Friday night I went to a lab dinner at my prof’s place, then drove down to CT. On Saturday we had two overlapping Bat Mitzvahs at different synagogues, and then the accompanying fabulous parties, for about 13 hours straight. Today I’m hanging out with some friends in Westport, and tomorrow is The Picnic. It’s quite a social weekend.


Sunday I inflated both tires with my new pump and adapter. I rode the bike to work Monday and Tuesday, enjoying the confidence of properly pressurized rubber. Coming out from work Tuesday afternoon, my rear tire was dead flat, like someone had stuck a pin through it. This is far from the first time this has happened. I must be doing something wrong, but I can hardly imagine what.

Either that or some vandal is sneakily pricking my tires.


The Polywell fusion device developers stopped making press releases some time last year, and basically disappeared from my consciousness. I imagined that the project was probably dead, given its tenuous position at the edge of crackpotdom. As it turns out, though, the US Navy is still funding them, in the form of a $7M+ grant from the stimulus grant boost approved last spring.

They’re under a publishing embargo, so they can’t talk about their work, which I find sort of hilarious. The project was the subject of breathless hype by its founder, whose premiere invention (the Bussard Ram Scoop) is literally fictional. It claims to solve a problem into which many billions of research dollars have been poured, without introducing any new technology, by a simple design. Its entire scientific claim to success rests on the detection of a total of nine neutrons. This sort of edgy science calls for intense public scientific discourse to resolve the mystery, but the terms of the grant make this impossible, leaving us to wonder.

Schwinn Multi-Function Foot Pump

Yesterday, on my errand shopping spree, I bought a Schwinn bicycle pump at Target for $13, including Presta-Schrader valve adapter. The pump itself is mediocre, it turns out, flimsily built with a pressure gauge of questionable utility, but that makes it only a touch worse than all the other bicycle pumps my family’s owned over the years. The included Presta adapter, however, is worse than useless. It’s all plastic, and at first appears to work, but in fact just wastes your time and does not actually effect the flow of air.

Proper Presta adapters are just a machined piece of brass with a rubber gasket, so small that I dropped the last one somewhere around Boston and didn’t hear it fall. Bike stores typically stock them in a big jar at the checkout counter, like mints. At the fancy bike shops around Boston you’re lucky to find them for less than $3, and I wondered if there might be a better price somewhere less bikey. I found it, but not where I expected: EMS carries them for $0.80. I was doubly surprised that EMS carries bicycle stuff and that they sell anything at all for a reasonable price. Of course, I’d just let all the air out of my bicycle tires.

I laced up my jogging shoes and made the most of the weather on the way.


After a long day of errands I went to meet a friend at Conga, a restaurant/salsa club, tonight. There was a lesson first, which was nice, since I haven’t danced salsa in years. Unfortunately, the lesson turned out to be about 10% of what I would have needed to be able to survive on that dance floor. This club, it turns out, is populated by a crowd of very skilled regulars. The dance floor was packed with people doing spins in double time, with signaling and foot positions so highly evolved that only a hint of the original basic remains. It was impressive to watch, and disheartening. I would have to take a whole lot of lessons before coming back to Conga on a Saturday night.

Oh, Microsoft

You are always good for a laugh.


In its HTML5 support, IE9 will support playback of H.264 video only.


In its HTML5 support, IE9 will support playback of H.264 video as well as VP8 video when the user has installed a VP8 codec on Windows.

Time between quotes: three weeks. I feel sorry for the engineers.

High volume dust filtration

Does anyone know of a system for filtering large volumes of air for dust? I’m not worried about allergens here, just visible dust particles. Everything in my apartment seems to get covered in dust, sometimes even vertical surfaces, and it only gets worse when we open the windows in the summer. I need something that can filter dust from a lot of air, fast.

I’m still of half a mind to duct-tape a big furnace filter to the front of my box fan and see what happens.

Alzheimer’s Vaccines

In 2002, Elan/Wyeth/Pfizer stopped a trial of AN-1792, an experimental Alzheimer’s vaccine, after a pattern became apparent in which about 4.5% of those receiving treatment developed dangerous encephalitis. (Most of them did recover, though.) The vaccine didn’t appear to be a very effective treatment, and Wyeth called the whole thing off.

