Tue Sep 22 12:29:12 CEST 2009

Constraint Programming from CTM Chap. 12

Combination of propagation and search.  Try local deductions first,
and solve the rest with search.  Choice can be introduced by
transforming a constraint program P into P^C and P^(~C).

Local deductions are implemented as propagators.  ( Equations,
i.e. sets of linear equations, are propagators that can be turned into
functions, but there is a more general relational framework. )

Choice is inserted implicitly using a heuristic ``distribution