I sometimes wonder if I am doing the right thing by spending time on projects for OLPC and Sugar Labs. I do my best to work on it only during free time, but then, some graduate students spend their whole life in lab, and simply don’t have free time. What keeps me involved, apart from the social ties, are writeups like this one, from Stephanie Selvick of the OLPC Senegal student team.

In the first session, I was paired up with a male teacher who was nervous about communication. He asked if I spoke French, and I apologized that I did not. He then continued to have entire conversations with me in English. He began by saying he didn’t think the teachers would use the computers in their classroom. This particular school received their XO’s 6 months ago, received their first teacher training about 2 months ago, and will be introducing the students to the computers for the first time on Monday. … After that, my partner began by pointing to each individual activity and asking how each could be implemented into his curriculum. Although I scoff at the ideology that “curriculum” is somehow not important, I began to understand the dilemma. I said we should begin by opening one activity and we can answer those questions about curriculum after.

Distance was his gem! The distance activity measures the distance between two XO computers. It sends sound waves between them and a measurement in meters pops onto the screen. He loved it. He began brainstorming assignments for his kids to measure the distance around their homes or rooms and said they could also figure out area from those numbers. Although it took us the full 90 minutes for him to be comfortable with opening the program and inviting another XO to share the program, his enthusiasm was quite the relief. When each person shared what they learned from the first session, it was great to see his enthusiasm about distance in other teams about other activities. The goal of making teachers the experts for each other felt underway.

The other conference

All this week was the NAMIC-NCIGT 2009 Summer Working Session, or something. I can never remember its name. It was, in a way, one of the best conferences I’ve ever attended. It was held at MIT, which is much closer to my apartment than my usual office, and well within walking distance. That was especially convenient given my present lack of a working bicycle.

The conference was essentially just one big room at MIT, full of tables, chairs, and power strips. Over 120 people came from across the country (and a few other countries), all funded by one of the NAMIC or NCIGT grants. The plan was simple: take out your laptop and get to work. If you have a question, chances are that the person who knows the answer is in the room. Go find her and ask.

The purpose of the conference was just work, and so there were essentially no talks, apart from the first day when people stated what they intended to work on, and the last day when they reported their progress. There were optional tutorial sessions on technical questions, but most people just kept on working, breaking only for the gourmet catering, serving breakfast, lunch, and and afternoon snack.

I learned a lot by talking to the experts in the room, and got a fair bit accomplished with their help. It was unquestionably productive.

The Open Video Conference

The (“inaugural”) Open Video Conference was an unusual affair, with participants spanning the full range of political theorists, artsy filmmakers, and numerical algorithms gurus. As such, it’s difficult to pin down a single purpose or thesis for the event. The closest I can come is to note that a single word, “democratizing”, was used ad nauseam.

The idea of democratizing video is simple enough. For the past 30 years, moving pictures have been the dominant force in our social discourse. The accessibility, immediacy, and veracity of a video make it the ideal medium for the discussions a society has about itself. Our great social debates happen in movie theaters and on television screens, far more than in books or newspapers, because viewership is so much higher than readership, and compelling news footage beats prose any day of the week.

We have had a mature industry of moving pictures for about 100 years now, and for all that time, video production has been essentially “industrial” in nature. Until the 1980s, producing video footage of any quality required equipment so expensive that only large corporations could afford to build a television studio or film a movie. Even in the 1990s as camcorders approached reasonable prices, worldwide distribution remained an oligopoly of the major networks and studios. An independent filmmaker who produced a documentary could not hope to have more than a few people see it unless she convinced a major movie distributor or television channel to show it.

In the last few years, the barriers to distribution have been crumbling. With the advent of YouTube, it is possible for anyone to make a video, using equipment that costs less than $200, and upload it onto the web for all to see. For the first time ever, it is now possible for anyone to make a video, publish it, and have millions of people watch it, without any negotiations with distributors, networks, or studios.

The social implications of this are dramatic. The most powerful tool we have for emotional argument and revealing truth is available to any ordinary citizen. If you have something to say, and people want to hear it, you can, and they will. This represents an unprecedented decentralization of the means of persuasion. The media’s old guard, the gatekeepers of television (and even radio), are losing their stranglehold on public opinion. We are about to find out what the people really think, whether we like it or not.

