Acknowledgments

My years at the Laboratory for Computer Science have given me a remarkable opportunity to work with and learn from a range of bright and interesting people. This thesis draws upon that background not just for technical issues, but as a standard to live up to.

Steve Ward, my thesis supervisor, provided an important part of the support for this thesis. As well as starting the NuMesh project, he encouraged me to explore issues that were a little more fundamental and a little less mainstream, and both directed and participated in the chaotic and productive group meetings out of which came many of the most interesting results of the NuMesh project.

Thanks are also due to the other two members of my thesis committee, Gill Pratt and Frans Kaashoek. Their insistence on keeping the big picture visible, making the experimental framework of the results clear, and being precise in drawing conclusions, helped to make the thesis as readable as it is.

The other graduate students in the NuMesh group, along with the Alewife and CVA members of the overall Computer Architecture Group, gave me much of the day-to-day social and intellectual life of the lab. In particular, I would like to thank Dave Shoemaker for great collaboration and for refusing to allow me to ignore hardware, Pat LoPresti for helping me sound out ideas on languages and scheduled routing as well as putting up with continuous Tadpole bug reports, and Anne McCarthy for her competence, warmth, and supportive fire-fighting. My officemate Kirk Johnson was a source of many an interesting conversation, late night systems hacking, and friendly support. Thanks also to Frank Honoré, Russ Tessier, John Nguyen, and John Pezaris for many interesting times.

Thanks to the people who volunteered their time to read through drafts of this thesis and volunteer suggestions, Michael Littman and Joe Morrison, and Dave Shoemaker.

My parents provided me with encouragement throughout my thesis, giving me perspective from the other side of the Ph.D. fence and letting me know I was in their thoughts as I passed the various milestones of the thesis path.

Finally, and most importantly, thanks to my wife, Elaine, without whom I'm not sure I would have made it through--and without whom I certainly wouldn't have started at MIT when I did! Her support and encouragement were a large part of what kept me going; her advice helped keep me focussed on making progress week by week; and her shared happiness at every milestone passed was a joy. Just as important, even at times when my thesis didn't seem to be going the direction I wanted it to, her presence in my life provided a fundamental sense of balance and happiness.