Integrating pointer variables into one-way constraint models

Abstract

Pointer variables have long been considered useful for constructing and manipulating data structures in traditional programming languages. This article discusses how pointer variables can be integrated into one-way constraint models and indicates how these constraints can be usefully employed in user interfaces. Pointer variables allow constraints to model a wide array of dynamic application behavior, simplify the implementation of structured objects and demonstrational systems, and improve the storage and efficiency of constraint-based applications. This article presents two incremental algorithms—one lazy and one eager— for solving constraints with pointer variables. Both algorithms are capable of handling (1) arbitrary systems of one-way constraints, including constraints that involve cycles, and (2) editing models that allow multiple changes between calls to the constraint solver. These algorithms are fault tolerant in that they can handle and recover gracefully from formulas that crash due to programmer error. Constraints that use pointer variables have been implemented in a comprehensive user interface toolkit, Garnet, and our experience with applications written in Garnet have proven the usefulness of pointer variable constraints. Many large-scale applications have been implemented using these constraints.

Lessons learned about one-way, dataflow constraints in the Garnet and Amulet graphical toolkits

“…The two schemes used for one-way constraint satisfaction are the mark-sweep strategy [Demers et al 1981;Reps et al 1983;Hudson 1991;Vander Zanden et al 1994] and the topological-ordering strategy [Reps et al 1983;Hoover 1987;Alpern et al 1990;Vander Zanden et al 1994]. A mark-sweep algorithm has two phases:…”

“…In collaboration with a number of other researchers, Hoover later devised a second topological-ordering scheme that kept constraints in precise topological order and evaluated each constraint at most once [Alpern et al 1990]. A variation of this scheme that accommodated pointer variables and arbitrary code in the constraints was devised for Garnet [Vander Zanden et al 1994].…”

“…Since topological-ordering algorithms cannot be used as lazy evaluators (see Section 5.2.1), mark-sweep algorithms gained widespread usage in graphical interfaces. Mark-sweep algorithms that accommodated pointer variables and arbitrary code in the constraints were devised for both Garnet and Amulet [Vander Zanden et al 1994;Myers et al 1997].…”

“…(2) Pointer variables: A constraint can reference variables indirectly via pointers [Szekely and Myers 1988;Vander Zanden et al 1991]. For example, an object can be made to appear 10 pixels to the right of the previous object in a list by writing the constraint left = self.prev.right + 10 where self is a pointer to the object containing left, and prev is a pointer to the previous item in the list.…”

“…For example, an object can be made to appear 10 pixels to the right of the previous object in a list by writing the constraint left = self.prev.right + 10 where self is a pointer to the object containing left, and prev is a pointer to the previous item in the list. 1 (3) Automatic parameter detection: A constraint's parameters are automatically deduced as the constraint executes, so the programmer does not have to declare a constraint's parameters [Vander Zanden et al 1994]. (4) Support for cycles: A constraint is evaluated at most once if it is in a cycle.…”