Is open source just for the poor?

Just not here.

Pradeep Tagare, a senior investment manager with Intel Capital  is quoted in the Indian Express saying that Intel will launch an "open source incubator" in Bangalore and fund 3-5 Linux start-ups there.  Tagare also told the LinuxAsia conference last month that open source can be a success in India, based on service contracts and the value provided by it.

Reading this I had two criticisms. While there is government support for open source in India, much of that support is predicated on the idea that open source is free, and in fact most of the Indian programmers I know use Windows. Besides, what about open source start-ups in this country?

Then I thought again.

Linux still faces the chicken-and-egg problem, in India and elsewhere. Without hardware that is in demand, there is no demand for software at any price. Intel is a hardware company. Most inexpensive PCs -- Linux or otherwise -- have flopped.

There is a good reason for this. When you sell something as a "PC for the poor," your design constraints become extreme. When the goal is mainly to keep costs down, you wind up with a machine that doesn't deal with the computing mainstream.

Maybe I'm just being a crank. There is something Tagare's fund might work on that could be very useful and profitable. India has 216 languages in all, 22 listed as official ones. Deliver software applications and training materials customized in each of these languages, designed to meet real needs, and you will really have something. Just in Bangalore there are language problems, with Tamil and Kannada speakers vying for political and economic power. (Tamil has most of the economic power, Kannada the political whip.)

India doesn't need cheap software or cheap hardware. It needs stuff that works, computers that talk to one another in a common language but to people in their own. That is the "killer app."