Wednesday, October 15, 2003

When I was young, I used to respond to inquiries about my day by saying We ate lunch; and we had recess... Since then, I've added a few things to my list, like I had coffee, and I attended a seminar. But if asked what I did at the end of the day, my inclination is still usually to blink owlishly and ponder for a while. If pressed, I'm likely to blunder and talk about whatever I was thinking about most recently, whether or not it had anything to do with the bulk of my day. It takes me a while to digest experiences.

I remember food, though. So far, this has been a good week for food. I spent Saturday with Winnie, and we had Zachary's pizza for lunch and black bean soup with sausage for dinner. The pizza was thin-crust, not the usual deep-crusted Zachary's pie with all the fillings on the inside. They didn't have the deep dish pizza by the slice, and I still owe Winnie a deep-dish pie at some point, but the regular slices were still good.

The rest of the activities of the afternoon were like lunch: they didn't go as planned, but they turned out well in the end. We went to Sausalito and spent some time walking around there, and planned to then walk over the Golden Gate Bridge. But we missed the turnoff at the beginning of the bridge and ended up stuck in traffic in San Francisco. This weekend was the Fleet Week celebration, and about a million people went to San Francisco to see the Blue Angels stunt flying. Consequently, we were in traffic for a long time. But we listened to Prairie Home Companion on the radio while we waited, and had a very satisfying dinner when we got back: bean soup, bread, asparagus, potatoes, and chocolate turtle brownies for dessert.

Sanjay had some of his students over for dinner on Sunday evening. The food and company were good, and the kids were cute. Dessert was a fancy-looking chocolate cheesecake, though Sanjay claims is simple to make. Whether it was simple or difficult, it was delicious. On Monday, Dave D. and I had ceviche (a Mexican fish dish) and tacos and talked about a change-of-basis problem. We finished the brownies from Saturday for dessert. And later I had tea biscuits with my tea, and ate some of the chocolate turtle candies that Esther bought.

I seem to be eating a lot of dessert recently.

Perhaps there's a correlation between more desserts and more computer failures, for surely that has been another characteristic of this week. Hardware that has thus far been quite reliable -- my laptop, for instance, and the router that Patxi and I use -- has decided this week to show its cantankerous side. My laptop usually goes for weeks at a time between reboots. So far this week, though, I have rebooted at least once a day. You'd think I was running Windows, but I'm not. I believe I know what the problem is, though. There are documented peculiarities in the video hardware on this laptop, and I've been stressing that hardware more than I usually do in the preparation of my fluids homework for this week. Perhaps I just need to upgrade the driver.

While I think I may know the issue with my laptop, I have no idea what has gotten into our wireless router. Occasionally the router just hangs; at other times, everything but the DNS redirection seems to work. Rebooting the router usually fixes the issues -- though not always -- but it's still troublesome. Perhaps I should update the firmware for the router as well as the video drivers for my laptop.

The matrix computations seminar today was made more entertaining by a hardware failure, too. The overhead projector would not project, for no reason that any of us could ascertain. So we all tested our eyesight by looking at the monitor and at the whiteboard until one of the media technicians could come to fix the problem. He did come in the middle of the hour, but eventually declared, Sorry, folks. I'm going to have to look at this later. It just doesn't seem to respond to commands. At that comment, Ioana leaned over and whispered, We could have told him that. Which is true, but probably beside the point -- I doubt he would have believed us.

I've prepared a computer presentation for the talk I'm to give Friday, but perhaps I'd better print overhead slides tomorrow as well, just for insurance. This does not seem to be a week to test the determination of whatever spirits enforce Murphy's laws. I imagine that I'll seem curmudgeonly enough by the end of my talk without any projector failures, though. My talk has three parts: an overview of resonant microstructure design; some ideas underlying eigenvalue computations; and advice about how to be a wise consumer of numerical software. I've tried to leaven the last section with humor, but I think people draw some really boneheaded conclusion from poorly conceived numerical experiments, and I expect my opinion of such shenanigans shows in the slides.

In reviewing my slides, I'm also reminded that graduate students supposedly turn into their advisors over time. I hope that I'm combining the better parts of Jim and Kahan's respective presentation styles, and not taking the worst of both and adding a few new flaws of my own. For better or worse, I can see the influences of each of them in this talk.

Usually, I drastically underestimate the time it will take me to finish a presentation. I've gotten better at recognizing my optimistic tendencies and compensating for them, but I somehow managed to finish the slides for this presentation in about the time frame I had planned. Perhaps to compensate, I spent far too much time this week on my fluids homework. The problem wasn't particularly hard, but it was fun and it generated pretty pictures, and so I spent a lot of time exploring. In essence, we were looking at the movement of fluid around two propellers, each active in turn. It's a little like the flow generated by an egg-beater; the material is pulled around the propellers, and at the same time it gets folded up, so that very quickly it becomes interestingly complicated. So I spent far too much time generating and admiring pictures of what happens when you draw an initially square grid in the fluid and then stir it up with the propellors, or of what happens when you drop a string into the flow and then let it get stretched and folded. The problem is due Friday, but I turned it in early this afternoon, purely out of defense of my productivity for the rest of the week.

Also out of defense of this week's productivity, I have decided to skip the IEEE meeting for the month, which takes place tomorrow. I really want to get some things done by the week's end, and it does not seem wise right now to take a day off to pick floating point nits. If I have some extra time during the week, I'd much rather spend it on books than on meetings. I still have not finished To Say Nothing of the Dog (Willis), and there are new books I would very much like to read some time soon: Quicksilver (Stephenson) and Paladin of Souls (Bujold). I'd also like to spend more time working my way through the technical texts I bought at the beginning of the semester. Reading and writing and thinking seem much more important right now than trying to stay awake through meetings does.

On the whole, I think I've done well recently at balancing time reading and thinking with other demands. I spent some time this afternoon hiding at a cafe in Berkeley and reading through Kahan's notes on Laguerre's iteration for finding polynomial roots. And last night I spent some time reading the most recent SIAM News issue. Of course, the thing I remember most from reading SIAM News is that there is a new edition of Allgower and Georg's classic Introduction to Numerical Continuation. Their old edition was released into the public domain some months ago, and so I returned the aging copy I'd checked out from the library and switched to the electronic version. But if there is a new edition with better typesetting and new material, perhaps I'll invest in a copy. It's a book I already use, after all.

I'm convinced. I think I'll go buy that book, now. And then I'll make some food. And perhaps take a short recess.

  • Currently drinking: Green tea