03 August 2007

One Laptop Per Child - the most important development in computers for a decade?

The recent decision of the One Laptop Per Child (OLPC) project to go into full production for the first time marks a watershed in computer usage in the home. For the best part of two decades, home computers have been dominated by one company - Microsoft - and although many incremental improvements have happened during that time there hasn't really been a seismic shift in the approach to computers since the rise of the internet. I believe that the OLPC project, if it succeeds in its aims, will mark a new era of computer use. Let me outline the reasons why I believe this is the case.

1. It brings computers to a completely new section of society
The intended recipients of the OLPC are schoolchildren, mainly from poor nations, most of whom will never have had the opportunity to use a computer before. Most of these will benefit greatly from their exposure to modern technology and what it has to offer in terms of communication, learning and an ability to see what's happening in the wider world. A small proportion of these children will grow up to have successful careers in computing and related technologies, which as well as empowering these individuals is likely to invigorate the economies in which they operate. This simply would not be possible if they never had the opportunity to learn in this way. A much larger proportion will be empowered to go on to careers where computers are an essential tool of the job rather than an end in itself - virtually all office-based jobs involve computers, and nearly every company can benefit from crucial functions (accounting, stock-keeping etc.) being computerised. Whilst these functions already exist, the OLPC project provides a means of bringing these skills to people who otherwise would not have this opportunity.

2. It brings a new, fresh approach to the desktop
The desktop paradigm has changed very little on the last 10 years. People are familiar with icons representing files or applications and possibly 'folders' (directories) but the approach of the Sugar interface used by OLPC is more human-orientated - icons represent other individuals in the user's sphere of acquaintance who are online. The tools it provides are geared very much towards learning, so for example many applications have a "View Source" capability to facilitate this. The 'source' might be the source code for an application or it might be the raw HTML of a web page for example, but it encourages children to look beneath the surface and investigate. Many of these children will develop their own enhancements for their own purposes, and share these with their neighbours.

3. The hardware design spec is radically different from the current 'norm'
For as long as I have been on this planet (and even before), Moore's law has held true. There has been an incessant push for hardware to become more and more powerful which shows no sign of abating (although it's clear that the method by which this is accomplished is changing now that multi-core processors are the norm). This has led to a large number of benefits for users. However computers have stayed resolutely in the office, or at least with laptops in a clean, dry environment where mains electricity is available. The OLPC breaks away from this totally - it is designed to be rugged and used where mains electricity might not be available or at least not in every home. The display has been specially developed to be usable in full sunlight, an important consideration for children who receive their schooling in the open air. The focus is on power saving and maximising battery life - a simple processor is more than powerful enough to run desktop applications.

4. It will mark the introduction of mesh networking to the mass market.
Mesh networking is not a completely new technology, but in the wealthy parts of the world the wide availability of broadband internet at affordable prices means that there just isn't a huge requirement for using it. However in rural locations in developing nations this is not the case. The ability for a single internet connection to be shared among potentially hundreds of users makes the difference between having and not having access to the internet, and all the services and benefits that come with it. Mesh networking enables the reach of the internet to be massively increased without the need for a large amount of expensive infrastructure to be installed. It may not represent a totally reliable, always available internet at any particular location but this is less important as the children will still be able to research and communicate at least some of the time. Their eyes will therefore be opened to the wider world.


5. By increasing production volumes of lower-end components, it will bring the entry-level cost down thus reducing the barriers to adoption of home computers around the world.
By reaching to a sector of the market that is, frankly, huge - economies of scale should mean that the cost comes down. The aim is for the laptop to have a bill of materials of less that USD100. Currently it is more expensive than this but over time it's realistic to expect this to come down. However this will also have knock-on effects for other products aimed at different markets. Ultimately, every market sector should benefit from this.

6. It actively promotes open learning.
OLPC is built upon a version of Red Hat's GNU/Linux distribution, which can be freely downloaded by anyone and shared out. The fact that every part of it is Free Software is vitally important here - no one company can lock in the users to their products (in the way that Microsoft has done very succesfully, albeit illegally, with Windows and related products). Red Hat is just one of a large number of organisations developing GNU/Linux-based offerings. However due to the innovative licensing of these products, enhancements developed by one organisation can be used by any other organisation. The net result is a bit like a mutually beneficial cooperative. My personal favourite GNU/Linux distribution is Kubuntu (a version of Ubuntu using the KDE desktop rather than Gnome which is standard in the default ubuntu). However the vast majority of the software is common between Ubuntu and Red Hat (and indeed other distributions). Each distribution takes what it believes to be the best applications, makes them even better then makes money by providing support services. Expect to see other distributions using variants of the Sugar interface - although it was developed specifically for OLPC, other projects can tweak and optimise it for other, different markets. That's a win-win situation, both for the consumer and for the individual developers. Some of the enhancements made by others will be adopted by the OLPC project itself in future versions.