Latest news on the OLPC

This, from Slashdot:

Yesterday Nicholas Negroponte, (…) current head of the nonprofit One Laptop Per Child project, gave analysts and journalists an update on the OLPC project. Two big changes were announced — the $100 OLPC is now the $175 OLPC, and it will be able to run Windows.

Toronto Democamp 13

Last night was Democamp night –the thirteenth in Toronto, to be exact, and a nice improvement over some of the previous iterations.

To start, Mike Beltzner (User experience lead from Mozilla) and Betsy Weber (from TechSmith) gave an invited 15-minute talk each. Mike Beltzner’s comments about managing a large open source project were insightful, and reminded me of Fogel’s book Producing Open Source Software. (Mike also has the best CV I have ever seen.)

We then had a bunch of short (5 minute!) demos. I lost count of how many (David Crow says 11, including Betsy’s), and though the quality was uneven, there were some pretty good ones: Walkah’s OpenID presentation and myhood.ca were my favourites among several others worthy of attention. OpenID is picking up speed and we’ll be hearing much more of it soon; myhood.ca is a nice integration of Google Maps with rental information in the same tune as JobLoft, a previous Democamp presentation.

On the negative: No Regrets, the venue, was uncomfortably packed, and the number of presentations dragged for a bit too long. But overall I enjoyed the new format pretty much. David Crow’s post throws some cool ideas on how to keep improving the format –I’m looking forward to seeing them in action.

Great lip-sync video

I just have to share this music video. Starts slow, but give it a minute. I love it when people do this sort of things.

(WordPress won’t let me embed Vimeo videos here, so the link will have to make do).

Old Toronto map

Being an immigrant can be an alienating experience: you come to a strange city and you just take it for granted –you’re there but you could be anywhere, the city’s history is a blur, it’s just a place where you sleep, eat, and work, it’s a space that simply is not home.

Making it home is laborious, but fascinating if you’re up to it. I’ve been in Toronto for more than 3.5 years, and I keep discovering corners and stories that make me appreciate it all the better.

Through the wonders of ebay, I got myself an old atlas map of Toronto –from 1898. That makes it 109 years old. I think it’s amazing:

I’m far from finishing exploring it, but I thought I’d share a few things that I’ve found.

I was intrigued by this piece of the map. Where we now have the Old City Hall, the map says it’s the “Site of New C.H. (Courthouse?) & City Hall”. Apparently construction of the now Old City Hall was finished in 1899 – one year after this map was printed.

But if back then that was the New City Hall, where was the old one? The answer really surprised me, like the murderer in a mystery novel: Why, but if it’s old and yummy St. Lawrence Market!

Wikipedia confirms it: “It was established in the early part of the city’s history and (was) home to Toronto’s first permanent city hall and jail house from 1845 to 1899. Designed by Henry Bowler Lane, the first floor was formerly Police Station # 1.”

Why did nobody tell me?!

Two more findings from the map, and I’m gone. First, the corner of my current apartment building (Charles and Balmuto) didn’t exist yet: Charles ran from Yonge eastwards, not to both east and west; it’s continuation from Balmuto westward to Queen’s Park was called Czar St. If you’re around the area, you can now spot that Charles was two different streets, since the line from Czar St. ends a bit to the north of Charles:

Finally, of course there was no subway in 1898 Toronto. But there was public transportation. The dotted lines you see in the previous three maps represent horse railroads:

I feel a bit ignorant, but I don’t think I even knew these things existed:

They ran on Yonge, King, Queen, College, and other streets. Great stuff. Back to study the map…

—

Photo of St. Lawrence Market by SimonP. The horse car picture is not from Toronto, but from Stamford, Connecticut.

A friendly criticism of the One Laptop Per Child project

My family’s first computer was a TRS-80 Color Computer 2. They didn’t sell them back in Mexico –there were no Radio Shacks, the import taxes were prohibitive, and since few people there knew of its existence, you couldn’t find it in the black market either. If you wanted one you had to smuggle it from the U.S. yourself. My father bought one in 1984, I believe, for what surely was a significant amount of money for our family. I must have been 7.

I thought it was one of the coolest things ever. You’d plug it to the TV, which acted as its monitor, and when you turned it on you’d get a BASIC editor to work in. With the help of my father, I learned what “LET”, “IF… THEN”, and “GOTO” meant. While he and my uncle programmed the computer to serve as a rudimentary PowerPoint, or to guess which person a user was thinking of (which was great for family parties), I started writing simple programs that I thought were amazing: for example, you could tell the machine as many numbers as you wanted and it would tell you their sum, or their product, and it knew how because I taught it! I guess I was learning to reason in (garbled) English too: this was nothing like the boring vocabulary lists of my Elementary school English lessons –this was actually useful! I was soon writing my own games and “Choose Your Own Adventure” stories in it, or picking new tricks from some programming-for-kids book.

