Went to hear Christopher Blizzard speak about One Laptop Per Child at the NY Linux User’s Group monthly meeting held at Google‘s offices in NYC. I don’t actually use Linux though I do use a lot of Open Source applications. DRM in Vista means I will probably be making the switch soon.
For those who don’t know the One Laptop Per Child project aims to address the growing gap between technological haves and have nots in developing countries.
Below is my summary of his talk. I admit I don’t necessarily understand the more technical aspects of it:
The first thing he pointed out it is a humanitarian effort. It is not getting done for the sake of invention (though they are doing neat stuff) and not for the purpose of selling the machines being developed. He showed an image of an inventory of supplies from a Nigerian school. 1 first aid kit, 1 lever (for science lessons) a handful of items – these are their tools to learn.
The countries they are specifically targetting for now are Nigeria, Brazil and Libya. The project was originally the brainchild of the MIT Media Lab, whos goal 25 years ago was to get laptops into the hands of normal people. Now their goal has shifted slightly – to get laptops into the hands of kids, cheaply. There are technological problems with doing so, the details of which are later in the post. First, a little more about the project itself
OLPC is now a separate organization from the Media Lab and employs 14 people fulltime. (They reside across the street.)
They have some huge partners on the project: UNDP (UN technology group) eBay, Google, AMD, Nortel, News Corp., Quanta Computers (Makers of 1/3 of computers made today which are then rebranded – it is an entire city in Taiwan.), SES Global, Redhat, Brightstar, CNI, MEI, Marvell.
Of special consideration is the environmet the laptop has to be delivered to. Issues include, occasional power, occasional connectivity, physical environment that can be very dusty and hot (hottest recorded temperature in Libya was 58 Celsius = 136.4 Fahrenheit, cost has to be manageable (100$ pricepoint goal).
Unique attributes of the machine related to these issues are that it can be used in direct sunlight, it has low power usage, connectivity is via a mesh – each machine is a router in addition to a client, rugged construction and low cost.
Display is 71/2 inches across. Can be used in color or grayscale, is low cost, is 1200×900 at 200 dpi. He explains that most displays are powered by halogen when backlit – when this goes out the display is gone. Also uses a lot of energy – 20 watts. Here they are using LEDs instead of halogens and can get full brightness at 1 watt. In reflective mode it is a tiny amount.
In terms of power consumption they have a goal of 2 watts during average use. Partly due to the unique display, also the machines suspends “early and often.” In response to a question from the audience he says it takes less than 100 miliseconds to suspend and 50 miliseconds to resume. It can handle different kinds of power where most laptops are sensitive and need to use a specific plug and source. Here you are in places where people may be carting around car batteries as a portable power source. It regulates “dirty” energy well; DC not AC. There will also be human power devices. They will better than the hand crank on earlier models, which Koffee Annan broke it off during a demonstration….oops! It will be something like a yoyo or chord on a lawnmower.
With regard to networking: it uses mesh networking, has an autonomous wireless chip, bunny ears (people underestimate the power of a good antenna) ad-hoc connectivity, UI for networking environment. While in the Australian outback someone got it to work at 1.3 kilometers away from another point, usual would be about 600 meters. In response to a question the network chip is from Marvell, it has a driver and its own OS and when it is off it will still forward packets for others. He is hoping they will release the code at some point.
The machine uses 800 miliwatts in suspend mode, It gets 10 hours per charge at normal usage.
With regard to the rugged construction: it has no moving parts, no fan, very few connectors, internal bumpers, rigid construction and it is sealed (dust issue) also there is a rubber keyboard which is actually more expensive so it can be washed. He hopes that one day they will be able to throw it in a dishwasher – have to solve issues of disipating heat and sealing even moreso.
In response to questions: It will have 802.11 style bandwith, it should be better than cellphones, would like each machine to have an individual ip address, but there is a shortage in ip addresses, there are some universities in India where the whole school shares an ip address. Goal: ipv6 internet.
Currently they have reached an approximate $150 price point for the first generation. It is a different economic model than with most technology. They don’t want to add features. They are happy to use older processers to drop the cost. It weighs about 4.7 pounds – most of which is the batteries.
With regard to issues involved in designing for kids: He admits they are doing something new and it may not work. They have a desktop that is very iconic since text is hard to internationalize. (There are different keyboards being built for different countries.)
Elements of Sugar (the OS). It has an activity model (there aren’t applications rather activities.) It has a zoom metaphor, presence model, frame and journal.
Zoom: kind of like in a game where you can see the map view or actual view. You can zoom in to your activity. A bit further out you see only you and the options you have. A bit further out you can see your friends. Further still you can see everyone represented on your screen and what activities they are a part of. There is chat built into all activities (uses XMPP chat protocol).
Frame: a rectangle border around the screen. Top has zoom features. Right side has representations of your friends (can hover over their icon and a name and picture will appear). Left has a clipboard. Bottom has activities you can take part in. All icons based.
Responding to audience: OLPC is part of a vision. It is not about one in a classroom or five per school. An entire country or region has to commit to one laptop per child.
Ages 5-6 can begin with it. The goal is to have the activies scale to different skill sets as the kids grow with the machine. For example TamTam is a music player with 3 versions. 1. player 2. track editor 3. synthesizer. It is an important concept having a low barrier for entry but no ceiling.
How hackable is it? Will 7 or 8 yr olds be taking it apart? It is all in Python and is simple to hack and add stuff. Can get at the nodes. Should be able to change the UI over time. It would be the same model as firefox extensions.
