Minimizing flow time on a single machine with job classes and setup times

We examine the static sequencing problem of ordering the processing of jobs on a single machine so as to minimize the average weighted flow time. It is assumed that all jobs have zero ready times, and that the jobs are grouped into classes with the property that setup tasks are only required when processing switches from jobs of one class to jobs of another class. The time required for each setup task is given by the sum of a setdown time from the previous class and a setup time for the new class. We show that an algorithm presented in the literature for solving a special case of this problem gives suboptimal solutions. A number of properties of the optimal solution are derived, and their use in algorithms is evaluated. Computational results are presented for both a branch-and-bound procedure and a simpler depth-first search.     

The symbolic politics of ballistic missile defense: seeking the perfect defense in an imperfect world

Ballistic missile defense (BMD) politics present an interesting evolution in how the USA, especially Congress, has come to think about BMD both as operational reality and as a symbolic policy. The argument here is that BMD's operational reality is increasingly overshadowed by its symbolic aspects. Such a status arose from rapidly changing international and domestic politics. The end result is a situation in which BMD policy in a sense floats above the question of its actual combat effectiveness. Its primary mission in part is sustaining US capacity to remain a global power and support its allies.