Talks at the conference covered many aspects of this change. Some talks discussed its history and current status, in the US, Europe, and elsewhere. Some discussed the future, and its implications for our society. One oft-mentioned example was the video of protests coming out of Iran, recorded by average Iranians on their mobile phones, and uploaded to YouTube for global distribution. Obviously, this video would never have made it out of Iran ten years ago, when the only mechanism of distribution was a state-controlled television broadcaster. Even US television stations haven’t run most of the footage, for fear of offending their viewers sensibilities (and perhaps the Iranian government, which is essentially holding their reporters hostage).

Most sessions, though, seemed to focus on the imperfections in the system, and what we can do to prevent the formation of a new gantlet of gatekeepers. One obvious example is YouTube. Most people who want to achieve wide distribution of their video post it on YouTube; they may not even retain a copy of it themselves. This makes YouTube a dangerous single point of failure. In fact, in order to avoid lawsuits, YouTube allows major media conglomerates to remove videos from YouTube instantly, without any human intervention, even if those videos are entirely legal according to US copyright law. As it currently stands, then, YouTube is far from manifesting the full promise of Open Video: that no one can censor you.

Another major limitation is the licensing of the technology used to distribute videos. Most video distributed today is in MPEG formats, which are heavily patented. As a result, anyone who wants play back video, such as the digital signals now broadcast by all US television stations, must buy a patent license for the technology. This prevents developers of Free Software, like the Firefox web browser, from including support for these video formats in their products. To solve this problem, a group of engineers called Xiph has been developing a video encoding system called Theora that does not infringe on any patents, and so can be used royalty-free. The next version of Firefox, due to be released in a few days, will include support for Theora video, already available on sites like dailymotion.com.

Other presenters discussed the cost of internet bandwidth and how to decrease it, the next generation of free tools for movie editing, and many ways to make use of the Open Video phenomenon. In fact, I only know what was said in about a quarter of the sessions, because there were many parallel tracks.

It was a remarkable conference. I learned quite a bit about the social scene, and I hope many people learned a lot about the technologies. I can’t predict where this phenomenon is headed, but it is definitely worth watching.


I’d love to tell you about the Open Video Conference, but I spent all day at my conference for this week, and tomorrow morning it starts again at 8:30, so I haven’t had any time to write.

Maybe someone will publish a good article about it, and I can just post a link. That would be very convenient.


I spent the weekend at the Open Video Conference, held at NYU in Manhattan. I’d love to tell you the whole story, though it was such a bewildering affair that I’ve had trouble synthesizing it. However, I’m still feeling sleep-deprived, and I’ve got another conference tomorrow, so I think I’m going to go straight to bed, and see if sleeping on it will help to distill the experience.

It will certainly be helpful to start the next conference, the National Center for Image Guided Therapy Project Week, on the right foot. It’s going to be a long week.

Sacred Harp

I was at a party Saturday night, and over the blaring dance music I struck up a conversation with the girl crammed in next to me in the crowded apartment. At some point she mentioned that she sings, and so we started talking about singing. She informed me that she sings in a style called the Sacred Harp, which I had never heard of. There was another Sacred Harp singer there that night, and so they gave us a small demo in another, quieter room. They also implored me to come try it out, at a “singing” on Sunday.

This demands a bit of explanation. Sacred Harp is a “non-performance” art form; there is no audience, and so there is no distinction between performance and rehearsal. The singing on Sunday was at a church, as they almost always are. It was near my lab, and since I was planning to be in lab anyway it seemed like a convenient enough experiment.

The Sacred Harp is a 550-page book of sheet music for hundreds of songs. The scores are all written in four part harmony, and, unusually, in shape-note form. Shape-note music is a sort of mnemonic device, designed to improve the ability of nonspecialists to learn music quickly. Each note is assigned to one of four geometrical shapes, which are substituted for the standard note-heads. Each shape is also associated with a syllable drawn seemingly arbitrarily from the classic solfege scale. When singing each song, the singers first run through the music singing these syllables, in order to get the melody into short-term memory. The song is then sung again, with the actual lyrics. The system is very interesting, and works quite well, largely because virtually every song is shorter than a minute.

The rehearsal took place in the nave of the church, with perhaps 25 participants, seated by voice part, facing the center. Any singer could choose a song, which they would do by calling out its page number. We would then turn to that page (the books are entirely standardized), and the singer who chose the song would conduct it from the center. Rather than tune the group from a pitch pipe, anyone can call the pitch from memory, and nobody cares if the whole thing is transposed a few steps one way or the other. The rehearsal has a rapid-fire quality; over the course of three hours we must have sung a hundred songs, each one called out by its number.