Squeek/Small Talk has an etoys distribution for it for drawing. A choice was made to use Python because it was quick to do. Some will be redone in C to make it smoother. Always with an eye to be accessible and open.
Have you considered working with poor communities in developed countries? Not yet. Some talk of working with aboriginal communities in Australia and Canada but it is not the first goal coming out of the gate. In the developed world you have to support everything – AIM, Windows, Office. We can affect millions of kids in the developing world if we stay focused. Libya has 1.2 billion million kids. [source for number] China and India were too large to start with.
The UI is it ASCII or Unicode? Have you considered making it a tablet since people are familiar with using pens? It uses Unicode, there is a different keyboard for each country. There are issues with languages. Spanish, Portugeuse, Arabic. Arabic is hard because of combining characters. Thai is nearly impossible – there are no spaces between words, you need a dictionary for linebreaking. But there is a good team working on the languages end. It has a dual mode touch pad on this demo but it has electrical problems so they are not sure if they are keeping it. 1.5 or 2.0 should have a touchscreen.
Tech support? Kids should be ale to do it themselves or just replace the machine at $150. Designed so it can be fixed but at that pricepoint you may not need to.
How do things like files, moving things, names work? You save things in your journal. It is time-based like a weblog. You cans end items in your journal, make them public, can add metadata so you can describe and search through it. Top of frame also has a search bar. Learning for kids is about creation and exploration. Journal replaces a file system. Visibility of an item helps remind how current it is. There is a 1/2 gig of storage. The model is one of working with a full hard drive at all times where older items naturally fall off. It has 3 USB ports so a USB key (pendrive) may make the journal bigger.
Could you boot another OS off the key? Yes. Has run Eudora Fedora on it in the right configuration.
Zoom metaphor works with the search too.
Activity bundles. They are self-contained programs. Easy to give to someone else. Would contain their own dependency files. (with Debian packages can have dependency issues, needing to download tons of other stuff to get something to work) With limited storage have to consider what you are adding. Model is of a toolbar. You chould be able to drag it and give it. It is free so you wouldn’t have licensing problems.
How do you build Sugar? It is not well-documented yet but there was a recent article in Redhat magazine.
In short it is about learning. Encourages sharing (people work in groups on activities, share their journals, extensions), creation and exploration. It is easy for people who haven’t used computers but can grow with the person.
How to get involved: wiki.laptop.org, Sugar mailing list + HIG (Human Interface Guidelines), read the article in Redhat magazine, write an activity, Development Mailing List, Fedora (it is based on Fedora and constantly trying to fix dependency issues)
Isn’t this ambitious? First thought: Crazy, insane, may be throwing away my career, then Holy crap this is gonna work.
There are about 12-14 fulltime employees with an army of programmers around it. There are 6 Redhat people and 150 Quanta people. We are small, not corporate, focused on this humanitarian nonprofit venture. Vista is for a monopoly in business and entertainment users.
Will it be used by adults too? In Cambodia, Nicholas has seen laptops used as a primary light source! Access to information will be there for adults too.
Can you switch to console mode? Yes, but there should be so many activities so you never get bored.
Has it been tested with kids since 10 or 15 years ago Apple did a test with kids with a mouse and the GUI (graphic user interface) and most couldn’t figure it out. Beginning to do some testing in Brazil. Also, laptops will be dropped off with a person who will train local people for about a month about the mouse, frame, zoom and interaction. Maybe it will work maybe it won’t but trying. It is about the kids and the organization learning.
How many are projected to go out? 3-4 Million in the first year. 50-100 million in the next year. For perspective 40 million laptops were made last year; 1 billion cellphones.
The laptops are darned cute. It was really fun to see all the adults in the room go up to play with them and trying to figure out how to use the camera in it – they have cameras too!
I asked Christopher if he was concerned about laptops being stolen and sold on a black market. He said they are working on implementing solutions such as having the laptops go dead if they are away from the school base for too long.
He was a great speaker and a nice guy. He was happy to answer questions in the middle of the presentation and wasn’t thrown off. It went a little over time. The NYLUG then went to a bar – I went home because I hadn’t really planned on being out. It was really interesting though and I’m glad I went. I was worried that it would be too techy for me to follow. It made me wish I could program…we’ll see how long that lasts – I’ll probably just ask Kevin to do stuff for me and he’ll say no :-) .
One of the words Christopher used to describe working on the project was creepy… I asked if he meant because it was such a large idea that can have an impact? He said not exactly. Maybe it has to do with an earlier comment he made about being a bunch of geeky tech guys who have no notion of the environment these machines will go into on a personal level. [Update: His answer here.]
On a different note it was a geekthrill to be at Google. They provided some nice food and drinks (including alcohol served by people in suits and foody uniforms.). Even the woman who greeted people at the door and showed people where the restrooms are had a big smile on her face. They must be a good employer…
This is an amazing idea. And the technological advances will benefit everyone (I know that’s not the goal) but we all want a screen that would work in direct sunlight and use less power no?
I think it makes sense to do twinning communities or children with the OLPC project. A wealthier community could buy one for the price of two and then those children in the communities would have a connection and ongoing contact. This would be like Jewish Federations twin cities or how in the 80′s people had a bar mitzvah with someone in the USSR who couldn’t have one. Another thought could be getting people to buy an OLPC laptop when they buy a computer for themselves or their kids. Maybe some of the corporate sponsors of the projects could offer credit for doing so.
This is another one of those moments where I think – this is the future – I want it now. :-D
