OLPC – where the money meets the road
I have not written about the OLPC project for many months. The reason is there has been little to write about. The hardware has not changed. It is still amazing. The software has not changed (other than a number of development releases) and still is missing critical function and numerous fundamental bugs. Most of the G1G1 community has been tinkering but not too much “usable” progress has happened beyond the initial quick gains.
The reason is clear – resources. It takes a lot of developers and testers to write good software. The OLPC uses a custom version of Fedora. The “custom” part of that is the issue. There were significant design decisions made that require significant development to accomplish. The “journal” storage mechanism, the Sugar interface, the chipset power management features, and the rotating screen.
In a “for profit” environment, there would be a cost/revenue analysis. The XO hardware warrants the investment but OLPC is not a “for profit” organization so there is no “revenue” in the equation since the organization is avoiding passing the development costs along to the consumer. While this is a laudable goal, it has created shackles that have jailed progress.
Microsoft has more than a 1000 times the resources available to apply to these types of challenges. It also has the money to “spend now and reap rewards later”. This is exactly what they have done. To their credit, they appear to have done a pretty good job. They made business centric choices. They developed support for the chipset power management and the screen but did not consider the “journal” or the Sugar interface. The resulting lean Windows XP doesn’t fit on the internal 1GB solid state drive so it goes on an SD card in the one available slot.
The video is compelling. I have two XOs and if given the chance, I will run one with XP.
Note: I shouldn’t detract from the topic of this post but I can’t avoid the realization that when “XP4XO” is released, there will definitely be a flood of hacking of drivers to make it a generally usable on many different hardware.
Just yesterday I was walking past a realty office near Harvard Square and they had an OLPC connected to a video camera that displayed the passers-by. I didn’t have time to stop in and ask them why they had one.
About a month ago I went to a talk on ethics and development and one of the speakers suggested that development projects should be structured more closely to for-profit projects – they should employ viable business models and be self-funding. “No country was ever lifted out of poverty through donations,” was his thesis. The book is on my shelf at work, but I can post the name and author later today.
Without sounding like a press release, Microsoft has launched a significant program focused on this notion that “for profit business models” is the only viable solution. It all comes back to “give a man a fish and he eats for a day. teach a man to fish and you feed him for a lifetime.” In it’s “Unlimited Potential”, Microsoft is attempting to look long term (novel thought) where investment in education and creating entrepreneurs will result in growing businesses and commerce in the future – “enable sustained social and economic opportunity”.