Filemaker 14, Windows
I am using an index field in a cross table which is calculated as >>> id1 &"_"& id2
I want this field to only contain unique values - which would mean that never two id numbers together are the same. In other words, an instance of a relationship defined between the adjunct tables must only be there once.
It is possible to open the validation tab in de field definition window of Filemaker and define the validation to be UNIQUE.
But when closing the window tab where validation rules are defined, this setting for unique is NOT stored. And the validation never worked.
Either Filemaker is not preventing me setting the value for the validation of unique values, but then discharging it, or something else I did not see.
So, how do I get the internal Filemaker validation engine to work on a text field that is using calculated entries from two id fields?
(The alternative could then be using a script checking for duplicates, or using executeSQL to do the same. But I am in the mood of keeping things are simple as possible, still catching unique validation errors and gracefully letting the user know that he should not enter the same tomato twice.)
(By the way, concatenating two or more id fields inside another indexed field is also a good way of filtering values.)