Why Linux Makes Sense for India

"The localisation of Linux to Indian languages can spark off a revolution that reaches down to the grassroots levels of the country," writes Prof. Venkatesh Hariharan. Read the rest of his informative essay Below .

Why Linux Makes Sense for India

Falling costs have made computers more affordable to a larger section of India's population. At the same time, the Internet has made the PC a compelling proposition for fulfilling communications, education, entertainment and information needs. Based on these two trends, the market for Information and Communication Technologies (ICT) is likely to take off significantly in India.

Yet, India faces a peculiar problem in that almost all popular operating systems and applications packages are available only in English, a language which is spoken by a mere ten percent of the population. The lack of "Indianized" software is therefore an issue that seriously hampers the growth of the Indian computer industry. For almost 915 million Indians, the lack of Indian language interfaces is one among many issues that hamper their ability to reap the benefits of information technology. This is creating a new class of people who live in what can be called as "Information Poverty" even as technology becomes cheaper and cheaper.

At the infrastructure level, the barriers to information access are dropping dramatically with new ISPs coming into India and several players jockeying to provide bandwidth and other back-end services. However, without operating systems, applications and Internet content in Indian languages, key benefits of the digital revolution-e-commerce, low cost communication through e-mail, access to information databases, telemedicine services etc are denied to the Indian masses. Giving Internet access to an Indian who does not know a shred of English is like giving someone the keys to a car when there are no roads to drive on!

One development that can help India out of this deadlock is a national-level, collaborative effort to localise Linux to Indian languages.

Linux is a free operating system that has gained phenomenal popularity in recent times because it allows users to modify it to suit their own needs. Linux is a collaborative effort of thousands of programmers interacting over the Internet and is therefore not owned or controlled by any one company. In this article, we outline the economic and cultural imperatives for the localisation of Linux.

Free operating systems have several advantages for developing countries because most software packages today are developed in the west and then sold in developing countries where the parameters of affordability are completely different. The Bangladeshi activist Shahidul Alam expresses these differences poetically when he says, "A modem costs more than a cow." The benefits of free software multiply exponentially when we look at large-scale implementations. The Government of Mexico is estimated to have saved close to $125 million that would otherwise have been spent on proprietary systems when it signed up Red Hat to implement Linux in more than 140,000 schools and colleges across Mexico. In India too, large operators like World-Tel (which plans to have a thousand Internet Centres in Tamil Nadu, with each of them having between two to 20 PCs each) have expressed their intention to go the free software way. The company is negotiating similar deals with several other state governments. Organizations like World-Tel, Internet centres, schools and homes etc. can be expected to be significant users of Indian language operating systems.

The growth of content in platform-independent file formats (HTML, MP3 etc) has also reduced the dependence on a specific operating system, making Linux a viable option.

Apart from these, there are cultural reasons that make Linux attractive. The existing user interface paradigm of files and folders evolved because computers were essentially designed for a western audience familiar with real-life files and folders. There is no reason to assume why the same paradigm should apply to a trader in Tamil Nadu or a farmer in Madhya Pradesh.

The openness of Linux (and other free operating systems like Free BSD) allows local linguistic groups to customise user interfaces in ways that are far more culturally sensitive than any centrally controlled approach. Linguistic groups that may be considered too small a market by vendors can also take their destiny in their own hands by customising the Linux interface to their own needs.

It is therefore clear that Linux is a very attractive long-term solution to India's computing needs.

Localising the user interface of Linux to all the 18 official Indian languages will involve changing the menus and help-text to Indian languages and creating a whole stack of applications and tools (word processors, browsers, spell-checkers etc.) to enable computing in Indian languages.

This is a task that involves both technical and linguistic challenges. For example, should "File" simple be called "File" but written in Indian scripts because it is now a part of popular usage? Or should we find Indian language equivalents? In some cases it makes little sense. For example, how many people know that the Hindi word for computer is "sanghanak"? Or what is the Hindi equivalent for "Internet"? A very sensitive balance has to be struck between practicality and preserving Indian languages. However, Indian linguistic groups will have to wake up to the fact that their languages will become outdated if they do not become a part of the digital age. In fact, the Internet can be one of the finest means of recording, archiving and propagating Indian culture. Since culture is embedded in language to a significant degree, the ability to compute in one's native language can give Indian culture a significant boost.

