ABSTRACT
For this, the final chapter of the festschrift for Ben Murdock, our assigned task was to somehow summarize and debate, critique and discuss, or relate and integrate the preceding four chapters on distributed memory. Because one of us is an editor of this volume and also contributed a relevant chapter, we shied away from the direct approach to commentary chosen by the critics of the other sections, despite the fact that this meant foregoing an opportunity to relate jocular anecdotes and personal reminiscences involving Ben Murdock. Instead, we chose to discuss an issue that is of at least implicit concern to all chapters in this section; namely, how one might best measure retrieval performance of a distributed memory model.
