The latest from Lascahobas

Currently at Lascahobas. I spent the first two days in Haiti visiting Silar's orphanage and getting our solar power equipment from different places in Port-au-Prince. I was mostly along for the ride - one of our teachers, Jeanide, led the way. We also visited a tablet manufacturer/assembler and talked to their head engineer, who talked me into buying a solar panel with integrated mobile backup battery. Yesterday Sora came into Port-au-Prince, then we traveled out into the countryside to visit the school in Lascahobas. Sora and the teachers discussed the project in Creole until after dark. I debugged some issues with internet, and went on the roof to check on the solar panels. I was then up past midnight writing some writing helpers into the iLoominate eBook project. Long day! Today is the first actual workshop day - our software is running on the server, teachers are meeting, Sora has ordered lunch for everyone. I'll be taking a look at the solar power system, with George and Tim joining by Skype

Going to Silar's and Lascahobas

Hey, this is Nick! Haven't been laptop-blogging in a while. I'm going back to Haiti tomorrow (the 14th). It's only for a week, but I get to see two awesome schools: Silar's and Lascahobas!

For the first couple of days I'll be at Silar's, inside Port-au-Prince, and after that I will be helping facilitate a teacher workshop in more rural Lascahobas (the upper right XO on the map). Along the way, I'm planning to meet a company making Android tablet in Haiti, help repair a solar power system, and map a couple of health centers.

About that tablet project...

One of the highlights of OLPC-SF was an update on OLPC's Ethiopian tablet experiment, directly from the engineers working on the project. This is an ambitious project where OLPC has taken great pains to be as hands-off as possible, while collecting meaningful data. They built resilient hardware in a village with perhaps no literate adults. They also developed and collected technology that speaks to kids ( taking inspiration from The Diamond Age, which was a fascinating read on my OLPC Uganda trip ).

Nicholas Negroponte's media circuit, though, doesn't sit right with me.

In an article reaching the front page of Reddit and Hacker News, we are told "OLPC workers dropped off closed boxes containing tablets... Within five days, they were using 47 apps per child, per day... Within five months, they had hacked Android." The press repeats these numbers, these timelines, these factoids.

Stories like this are setting up the project for failure. I'll limit myself to five points:

1) OLPC built a solar charging station and instructed children to charge the tablets daily. The project depends on this training. How much time was spent on construction and training? Would it work better with twice as much training? If so, the "dropping off tablets" idea is more stunt than strategy.

2) 47 apps / tablet / day is a silly number - an average over an unknown time. If months into the program, children still use their tablets like this, then they are overwhelmed with choices. Tell us: how long do they use content? Are they engaged? Do they get better at a literacy game over time?

"Children there had never previously seen printed materials, road signs, or even packaging that had words on them."

3) No way. OLPC's engineers chose the Xoom tablet because its charger could not be repurposed for villagers' and truckers' mobile phones. Yes: remote African villages do have motor vehicles, phones, radios, and Coca-Cola. OLPC has outdone themselves in finding a location far more remote than I traveled in Uganda, and I believe what OLPC has said about literacy and lack of education in the village. Even considering that, I would be astonished if children had never seen letters before.

4) "Hacking Android": OLPC set up a monitoring system, the children got around it by activating some mode or setting, and OLPC doesn't know how. It's a little funny, but not learning.

5) There's some conflation between this project's use of the Motorola Xoom and the proposed OLPC tablet. The OLPC tablets exist, their hardware is interesting, but the software is not touch-ready. I've always liked the idea of OLPC's custom software disrupting education from within, but there's a reason that the Ethiopia project chose Android.

Please... this is such a cool project. Tell us meaningful metrics, good and bad. Talk about the engineering. Talk to us about how you see this experiment leading to learning. Tell us about the choice to use Android, and stick with it. That's all.

Posts from Map Wiki

I was checking my server logs today and was surprised to see edits to an OLPC Map Wiki. For a short time in 2010-2011, I had an OLPC Map Wiki link (named "Our Maps") in the Map Activity. Markers were public by default. There was a private group option, but it was never used. If someone is editing the map now, they must have downloaded the activity a while ago, and never have updated. Some are blank or nonsensical, but a few since April 2011 are actual labels:

club atletico chilecito - Chilecito, La Rioja, Argentina

EE.UU. - Saskatchewan, Canada

Catedral - Parroquia Catedral, Tacuarembo, Uruguay

Uruguay Sudamerica Amca Mundo - Uruguay

iglecia nuestra mi casa - Pan de Azucar, Uruguay

mi casa (and a short distance away) amigo - Ometepe Island, Nicaragua

eslindo - Tacuarembo, Uruguay

es lindo - Portugal

un pais orible - Antarctica

On the dawn of Apple Maps

Apple is launching a Maps app tomorrow, supplanting Google as the standard map on millions of future iPhones and iPads. It occurred to me during dinner that tomorrow may be the biggest day in mapping since the launch of the iPhone, or even of Google Maps itself. Months or years from now, people will be asking me "can we add this to the iPhone map?" or maybe "should our app use Google or Apple?"