Other than an unhealthy habit for programming with GOTOs (which I found really hard to kick when I finally learned Pascal), the computer’s influence in me was extremely positive. I believe I wouldn’t be pursuing a Ph.D. in Computer Science now if I hadn’t played (and enjoyed playing) so much with it.

But I’m now aware I was very lucky, considering my context. I was able to go down this path because I had this resource, the computer, that back then was extremely scarce in Mexico. Poverty or lack of proper mentoring closed this door to many other kids.

This inequality -this unfairness- is still widespread across the world, especially in developing countries. Far too many kids face huge obstacles to exploit their potential, to learn and hone their skills in tasks they love passionately. I’m not just talking about programming –art, publishing, design, mathematics, and many other disciplines, grew to depend heavily on computers. And it’s not just about having access to a computer anymore –it’s about having access to the Internet, and to all the communities it harbors. If you don’t have a computer, you’re essentially cut out. And there was little one could do about it, until now.

The One Laptop per Child project intends to do just what its name suggests. In case you haven’t heard of it: OLPC is a non-profit organization formed by very smart and good willed people that found out how to manufacture decent-enough laptops for roughly $100 US, and that wants to put one in the hands of every child. In case you have heard of it, you may be interested to know that the silly power crank is gone. It looks pretty fun, actually:

Recently, Mike Fletcher came to the University of Toronto to talk about the project, and to ask for help from interested developers. It really is a fascinating project, both in its scope and in its technical challenges. It’s also picking up steam, as there are already several countries signed up for the project –this is happening for real. If the project succeeds, millions of kids will have an excellent opportunity to learn and have fun learning with this computer, and without any financial stress to their families.

(If you have the time and the will, consider helping out. The project needs you, especially if you are a software developer. Let Mike know.)

Now, since I’m going to be critical of the project, I want to state this beforehand: I think the OLPC project has a lofty goal, and I wish it an absolute success. But I also see how this is a major endeavor, and it’s really important to get it right. As it stands, there are some issues about the project that bother me, and I need to put them out here in the hope that this will somehow help steer the project in the right direction. Here they go:

It’s not a gift, it’s a laptop/textbooks trade-off: Even at $100US per laptop, giving one to every kid in a country requires a substantial amount of money. Where will it come from? From the education budget of participating countries. But since participating countries are usually cash-strapped, the money will go to the laptops instead of going somewhere else –and if I got my facts straight, that “somewhere else” is going to be textbooks. That is, children will get a laptop instead of getting five years’ worth of textbooks. It’s not that they won’t get their textbooks’ materials (which will be stored in their laptops), just the actual dead-tree books. At least in theory, children won’t lose anything. But, as I’ll explain, it’s important to keep the trade-off in mind.

It’s a rather expensive textbook to misplace, destroy, or steal: The first ugly implication of the laptop/textbook trade-off comes with the laptop’s relative value against any single textbook. A child misplacing his laptop will result in either a high financial stress for his family (as I highly doubt the government will pay for a kid’s laptop twice), or in no laptop (and hence no textbooks) for at least the rest of his school years. And I find it very hard to imagine a brainy kid in a Mexican slum successfully protecting her highly visible laptop (she needs to bring it school every single day) from bullies and thieves for the entire span of her school years. One careless moment is all it takes.

It’s a bulky textbook to work with: This one is minor, but worth considering. For many learning tasks, nothing yet comes close to paper and pencil. Not even TabletPCs beat paper in my opinion; it’s doubtful that a device with less capabilities will.

It will be big in the black market: I predict that as the shipments of $100 laptops increase, their presence in local black markets will blossom, at a reduced price and with patches to circumvent security mechanisms. Most laptops on sale there will be thefts, of course. For parents in urgent need of replacing a misplaced laptop, this will be their only alternative ($100US is more than a month’s salary for most people), and the whole activity will generate a perverse cycle of theft.

It will garner many enemies: The laptop is designed to be the kid’s own domain –he’ll be able to do with it as he pleases, without adult intervention. Its designers seem proud of that. Now, I don’t have a problem with this –in fact I would have loved such freedom as a kid myself. But several groups will have a problem, and they’ll make sure we listen to them. In conservative societies, and in a few liberal ones, the idea that kids have a stash of pornography in their laptops, and that parents can’t do anything about that, will be enough to propel them into swift action against their Education ministry and the whole OLPC project. Teachers may react negatively too, as those that are computationally illiterate are pushed to the sidelines and those that remain feel their control over kids waning. Expect the laptops to be called tools of imperialistic control by many academics, and some governments retreating under the pressure.

