Table Occurences,TOG's and data seperation

[cinolas]

Greetings,

I am a long time user of Filemaker (4, 5, 6) and I have recently started to work in FM7. The changes in relational concepts, namely the TOs and TOGs, have me all turned around. A link to a good resource to get a solid understanding of how to properly organise TO's in TOGs would be appreciated.

I am currently working on a system in which the data would be seperated from the interface. In order to get familiar with the concept I examined the New Century Data seperation model (download it here) and I now have a couple of questions:

On the NewCenturyData data seperation model exemple, the TO's are linked in the data file in a way that resembles most the logical hierarchy of the data as opposed to the way they are linked in the gui file. However, the relationships established in the gui file are all that is needed to make the database work (the gui is functional even if you delete the relationships in the data file).

- Why would one organise/link the TO's in the data file if the interface file is the one actually specifying the paths that the data will follow ?

- It seems to me that all the data tables can be specified in the data file wihtout being related and that they can be related and organised on the relationship diagram of the interface file. Is a relationship diagram even necessary at all in the data file unless it is needed by scripts/reports/web publishing ?

Any thoughts on this would be greatly appreciated.

Please also point me to resources that may help me get a solid understanding of TO and TOG organisation and the proper way to do it.

Thanks

[Kurt Knippel]

- Why would one organise/link the TO's in the data file if the interface file is the one actually specifying the paths that the data will follow ?

For lookups and calculations which are a part of the data model.

- It seems to me that all the data tables can be specified in the data file wihtout being related and that they can be related and organised on the relationship diagram of the interface file. Is a relationship diagram even necessary at all in the data file unless it is needed by scripts/reports/web publishing ?

The only relationships in the data file are those needed for lookups/auto-entered calcs and calculation fields that are actually part of the data model. Many people will also build the complete data ERD in the data file, even if it is replicated in the interface file, just for refernces purposes.

[cinolas]

But of course !

Thank you very much ! Truly appreciate your help.