LONG POSTING: BusinessWorldIndia.com on Linux in India...

[Frederick Noronha <fred@xxxxxxxxxxxxxxx>]

Normally such a long posting may not be appreciated on the LIG list.
But kindly permit me to send in the same, in view of the relevance to
our the theme this group is working on....See the latest issue of
BusinessWorld (Dec 10) for a couple of boxed-items accompanying this
article.Frederick.
PS: The rather-embarassing references to me in the article credit me for
far more than I'm responsible for. Much of my campaigning has been
keyboard-based. Others have done the donkey work. Also, I am not a
'part-time' journalist, but rather a freelance one. There's a painful
difference between the two... with the latter taking me 13-16 hours of
work daily to earn a living exclusively from writing!  Maybe I could
arrogantly claim to be two full-time journalists...!

----------  Forwarded Message  ----------
Subject: BusinessWorldIndia.com
Date: Sat, 9 Dec 2000 23:07:26 +0530
From: Frederick Noronha <fred@xxxxxxxxxxxxxxx>

HAS LINUX WHAT IT TAKES? 
Charles Assisi 


It comes down to textbook economics. To become a real IT nation,
India needs a larger domestic market and a bigger resource pool. That
requires low-cost access to computers across the country. In primary
schools where the young first learn the basics; in colleges where minds
are moulded; and in the hinterlands where information comes at a
premium. Cost is the key here: high prices are one of the main reasons
why computer penetration in India is among the lowest in the world. 

The obvious solution is to make computer access affordable. But isn't
taking hardware and software to the masses mean considerable
investment? Not everybody thinks so. Not anymore. 

Take the case of Mexico. Government officials there are half-way
through installing 140,000 computers that run on Linux -- a free
operating system (OS) -- in primary schools across the country. They
reckon that with 20 users per machine, the project, when fully
implemented by end-2003, will educate 28 million children on how to use
the Internet and access the latest educational tools. These children
hold out the promise of making Mexico a computing superpower. 

The project, funded by the government and supported by the UNDP, is
being put into place on the back of old hardware 'incapable' of meeting
the demands 'new' software imposes. Incidentally, the Mexican
government saves over $700 million on software costs alone. The most
conservative estimates put savings on hardware at $150 million. 

The reason why the Mexican government supports Linux is easy to
pinpoint. In practice, 486s with just 740 MB of hard disk space and 32
MB of RAM can support Linux and meet the requirements of most home
users.  

If that doesn't quite excite users brought up on 'powerful' machines,
consider the New Internet Computer (NIC) being promoted in South
Africa. Larry Ellison, the chief of Oracle, was reportedly delighted
with the machine and called it the machine of the future. Designed to
run on Linux, it uses a 266 MHz microprocessor, has 64 MB of memory, a
56K in-built modem, a network card, 24x CD-ROM and a Super VGA screen.
Bung in two USB ports for external devices and it is every inch a power
configuration. But what makes the deal really attractive is that the
machine retails in the international markets at Rs 9,200 ($200) and
comes loaded with free desktop applications which otherwise cost a
bundle.    For Ajay Shah, professor of economics at the Indira Gandhi
Institute for Developmental Research (IGIDR), only the numbers count.
"That kind of economics," says Shah, "is the single most powerful
argument in favour of Linux. Damn everything else." His case is simple.
In India, an entry-level programmer earns $100 a month. Chances are,
the only access he has to a computer is at his workplace. Owning a
decent PC can cost at least Rs 35,000 ($750). But it comes with just
skeletal software. An office suite, for instance, is not included. The
choice is to either pay in excess of Rs 15,000 for a licensed version
or Rs 300 for a pirated copy. The latter is an uncomfortable option.
But at less than Rs 10,000 for a fully loaded machine that runs on
Linux, the potential it holds to attract millions of buyers is huge. 

The flaw in current thinking, points out Shah, is the focus on the
200 million existing computers, 80% of which run on Microsoft's OS.
Instead, it ought to be on selling the next billion. When it comes to
that, the $0 price-tag Linux carries is a critical product feature. 

What Shah's saying is that, in India, these are viable solutions.
Computing can actually be taken to the masses for less than what it
costs to buy a colour TV. What makes the situation so terribly piquant
is the fact that implementing these solutions has remained a mission
confined to a small band of people, fighting a system that chooses to
look the other way, even when economic realities seem to point
elsewhere.  