It’s a case of mismatched objectives: There is a common criticism thrown to the OLPC: You want to give laptops to kids that really need food or shelter. As the OLPC wiki responds, this criticism reflects an ignorance of the conditions of many developing countries, which have enough food and shelter, but not enough learning opportunities. Unfortunately, the software developers with the OLPC seem to have made an assumption as mistaken as that of the project’s critics: that what children in developing countries need is what our geeky selves would want if we were kids again. Instead of striving to design the best educational tool possible (and, remember, the best textbook substitute), they want to design a kid-hacker’s dream: Browseable and modifiable code (one should be able to see the code that runs any part of any application easily), private access (your laptop is your temple), extensibility. The software design seems to come from the geek in us, not from the pedagogue in us. Why would an average 6-year-old be interested in such features is beyond me.

I can’t help but think of “The Nightmare before Christmas“. If you haven’t seen the movie: The inhabitants of the Town of Halloween discover Christmastown and decide they want to organize Christmas too. But their own nature conspires against them and they get it all wrong: their carols sound somber, their gifts are spooky –definitely not what the children were expecting on Christmas Eve!

This is, in my opinion, the most critical of all the issues. It is essential that the developers understand the real priorities of their project:

• First and foremost, they’re building a textbook substitute. If the laptops excel at something, this should be it.
• Second, they’re building a fun pedagogical tool. It needs to help kids navigate through their educational material in an accessible and inviting manner. It must help them discover math, biology, music, let them experiment, intervene when necessary. (Some software activities in the OLPC try to do this, by the way.)
• And in a far-behind third should come the hacker’s dreams: open source, modifiable, and extensible code, privacy, and such. In my opinion, the pilot programs shouldn’t even be entertaining these requirements yet.

Again, I wish the project absolute success. I hope the problems I listed turn out to be non-issues after all. And a lot of good will come out of the project, even if it fails. It has that “putting a man on the moon” flavor: the hardware needs to survive inclement weather and consume minimal power, the software needs to be extremely efficient and accessible, it needs to cost $100 US, it shouldn’t depend on maintenance, and so on. The whole project needs to resort to wildly out-of-the-box strategies (my favorite: carrying electronic messages by bike courier!). But since it’s worth doing (and worth investing very significant sums of developing countries’ money), it’s worth doing right. Here’s to its successful outcome.

—

Photo credits, in order: Wikipedia, Greg Lapouchnian, Full Metal Toys.

Virtual City Toronto

This web application may become quite cool given enough time. Enter an address, or ask for directions as you would to Google Maps, and you get photographs of the area or route of your result. The interface of the beta is still clunky, but the principle is exciting: It’s far easier to find out how to get somewhere if you have an idea of how the place looks.

It only works for Toronto so far. Found via Spacing.

Buildings shift like clouds

“What is suggested here is a new dimension—architects maturing from being just artists of space to artists of time.”

–Stewart Brand, “How Buildings Learn”

We think of buildings as some of our most immutable creations. Throughout their lifespan, their basic shape, their room distribution, and their purpose, remains mostly constant over time. They may last for centuries if built properly. Now and then they suffer some additions or renovations –an extra room here, a wider door there–, but these once-in-a-while changes do not detract from our perception of buildings as solid, permanent, inert structures.

This is because we see life with a short-term, day-to-day lens. Just like we fail to perceive the cumulative growth of a child we see often, we miss the slow changes in the world around us. Not even the constellations stay put, but you won’t notice unless you pay close attention to them. What happens if we cast a long-term view on our seemingly immutable creations –on our buildings? Do they change? How? Why?

This is what Stewart Brand sets out to study in “How Buildings Learn”, an eye-opener of a book that shocks the reader with its detailed comparisons of archival photographies and the reasons behind many building modifications. His long-term view transforms buildings into quasi-living entities that learn from their inhabitants and constantly evolve to suit their needs. Looking at some of the book’s photographs of the same building throughout the decades is like watching the ever-shifting clouds in the sky: You can tell it is the same cloud you saw a minute ago, it feels the same, and yet you can barely recognize it now.

Brand charges against most of modern architecture, which likes to produce disposable buildings to be torn down in about 30 years. He accuses architects of obsessing with striking looks and neglecting the needs of their tenants, most of which hate the buildings they must occupy.

Instead of focusing on looks, Brand proposes to focus on functionality, and on giving a building enough flexibility to adapt to the needs of its users easily. Buildings will change, he says, so let’s design them so as to facilitate their changes. And he does not stop there: he wants to modify not just buildings, but the way they are designed: Greater customer involvement, iterative development, scenario-based analysis, documentation for users.

It’s a very enjoyable book, and relevant not only to architecture but to any design field. As an Amazon commenter pointed out, simply substituting “building” for “software” in the text gives you an excellent essay on software development. I can imagine it offering the same benefits for many other professions.