However, one of the greatest roadblocks to computing in Indian languages has been the lack of widely accepted standards. If millions of people are able to freely e-mail each other, it is because of a widely accepted standard called ASCII (American Standard Code for Information Interchange). It is sad that in spite of claims that India is a software superpower, we cannot harness IT for the benefit our own nation's citizens and the greatest stumbling block is a lack of agreement on standards. Check out ten different Hindi newspapers on the Web to see for yourself. You'll end up downloading and installing ten different fonts that (in most cases) can be used for browsing that one site and nothing else. It is because of this reason that Hindi, despite being one of the largest spoken languages in the world, has a negligible presence on the Web. Informed sources feel that the Unicode standard (which Microsoft has adopted for the upcoming Windows 2000 operating system) will soon become the de-facto standard settling the language standards issue once and for all. If this prediction comes to pass, it will significantly increase the domestic market for hardware, software and services, which is restricted only to a small fraction of India's population that understands English.

There are several initiatives that are underway in order to make this possible. The National Centre for Software Technology has submitted a proposal to the Technology Development in Indian Languages of the Government of India. TheIndian Institute of Technology, Madras has already started work on localising Linux to Malayalam and Tamil. My own institute, the Indian Institute of Information Technology, Bangalore has committed resources to this the "IndLinux" project and started a collaborative effort to realise this goal. IndLinux has attracted the interest of organizations like FreeOS.com and many individuals located around the world.

In conclusion, it has to be said that the Indianisation of Linux is probably one of the most practical ways of making information technology available to millions and millions of Indians. It is now upto linguistic and technical groups to collaborate and make things happen.

-0-

Prof. Venkatesh Hariharan is with the Indian Institute of Information Technology, Bangalore. He can be reached at venky@iiitb.ac.in.

Comments

Files and Folders are not particularly western

[Effugas]

Apart from these, there are cultural reasons that make Linux attractive. The existing user interface paradigm of files and folders evolved because computers were essentially designed for a western audience familiar with real-life files and folders. There is no reason to assume why the same paradigm should apply to a trader in Tamil Nadu or a farmer in Madhya Pradesh.

Interesting that the author brings this up. I was recently reading about a very specific form of brain damage, caused by an operation to remove a tumor from the brain. A very small but highly focused amount of damage was done to a patient's language system:

He could name people.
He could name objects.
He could name cities.

He couldn't name a living animal. He would consistently mix up dog, cat, and any other term belonging to the family of "living animal".

If there's one thing linguists have found, it's that the core roots of language are not cultural--they're genetic. The base objects of communications--nouns, verbs, and so on--are by no means the only theoretical communication paradigms, but they're shared by every non-artificial human language.

You might wonder why I bring this up: In designing a method for interacting between a human and a computer, the properties of language are indeed important for establishing relationships. While there may not be literal files and literal folders in Indian culture, the concept of items existing within the branches of a tree is engrained deep within the structure of the human brain.

Now, "File" and "Folder" themselves are western analogies, to be sure. But there's a difference between recontextualizing an idiom and dismissing a natural paradigm.

Yours Truly,

Dan Kaminsky
DoxPara Research
http://www.doxpara.com

And this is the way the World is won...

[FFFish]

...person by person, country by country -- and starting with the corners of the world that are ignored: where there are barriers to language, barriers to affordability, barriers to access, barriers to the type of application that is needed.

China (slashdot story), Mexico (slashdot story), India... those are significant populations. It's a step in the right direction.

Remove language barriers: internationalize all parts of Linux.

Remove affordability barriers: release up-to-date packages that are designed to be useable on old 386 systems. In a lot of countries, Pentium are unaffordium.

Remove barriers to access: donate your old hardware to third-world countries. Help get Linux distributed -- donate a diskette-set to an emerging-world school.

Remove application barriers: internationalize applications. Identify what old software (running on DOS, Commodore, other old iron) should be ported to Linux.

Support the world outside of the little space you inhabit. Think outside the USA, think outside white Europe. Most of the world isn't like you. Look after them, and it'll pay back a hundredfold...

--

Re:And this is the way the World is won...

[Surak]

Remove language barriers: internationalize all parts of Linux.

This brings up a good point: do you internationalize the source code?

Seriously, think about it: the free software movement is all about bringing access to source code and the ability to modify that source code.

Do you internationlize the source (meaning comments, variable and function names, etc.)?

I think for the sake of complying with the open source movement, I think maybe it needs to be discussed and thought about.

The problem that becomes classic is that international versions of programs tend to be much further behind than current, U.S./English versions. It would be even more so if you decided to internationalize the source code, because while the original source is very much English, international versions would need to be translated.