Ask Frederick Noronha. Based in Goa, the part-time journalist is an
evangelist for open source software like Linux. He figured the Mexican
experiment could be emulated in India if he could get second-hand
computers of the 386/486 generation. They could well be loaded with
Linux and distributed to schools which didn't have the resources to
afford computers. His thinking spurred a number of Goan NRIs to come
together. They managed the machines and put them on a container to
India. 

Unfortunately for Noronha and the schools these machines were meant
for, the computers still lie in the customs godowns because duties
can't be paid. At institutes where an outdated machine is a luxury, the
idea of paying import duties is not practical. The wranglings continue,
but the reasoning behind the project was economically sound. 

Second-hand machines can be had from generous corporates and
philanthropists anywhere in the world. It is the operating system --
Windows which runs 80% of these machines -- that is a problem. Simply
because for Microsoft the idea of giving things away for free is
unacceptable. The company insists that new users of second-hand
machines pay for a fresh licence. Given these realities, the choice for
people like Noronha is stark. But even as this choice is being made
across the world today, some questions need to be answered.  

Linux made its debut in 1991 on a single desktop -- that of its
creator, Linus Torvalds. Its success at the enterprise-level has been
well documented, but what about its efficacy at the desktop? Given the
stubborn resistance Linux developers have displayed in the past towards
a 'dumbing down' of the OS, the concerns are valid. 

Recent developments show that they have been addressed. Take a
classroom environment where word processors are required. One of the
options available today is LyX, a software that can be used to create
text documents. Then there is Corel WordPerfect 8, which offers pretty
much the same features that Word offers. Corel allows files to be saved
in a format that can be read across platforms, including Windows. 

As for spreadsheet applications, Microsoft Excel files can be
imported into Gnumeric, a part of the Gnome Desktop environment that
comes bundled with Red Hat, one of the more popular Linux
distributions. Star Office, an office suite developed by Sun, in spite
of its perceived shortcomings, is also a viable alternative. More
importantly, a user has a choice of 14-odd graphical user interfaces
(GUI) for the desktop environment. Two of them, Gnome and KDE, are very
similar to the Windows interface and equally easy to use. 

Linux is already the OS of choice at an advanced level. Open source
science software ranges from neural networks for artificial
intelligence, to solving n-body problems for physics, DNA simulations
and chemical compound visualisations for chemistry and biology.
Virtually all of India's premier educational institutes made the
migration to Linux a couple of years ago. This includes the IITs, the
IIMs, the Indian Institute of Science (IISc), the University of Pune,
the National Chemical Laboratories (NCL), BARC and the Tata Institute
of Fundamental Research. Ironically, most engineering colleges --
without access to the kind of resources the premier institutes have --
still work on expensive software that runs on the Windows platform.  

Limitations, however, exist. Nagarjuna G., a science educationist at
the Homi Bhabha Centre and president of the Linux Users Group (Mumbai),
points out that there are a number of older generation multimedia
applications that cannot run on Linux. While audio files can be dealt
with easily, it is still difficult to play a Video CD on the platform.
So, a Sesame Street which children love so much goes out of the window.

In the same breath it ought to be mentioned that Linux supports
multimedia CDs that are Java-enabled. More than that, similar
multimedia applications are supported on Internet browsers where the
power of Linux is unparalleled. And, given the fact that the Internet
is a source for rich content, such issues are being buried. 

And it is not just schools that stand to gain from Linux. A point the
Linux Users Groups are trying to prove in Kochi and Thiruvananthapuram.
The need for low-cost computing solutions propelled these groups into
lobbying actively with the Kerala government to set up computer nodes
at the panchayat level. The idea being to use the Net and establish a
communication channel between the people and the government. Pretty
much what N. Chandrababu Naidu is working at in Andhra Pradesh. The
difference lies in the economics of cheap hardware and free software.
While Kerala's efforts have met with a fair degree of success and
information trickling in indicates some acceptance on part of the
government, a lot of ground still remains to be covered. It could be
the first Indian state to adopt Linux at all levels.    In another
recent initiative, Chandigarh announced plans last month to launch a
pilot project in 12 villages, again, with Linux. Here too, the primary
driver has been economics. 

