time and use the Add FileMaker Data source option to select the other copy of the file and the matching table within it.

          Once you've updated every single table occurrence to link to the matching table in the other file, go to the Tables tab and delete them all from the interface file.

          You can stop at this point if you want or go further. You may want to retain layouts and scripts in the data file, at least at first so that you can still check some things out.

          If you want, at this point you can analyze the relationship graphs in each file and edit them so that relationships that are only needed at the data level (so that calculation fields that combine data from other related records evealuate) are kept in the data file and relationships only needed at the interface file are kept there, but this is not necessary, takes a bit of time and a mistake here could break your system fairly dramatically.