Development work should be kept separate from "live" data, using the Data Separation Model ([i]).
The Anchor Buoy method
Carsten Levin 16-Jun-2015 15:20 - From 7.0 and onwards this has been recommended by many of most serious developers. And FileMaker did specifically recommend rewriting FileMaker 6 solutions to take advantage of the separation model.
Let me quote: Create a New Solution (Entirely Rewrite) Starting with new files is most appropriate for a new business problem or an isolated information domain. It could be appropriate if you want to take full advantage of the new FileMaker Pro 7 possibilities and you can afford to wait to complete your redesigned solution before deploying. It is also appropriate if the features of the original solution are obsolete or poorly designed in the first place. The foremost requirement when designing a new solution is to determine your solution architecture. The articles on The FileMaker 7 Relational Model and The Separation Model: A FileMaker Pro 7 Development Model are both essential reading.
The Separation Model: A FileMaker Pro 7 Development Model – Colleen Hammersley & Wendy King
FileMaker Pro 7 is a dream come true for those who have long advocated the separation of data and interface. Colleen and Wendy present the new application model in all its glory.
http://www.filemaker.com/downloads/pdf/techbrief_fm7_foundations.pdf
Just for those who may need to know a little bit more about FileMaker and separation: It is a long discussion. The main school is to have all presentation and functionality (programming etc.) in one or more interface files with the main relational diagrammes. And then one or more data files with the tables and only very few relations supporting basic calculation fields. Some are going all the way and are even avoiding calculation fields and their relations in the data file, we are not that rigid at my company.
https://www.geistinteractive.com/2014/01/16/filemaker-separation-model-splitting-file/
Here from tidbits 2004 about the paradigm shift: https://tidbits.com/article/7587
The Separation Model -- When one file references another file, it references everything in that file. In other words, if one file, called My_Data.fp7 has 22 tables in it, another file called My_App.fp7 can access every one of those tables as if they were stored internally rather than externally. What does "reference" mean here? It means that My_App.fp7 can contain layouts that provide form and list views of each of the tables in My_Data.fp7. Not list as in "portal" - a limited interface element familiar to current FileMaker users - but list as in a real list, a list that can print nicely on page after page, a list that can be displayed with subsummary totals. My_App.fp7 can do this even if it does not contain a single table itself. It might contain nothing but layouts and scripts, all taking advantage of the tables in My_Data.fp7
And here are FileMaker's own suggestion regarding separation from the first version of FileMaker 7.0v1. http://www.filemaker.com/downloads/pdf/techbrief_fm7_foundations.pdf
The Separation Model: A FileMaker Pro 7 Development Model – Colleen Hammersley & Wendy King
FileMaker Pro 7 is a dream come true for those who have long advocated the separation of data and interface. Colleen and Wendy present the new application model in all its glory.
Start having a look at Todd Geist FileMaker Separation Model: Splitting Files - YouTube
Or Dwayne Wright (See: http://www.dwaynewright.com/filemaker-thoughts/2011/5/24/why-try-the-filemaker-separation-model.html)
Or FileMakerPapers.com - Data Separation: Ideal and easy with FileMaker by Pam Rotella
Or data modeling by Ray Cologon http://www.nightwingenterprises.com/Resources/filemaker_solution_architectures_en.pdf
http://www.filemaker.com/help/14/fmp/en/html/create_db.8.2.html#1126267
http://filemakerhacks.com/2015/07/06/fm-14-separation-model-data-mining/
http://filemakerhacks.com/2015/07/06/fm-14-separation-model-data-mining/