The music itself has a haunting quality. It is essentially all Southern Christian hymns; the form is particularly associated with the Primitive Baptists of Alabama and Kentucky. The chords are very open, and often minor. The arrangements mostly date from the middle 1800s, with few later than 1930. The style of singing is tremendously loud, with a strident tone, no dynamics, and an absolutely rigid rhythm. The effect is far from any modern style, but it is surprisingly beautiful. Between the chords and the nasalized tone, there is a tendency to generate audible overtones. To my mind, the obvious reference point was the soundtrack to O Brother Where Art Thou.

The lyrics are very … period. We spent the three hours singing that our only embarrassment was that we didn’t love Jesus enough, that we would all soon be dead so earthly pleasures are fruitless, that we wished we were dead already so we could be in Heav’n, etc. There’s little in the way of imagery or rhyme. It’s strong stuff, and would sound just about right coming out of the mouth a stereotypical early-20th-century southern church lady.

For me… by the end of three hours, I felt like I had used up my entire annual quota for church music. Maybe I would be more inclined to do it again if the subject matter were a bit more secular and diversified, maybe if there were a bit more poetry, and a wider (or any!) range of musical styles… and maybe if it didn’t feel quite so much like an art form created by the Sacred Harp Publishing Company. For now, let’s say I won’t go out of my way.

The makeup of the group was quite surprising. There were perhaps three people over the age of 50, with southern accents, for whom this could be called part of their culture from childhood. The other 20+ people seemed to be almost entirely in their 20s or early 30s, including plenty of fashionable hipster types and reasonably talented singers. I can only guess at their motivations: some mixture of the social experience, artistic fulfillment (without the stress of performance), anthropological interest, ironic amusement, and maybe, for some, a hint of a Christian connection, without the overt structure of organized religion.


I took my broken bicycle to a bicycle expert friend, who whipped out his portable bikestand and started investigating. After an hour or so, we concluded that most things are fine, including the frame, pedals, headset, bottom bracket, etc. However, the rack will just have to be discarded, the rear wheel will need to be trued, and the derailer tab on the frame will need to be bent back into alignment. I’ll also need a new brake housing and cable, but that’s pretty minor. I think the rear derailer is probably fine, but it’ll need to be verified.

This process has forced me to take a closer look at the damage… which make me even more confused about what happened. For example, the rear rack was severely contorted, despite being made of thick strong aluminum (I can’t bend it all by hand). There’s a white-colored scrape on the side that was not touching the ground, perhaps white transfer from whatever hit it… but it’s hard to be sure. It’s also hard to imagine what that object might have been. It must have been pretty heavy, or moving pretty fast, but that part of the bicycle was on the sidewalk, far from the kerb… and the whole thing was thoroughly locked in place; it’s hard to imagine that it was moved after being hit.

Oh well. Physics often provides a mystery.


Last night I went to a costume party at Pika, one of the MIT undergrad living groups. I wore my one costume, the Spartan Warrior outfit given to me by the Biophysics department, complete with helmet, facemask, chestplate, and cape. I got a lot of compliments on it, though since I didn’t make it, buy it, or even choose it, I can’t take a lot of credit for it.

Pika is pretty close to my apartment. Under normal circumstances I would have biked there, but with no working bicycle, I decided to just walk. I couldn’t easily carry the costume, so instead I wore it. Hopefully, I made a few people’s Friday night a bit more surreal.


A fellow Biophysics student had his (successful) Ph.D. defense today, and I watched. It was seriously impressive. The dude was on 12 papers by the time he graduated. I can’t even imagine it.

I have a toothache. A wisdom toothache. My upper wisdom teeth seem to be completing their descent, and it’s painful. My limited inspection suggests that they’re approximately where they should be, so I don’t know what’s so painful about it. I guess I should see a dentist… but somehow I feel like I can predict what they will say.

Bad Behavior is Great

This Bad Behavior plugin is fantastic. For the last year, I’ve had to spend several minutes deleting spam messages every time I went to write a post here. Now 90 or 95% of spam messages are caught before I ever have to deal with them.

As long as it doesn’t screw anything else up, I am incredibly pleased.