![]() | Constraint Map | Instance Property Settings | ![]() |
In the Inference section of the Object Control Properties editor there are a couple of properties concerned with Constraints.
constraintMode
The constraintMode property specifies how interference using constraints is to be applied. This should be set to <none> if you are not using any form of constraints with this object.
Instance Properties: Constraints are to be applied by using the instance properties of this object.
Knowledge: Constraints are to be applied by using the knowledge of this object. The knowledgeMode property then specifies whether this is from Cases or from a Tree.
Knowledge & Instance Properties: Constraints are to be applied by using both the instance properties and knowledge of this object in conjunction.
See the topic on Using constraints with Instance Properties.
oneWayConstraint
The oneWayConstraint property can enhance the performance of Constraint Inference by restricting the constraint domain. For example, if 5 objects are constrained in two domains (A,B,C) and (C,D,E) then any change to the run time value of any object will evaluate both constraint domains because they are linked by "C".
This requirement can be avoided by making oneWayConstraint = True for C. Provided that you are prepared to capture C first (on a separate dialog) and accept that C will constrain the other objects, but note that other objects will not constraint C. A typical example would be capturing the type of user up-front and then all objects that are constrained by the type of user will be not be linked into a single domain by the type of user. This would reduce the size of the overall domain which has to be evaluated by constraint inference.
selectionMode
The selectionMode property sets the method to be used for inference for the object.
Decision uses the Decision Tree or the Cases table and is dependant on the setting of the knowledgeMode property.
Pattern Match uses inference from pattern matching on the instance Properties. See the section on Using Inference from Pattern Matching.