The ORM Foundation

I keep getting this validation error  message, when I tried to put subset constraint between two roles. 

 "Constraint Requires exactly two role sequences. Currently :One sequecnce with 2 role(s)"

It seemed to work fine earlier, and I tried to correct the constraint to point from the subset to the superset instead of the other way round, even though there was no validation error, ( I have not put in sample population yet?)  Now even the earlier " wrong" way of attaching the subset constraint does not even work. What is my mistake? I am cliking on one role each time from two different fact types played by the same object to attach the constraint.

thank you for all the help! 

Hello Berhanu,

A subset constraint (along with an equality constraint and an exclusion constraint) is classified as a set comparison constraint. This means that these constraints compare the population of ordered sets of roles, not individual roles. (The instructions and error messages in NORMA refer to these sets role sequences.)

So, to enter a set comparison constraint, you click in sets of roles, not individual roles. The sets are separated by double-clicking the final role in the set. If your set comparison constraint happens to be single-column (one role per set), you still need to signal the end of the set by double-clicking to complete.

You can see constraint editing instructions by hover over the constraint shape while a constraint editor is active. The instructions will differ for set comparison constraints (subset, equality, exclusion) and other constraints that span multiple roles (external uniqueness and xor, disjunctive mandatory, ring, frequency, value comparison).

So, to enter a subset constraint between two individual roles you would do the following:

1. Click the toolbox (don't drag, or you'll have to activate the constraint editor yourself by double-clicking the shape).
2. Click on the diagram to place the constraint shape.
3. Double-click the first role.
4. Double-click the second role.

 Thank you Matt, It worked, Thank you 

 Thank you so much Matt, for the quick response , it worked!

thank you

berhnau