In which I go to Boston and Kate goes to Reedsburg.
I went to the 2002 Operating Systems Design and Implementation (OSDI '02) conference starting on Sunday, Dec. 8, in Boston. It was good to get out of Palo Alto for a few days. It was good for Kate, too, since it gave her a chance to work on final papers for her classes without me around to distract her.
The flights there were uneventful, even if I have a tendency to get ticked off by airline security. What I object to most is the undignified treatment. I don't like being ordered around by people whose salaries I am paying. I don't like taking everything out of my pockets, taking off my shoes, my coat, and my belt, taking my laptop out of my backpack, and then perhaps still being patted down or having some kind of explosives test run on my laptop or stuck with a bag search. I don't like having to go without my multitool or pocket knife unless I check a bag, either.
Anyway, we checked in at the conference hotel in Boston and found out that they had no rooms for us. This seemed odd, since we had prepaid reservations with an online travel site. The desk staff explained to us that online reservations, even prepaid ones, don't really reserve a room, and sent us across the street to the Radisson, which had a lovely room available. The next day the conference hotel scrounged up a room, which turned out to be not nearly as nice. By complaining to the manager, we did get a couple nights of free hotel stay out of it, but as Stanford was paying for everything anyway that didn't really matter too much in the end.
It was a good batch of papers. I particularly enjoyed ``Practical, Transparent Operating System Support for Superpages,'' ``Vertigo: Automatic Performance-Setting for Linux,'' ``Cooperative I/O: A Novel I/O Semantics for Energy-Aware Applications,'' ``Memory Resource Management in VMware ESX Server,'' and ``TCP Nice: A Mechanism for Background Transfers.'' The last is a way to make use of otherwise wasted bandwidth for background transfers using TCP. It solves a problem for which my research group needed a solution anyhow, and it does it in a simple, elegant way by modifying the congestion window backoff strategy.
Costa presented my group's paper, ``Optimizing the Migration of Virtual Computers,'' in the final session on Wednesday morning. He seemed a little nervous, speaking with some hesitation, but the talk went well in all. He handled the questions at the end well enough, although I would have answered some of them in a little different way. It was a success.
I met Shaya Potter for the first time in person at the conference. We are about the same age and both in the second year of Ph.D. work in computer science, although he is at Columbia and I am at Stanford. Both he and I have been involved in Debian since high school, and so we've run into each other by way of Debian mailing lists.
On Tuesday night, Jim and Ramesh and I walked over to MIT to have a look around. It seems very much a tech school campus, not all that exciting. I was surprised to see that the MIT newspaper has PhD in it too—I didn't know it was in anything but the Stanford Daily. Also, as we wandered through, a bunch of MIT students walked by and offered us homemade cookies.
We also visited Harvard. Its campus is very much my vision of an old European-style university. Red brick is to Harvard's campus as yellow sandstone is to Stanford's. It was late, about 9 pm, so there was nowhere we could go inside, but we did wander around for quite a while. No free cookies, so I concluded that it wasn't as friendly as MIT. Finally we took the ``T,'' Boston's subway system, back to the hotel.
All of us but Mendel were on the same flights home on Wednesday afternoon, so we had an impromptu meeting on the Chicago-San Jose leg for most of the 5 hours or so. Usually I get antsy after less than 2 hours in a meeting, but this time I was strapped in. We went through our future plans in some detail and figured out what conferences we want to target.
When we arrived in San Jose, I was still feeling alert, so I called Kate. She said she'd still be up, so I promised to go see her when I got back. I did, arriving half past midnight, and we spent a long time through the early morning catching up on things. We even spent all of Thursday together. I proofread one of her papers (23 pages!) and we went out to Menlo Park for lunch in the afternoon. We also visited an art store in Menlo Park. It had very expensive original art by Dr. Seuss, but it also had cheaper prints by an artist that Kate liked very much. She bought one of those and I bought a Christmas present for my mom.
Kate left for home in Reedsburg, Wisconsin on Sunday morning. The airport shuttle bus came to pick her up at 10 am. I helped with a couple of things around her apartment while she packed, and then she had to leave. It's been less than two days that she's been gone, but I miss her already. I'm going to visit her at her parents' house on Dec. 24, and the day can't come soon enough...