The economic rationale aside, there is another potent argument in
favour of open source Linux. It allows localisation. The Chinese,
Japanese, Koreans and Russians have used it to great effect. In India,
a bunch of Tamil programmers has built regional language tools, so much
so that KDE supports the language. The Centre for Computer Aided
Translation in Hyderabad, headed by Professor Syngal, is even now
working on a government-funded project to increase localisation. Its
applications will be supported across platforms. 

Check out the work being done by Mumbai-based FreeOS.com. Says CEO
Prakash Advani: "With a 5% English- speaking population in this
country, it would be foolish not to work on local languages." Very
recently, the company received funding from Boston Computers and has
set up offices in the US where, Advani says, the degree of awareness is
higher. FreeOS.com deploys solutions and provides support for
Linux-based applications. By the time the company is through with its
project on Indian languages, it will be possible to write programs and
create applications and interfaces in all major Indian languages. Sure,
Microsoft offers local language support too. But that comes at a price
and it doesn't work across systems, unlike open source. 

Yet, all this may not be enough. Illiteracy is rampant. Assuming for
a moment that computers are ubiquitous, how do you drive the masses to
take to computing? Colour TV got a shot in the arm when the Asiad was
hosted in 1982. Cable permeated the collective consciousness of the
country when Iraq attacked Kuwait and a nation rushed to wire up their
homes and watch the war on CNN. 

"Baby pictures," says Shah, "is a way out. Show the women that their
daughters can send them pictures of their grandchildren from wherever
they are on community machines. That is a bigger killer app than
email!" Shah says he doesn't mean to be facetious. People can be shown
that IT can be used for real... that it can impact their lives.
Provided there is cheap infrastructure available. 

But showing people how to send baby pictures and maintaining machines
that support such requirements needs trained personnel. Though it is
not difficult to use, right now there aren't too many people who
understand how an OS like Linux works. It is a very real hurdle and one
that Linux advocates acknowledge. Success, in getting around it, has
been limited -- largely because the Linux movement in India has been a
non-commercial one. 

Over the last couple of months, however, major Linux distributors --
Red Hat and SuSe -- have established their India offices. Red Hat's
entry into the country was followed by a tie-up with NIIT to produce
Red Hat-certified Linux engineers. SuSe is expected to follow suit and
make similar arrangements with a software training school. Prakash
Shetty, vice-president, SuSe (India), admits there certainly is a
problem with the number of trained people. 

At the same time, tie-ups of the kind Red Hat has entered into are
expected to generate interest and churn out graduates with knowledge of
the OS. According to Sunita Thomas, CTO at Infotech Financials, a
Mumbai-based company which deploys Linux solutions, most engineers in
her team came from a Microsoft background. Adapting to Linux and
providing solutions have been a non-issue at the organisation because
the migration is simple. 

Linux has the power of numbers and recent history on its side. From a
single desktop in 1991, it reached a stage where 16 million people were
tinkering with it by end-1998. Which means the OS occupied 2.5% of all
desktops in the world. The user base had doubled exponentially three
times each year, for eight years. Linux now commands between 6% and 7%
of the desktop market and remains one doubling away from breaching the
10% mark.     

And 10%, most analysts will agree, is when it becomes viable for
third-party software developers to jump on to the bandwagon. At those
levels, the market is far too attractive.  

The fact that the OS comes for free is a compelling driver even now.
Waiting for the market to take its course could mean delay. Which is
why the government has to step in -- because there is a danger in not
trying hard enough. India's stated objective of IT for the masses would
then remain a pipedream: its programmers consigned to writing millions
of lines in code to create modules that fit software designs
conceptualised in Silicon Valley.  

Look at it from another perspective. There is this search engine --
Google. Over the last one year, it has emerged as the engine of choice
for Net users. Not because it was built on the back of rocket science.
But because two students at Stanford University cooked up a set of
powerful algorithms with the power to search for content intelligently. 

It is the kind of algorithm that could have been written anywhere in
the world. That it wasn't, is testimony to a unique mindset which
exists only in pockets. The real task is to build that mindset in this
country. That is where Linux holds out a promise. And a challenge. Any
takers? 
-------------------------------------------------------