Saturday, October 14, 2006

How I stumbled upon open source

If this reads like an article, that's because it is one. I wrote this for Digit magazine.

Venky

Sometimes, one good question is enough to change our lives. For me that
one question came like a bolt from the blue in January 1997, when I
interviewed a distinguished professor, Ken Keniston, Andrew W. Mellon
Professor of Human Development in the Program in Science, Technology,
and Society at Massachusetts Institute of Technology.

“Why is it that you produce software for the rest of the world, but very
little for your own country?” Prof. Keniston asked.

The software exports business was in a gung-ho mood, fears of the Y2K
bug was driving business to Indian companies and the domestic IT
industry was a poor cousin of the dollar-earning software exports
business.

Back then, I was a journalist and was trained to ask questions so I
asked the good professor to explain what he meant.

Despite his age and seniority, Professor Keniston was not one of those
Americans who sticks to the safe harbor of five star hotels in Indian
cities. He had systematically visited several e-government projects in
rural Madhya Pradesh, Tamil Nadu and other places and tracked their
progress over the years. A social psychologist by training, Professor
Keniston has a keen interest in seeing how rural India uses computers,
email, and the web to promote development, political transparency, and
social justice.

“Your country is very strange. All software is available only in
English, an alien language spoken by a small percentage of your country.
There is very little software available in Indian languages like Hindi
or Tamil,” replied Prof. Keniston.

My first reaction (like most others, as I subsequently found out) was,
“Why should people who speak Hindi and other Indian languages, need
computers?” To my mind, computers were an urbanized, westernized
phenomenon and the idea of rural non-English speaking people using
computers was an alien thought.

The interview was published and we kept in touch over the Internet,
which was a new media in India at that time. I kept sending Prof.
Keniston news items on language computing that I found on the Internet
and gradually developed an interest in this area. A few months later, on
the front page of my news paper, I read about Harsh Kumar, the General
Manager of Konkan Railways who had developed Indian language fonts that
he was giving away free. Manu Parpia who was then President of
Manufacturers Association of Information Technology (MAIT) liked the
idea. Manu and Vinnie Mehta, Executive Director of MAIT helped us set up
a stall at the IT India/Comdex 97 event in Delhi in December 1997 where
we launched these fonts under the BharatBhasha initiative. That was an
experience that truly brought home the need that Prof. Keniston had so
astutely perceived.

At the event, Harsh and I were giving demos of the Indian language fonts
and those interested could download the fonts from the Internet. But, as
John Lennon said, “Life is what happens to you when you are busy making
other plans.” For many of the people visiting our stall, this was the
first time they had seen a computer working in Indian languages and they
were blown away by what they had seen. They wanted to know how they
could use their own computers in Indian languages and they wanted the
software NOW.

From bearded sadhus to sarkari babus, they all wanted a copy of the
fonts immediately. Most of them went to an adjacent hall in Pragati
Maidan (if you have been there, you know how big a place that is),
bought a whole box of floppies, gave us one and asked us to copy our
fonts onto it. Over the four days of the exhibition, we gave away
thousands of copies of the fonts and got tremendous appreciation for our
efforts at democratizing technology.

When Keniston came back to India in early 1998, he gave me a
recommendation to the Knight Science Journalism Fellowship at MIT and I
spent 1998-99 traipsing the corridors of MIT and Harvard looking for
appropriate, affordable technologies that could be brought back to
India. Most of the professors and students there seemed to be using a
geeky operating system called Linux. Of all the cool, fantastic,
futuristic technologies that I saw, the one thing that really stood out
as being beneficial to India was the collaborative model of open source
software development that lead to the development of programs like
Linux. The idea that sophisticated software programs can be developed by
people collaborating over the Internet and distributed for free was
highly exciting because proprietary software programs were (and, alas,
still are) exorbitantly priced for developing countries like India.

After returning to India, I joined the newly set up IIIT-Bangalore. The
remarkable Prof. SS Sadagopan, one of the most entrepreneurial
professors in the Indian academic firmament encouraged me to continue my
research on bridging the digital divide. It was at IIIT-B that I wrote
an article titled “Why Linux Makes Sense for India,” that appeared on
the popular geek web site, Slashdot. The thrust of the article was that
if the collaborative model of open source was leveraged to localize
Linux and other open source software to Indian languages, it could spark
off a grassroots revolution and truly take IT to the masses. The upshot
of this article was that Prakash Advani, who was setting up a company
called FreeOS, came forward to fund this effort. This lead to the
creation of IndLinux.org, a non-profit that localized Linux to Hindi and
worked with different language groups across the country. We hired a
talented young programmer, G Karunakar, who built such expertise in
localization that he was invited to Iran, Bhutan and Nepal to help with
their localization efforts. The noted Internet entrepreneur, Rajesh Jain stepped in to support IndLinux.org when FreeOS.com was not able to fund the effort anymore.

Looking back, working with Linux seemed very logical because our
objective was to take computing to the masses. The General Public
License, one of the most remarkable documents in the IT industry,
ensured that the work we did would be freely available to the community.
There was no point in even considering localizing a proprietary
operating system because the benefits of that would flow to the company
and not the community. [Even if we did want to localize the proprietary
operating system, who would entertain a bunch of geeks like us?]

Today, Linux and open source software like the Open Office word
processing suite are available in all major Indian languages thanks to
the work of dedicated volunteers across the country and support from
government organizations like CDAC and the TDIL group at Department of
Information Technology. Red Hat has incorporated five Indian languages—
Hindi, Tamil, Gujarati, Punjabi and Bangla—in its flagship software, Red
Hat Enterprise Linux v4 and eight more languages will be supported. The
localized language interfaces are now being deployed in panchayati raj,
rural development, education and other e-government deployments and
helping take IT to the masses. Within the next five years, we will see a
profusion of web sites, search engines, blogs, social networking sites—
in other words, an entire ecosystem—evolve around Indian language
computing.

Looking back, I am struck by the irony of it all. It took an American
coming from halfway across the world to open my eyes to the need for
computing in Indian languages! Such is life.

This work is released under the Creative Commons Attribution Non-commercial Share Alike license and can be freely reproduced. See www.creativecommons.org

1 comment:

Hari said...

I don't work on Linux or on Indian languages ,
but I appreciate what you have done!

good stuff!
Hari