Table Occurrences - a check on my understanding
I have inherited a FMP 11 db and I am trying to get my head around it. So we have established that in the relationship diagram we see the base tables and the table occurrences relating to those base tables.
There are about 100 base tables and many table occurrences. Some base tables have a single occurrence i.e. the base table only others have many occurrences.
A check on my understanding:
- All relationships should resolve to one to many (many to one).
- A one to one should effectively be a field and a many to many to be broken down to two one to many relationships.
- A parent is on the 'one' side of the relationship and a child is on the 'many' side of the relationship.
- If a relationship between a parent and child is created then:
- We can access a parent field and e.g. display it on a layout, from any related child record.
- We can access a given set of child records from a parent using a portal or by creating a table occurrence where the relationship between the parent and child is established. We can subsequently utilise various relationship functionalities in conjunction with scripts.
- Am I correct in assuming that the data only exists once in the base table a new table occurance merely establishes the relationship?
- I can see the advantage in terms of the layout of of the relationship table in not having relationships between the base table in that it rapidly becomes totally unreadable. However is there an overhead in approaching relationships in this manner? I ask as some layouts load slowly.
- Is this an accepted approach managing the relationships?
- Are we missing out on any FMP capabilities by omitting to have any parent / child relationships between the base tables?
- Is this efficient?