As an avid supporter of OpenStreetMap, I acknowledge their network of volunteers and applications, but there's no one day which defines them. It's also possible Apple will use OpenStreetMap data, as they did in iPhoto. If so, we can expect a wave of interest, contributions, and corporate data uploads to transform the site. Hopefully Apple has learned from the mistakes they made back in March:

Footpaths indistinguishable from roads on Apple Maps (Mar 2012)

I'm not worried for Google... in addition to Android, the web, and their API, they must have a plan to keep iPhone users on their system. But let's remember that Google Maps was one of the few apps you could have on the first iPhone in 2007-2008; Maps was one of its greatest selling points. Apple has decided not to let that continue any farther. Not because of iOS vs. Android, but because location is valuable, because (as I said a year ago) "Apple doesn't have a market for real world places". Apple is taking a bet that they understand mobile better than Google, and they can succeed there, like Instagram and Path taking on Facebook on mobile devices. I doubt they'll add their own reviews or check-ins or StreetViews... Instead they may surprise us with better location-aware apps, notifications, and smart caching for faster searches.

It's entirely possible that on day one, Apple Maps will look clunky and incomplete compared to Google. Apple failed to make a dent on social networking with Ping, and Nokia Maps never caught on in the US. But if Apple places itself in the Maps app, plus maps in iOS native apps, millions will be using it daily. And what happens if you aren't on that map?

Cities for Coders

At Code for America, a great deal of thought goes into what's possible with apps. But I miss the reflection which came with OLPC, the question of where programmers come from in the first place.

It's actually super-relevant to our work in cities, because not every city fosters a programming community. Most cities have a dedicated and hard-working IT department, but local coders provide the lifeline for new ideas and skills, an unofficial partner in apps and hackathons. Think of RAP Ceibal's work in Uruguay.

My city team's strategy focused on a local university. I made a few appearances at MUGTUG, a late-night CS meeting where students share projects centered around Google APIs. In the future, we hope to drive this group to develop a musical map of the city ( something technical, but not overwhelming ). But for the most part, CS students aren't taught languages of the web. It's so odd - I mean, the world is web. But two weeks ago, a Berkeley grad student told me that she'd quit Codecademy, because her professor wants her to focus on Python.

My own experience? I like learning while working - the start-up way. This year I have completely overhauled how I make websites. Even maps!

I'm also working on what might make Codecademy's JavaScript classes ring true for more people. You're invited to try Rainy Day Coder and learn JavaScript on rainy nights and weekends! You'll get mail at most twice in a week. You should get the right city automatically - if not, put in your zipcode.

Storytelling Workshop, Haiti

This week, OLPC school École Shalom was part of a multi-school workshop by Dr. Terri Bucci to improve schools' education methods and introduce new interactive lesson plans. One of their main focuses was storytelling. By reading out loud and giving the students individual copies to review, the teachers could spark discussions in their class. There were two interactive activities with each book: putting the pages back in order, and matching the text to the illustrations.



We met at this church:



During the read-aloud of one story about a mole and a hedgehog, one of the researchers asked if the teachers knew about the animals. They hadn't. A few explanations later, we were back on the tracks. It's a good reminder that some elements of a story might not translate the way we'd like ( another example: Khan Academy's math lessons with avocados ). But does a story need to be rewritten every time it crosses national borders? There was a superb point from another researcher that stories can be a tool for us find out about other places and their cultures. Hmmm.

I also met Dr. Kranch and a student who are developing a lesson-builder for the teachers and the schools' tech people ( like Junior and Lorinord here ). They took interest in the Fedora Linux / XS school server / HTML+JS stack we were using to deliver lessons and quizzes to our students. The HTML+JS side makes it compatible with computers they are introducing in a few other schools. I need to send them a write-up of how we run it, and what parts need the most streamlining.



Today I biked south to the point where the river disappears between two mountains, something I'd only seen in the distance in my last trip. Google Maps Link Heading that way, you reach sheer cliffs and a dam where the river is diverted into irrigation canals. Here and there you see people shoveling rocks into piles. Eventually I reached a point where I could go no further without walking in the river. Soon after I took the picture above, a mining truck appeared, driving down the middle of the river. No safe passage for bikes.