Follow-up studies on the 300 patients enrolled in those initial trials have begun to suggest a different story, that AN-1792 was quite effective, though perhaps not miraculous. This, for me, always seemed like an incredible missed opportunity. How could any drug company let go of a drug that substantially disrupts Alzheimer’s disease progression, just because it happens to kill a few percent of patients? Did the FDA block it based on very conservative safety guidelines? Did the company fear that they would get endless lawsuits, no matter how big the warning label?

Having had a few years to mull it over, I think the real answer is that they expected to get crushed in the market by the first competitor to make a safer version. As it turns out, the competition is already here. There are now clinical studies recruiting patients for experimental treatment with Novartis’s CAD106, which was announced in 2005 and seems to have passed its safety qualification already. I can well imagine that Wyeth’s industrial espionage was sending them hints about it already in 2002. If it turns out as effective as AN-1792, I expect it to be a huge success. There are also other vaccine efforts in various stages of development.

I consider myself optimistic. A vaccine that can slow down disease progression in patients with Alzheimer’s might prevent the disease from ever occurring if administered prophylactically.


Every spring the streets of Boston are retextured and repaved, in the hope that by June or July we will again have things that can reasonably be called “roads”. Cambridgeward off the BU bridge this evening I found Brookline St. to be in such a state. It seemed mostly unremarkable at first, an irritating exercise in dodging protruding plumbing in the dark.

First I noticed the bricks. Peeking through in patches were flat red bricks, clean-edged and neatly arranged, that paved the entire street. This was pretty cool. I always love the discovery a patch of cobblestone under a paved street, and brick is in some ways even more intriguing.

In the dim streetlight I could see a central ridge, or maybe two, but it took half a mile before I caught a telltale gleam, polished to mirror finish by the grinding machines. Running right down the center of Brookline Street. is a hidden, forgotten narrow-gauge tramway.

I can’t tell you anything about the history of the rail there, because the line is not shown in any historical railroad map I could find tonight. If you find disused public transportation (or perhaps industrial transportation) as eerie and exciting as I do, make sure to pass by Brookline Street. You’ll have to be fast, though. Soon (well, hopefully soon) the rails and bricks of old Cambridgeport will be hidden again beneath the pavement.

EDIT: I checked back in daylight. The rails are clearly visible from Putnam Ave. to Erie St.. North of that there were men with shovels, and bare dirt where the rails would be. At Erie St. the rails turn into two narrow lines of yellow bricks amid the red, suggesting that the rails were removed here while the road was still paved in brick.


Last night I transferred all my data from my old laptop onto the new one. I did it in the fastest way I could think of: hooked up a crossover cable and ran rsync. It worked like a charm, at 20MB/s once I turned off compression. Even so, copying many thousands of tiny files is a slow process, and I let it run overnight.

In the morning everything was ready. I wasn’t quite prepared for how complete the transfer would be. MyGnome panel icons, Firefox add-ons, desktop contents and font settings all copied over. The last was actually a bit of a problem, and it took half an hour to figure out that I needed to delete my old .gtkrc in order to let Ubuntu’s font settings work.

My real motivation for this was copying my several gigabytes of archived mail, currently stored in .thunderbird/. Running Thunderbird on the new system immediately loaded all my old account info and let me log in, despite being a different version (3) from what I was using before. It’s spent about 8 hours so far churning through those 150000+ messages, indexing them for fast search later. The indexing is pretty unobjectionable though, and I’ll be thankful later if the search is faster than Thunderbird 2, which made me wait for minutes.

It occurs to me that for those who don’t know about crossover cables, rsync, and dotfiles, copying settings from an old machine to a new one could be a real pain, in Linux or any other OS. We tend to think of a user’s documents as what’s valuable, but their archived mail, bookmarks, Zotero citations, and myriad other configuration data are also important. It seems like Ubuntu’s answer is Ubuntu One, but I wonder if there isn’t a more direct solution.