It would be an interesting discussion. Maybe an Ask Slashdot forum would be appropriate?

Re:And this is the way the World is won...

[QuMa]

While I agree with the 'this is the way the world is won' comment, I'm actually not to wild about internationalisation. I think internationalisation is going to put up barriers in the long run, not break them down. If everybody gets used to 'computers speak english', everybody can communicate. So untill a real technological babelfish is made (no, altavista's babelfish doesn't count), people who don't speak english will be cut off large parts of the net. English is any easy language to learn, (It isn't my first language, and yes, it might show, but I'm reasonably understandable), compared to all other 'major' languages. So I say, everybody learn english. Don't waste time internationalising.

Re:And this is the way the World is won...

[Menthos]

I don't think that everything about an application should be internationalized. It's the UI of the application, and the documentation, that needs translation.

Internationalization of comments? This would indeed be useful to the non-English world, but I don't think that it's a high-priority thing.
Many source code comments are not even understandeable by other people than who wrote them, and since the source may change very often (but not necessarily the UI) it would be hard to keep up. This goes also for function and variable names. Translation of these could also very easily break things.
Not saying that there isn't translation of programming languages out there - an example that comes to mind is the macro language (VBA?) in MS Office applications that is translated (function calls etc.) in localized versions. However, I don't understand the point in this - most programmers or other computer-savvy people in the world know English well enough to be able to read the name (!) of a function call in the documentation and use it. Again, it's the documentation that should be translated, not the function calls, variable names, etc, where translation could break things.

I'm translating programs myself (GNU utilities into Swedish) and there's a reason why we (translators) have a strict policy not to translate such things as command-line switches. This would break things. So "--verbose" should be left intact in the translated strings, no matter if it is Hindi, Swahili, Esperanto, Dutch or Swedish. But the explaination of "--verbose" in "--help", and the documentation, is translated.

To sum things up: Since many wonderful programs lack proper UI translation in many languages (not to mention the documentation!), I think this is where the main efforts should be concentrated.
Also, internationalization is more than just translation. For example, I'd love to have more applications understand localization settings and properly react to them - I hate when an application defaults to inches as a measure, weeks beginning on Sundays, Legal as the paper format, AM/PM clock, "," as thousand delimiter, etc, etc. All things that are NOT the "standard" outside the USA, and things that many programmers overlook. Even worse is when these annoying things are hard-coded - it goes from being merely annoying to a chronic pain, and often makes me refuse to use the program at all.

So even if programs are translated, there's still more left to do to make it internationalized.

Re:And this is the way the World is won...

[ianezz]

> This brings up a good point: do you internationalize the source code?

Just my 0.02 Euro here.

I'm Italian, and I live in Italy. I'm just a C/C++ programmer like many others. The source code I write has all the comments and symbols names in English. This mainly for 2 reasons:

1) Technical language: programming needs a whole new class of technical terms. Simple words like `formatting', `font', etc. didn't have equivalents in Italian when they were introduced to the mass in the former '80. Simply, we adopted (and distorted) the usual English words. Attempts to create a brand new tech gerg just failed. After 20 years the mass starts using the Italian form for "directory", but still it's a minorty. Just figure yourself when it comes to source code...

2) Technical documentation: good technical documentation is written in English. Programmers like me avoid translations, just because they make life more complex (the effort is doubled: first you read it, then you try to figure out how the terms used are related to the well known English terms). I'm translating the GNU Emacs manual in Italian in my spare time, and it's a damn difficult work (and it's not at all about programming - figure it yourself when it comes to programming books).

3) I18n: English is good for symbol names, just because you don't have to do with accents. Italian words just look ugly without accents, and using ISO-8859-1 is still not a viable option for sources... :-)

4) Last but not least: English is the de facto `lingua franca' for programmers, just like Latin was for the europeans in the passed ages. If you start writing your sources using one of the 18 languages of India for comments and symbol names, probabilities are that only indians will ever peek into them, just taking out the rest of the world. And since most people giving help and advice on programming out there speak English well enough to make one understand them (perhaps except Alan Cox, sometimes ;-), probabilities are that one's best choice is (for now) learn English.

Of course, I just use Italian in every other context, just like everyone else...

I hope this is a bit clarifying.

Language, Software, Web, and Microsoft!

[kbahey]

The article raises very good points about the issue of software localization / internationalization.

Many of the points above are shared with other non-Western languages (lack of a single standard character set, the issue of linguitics, user interface, ...etc.)

Here in the Middle East, we face a strikingly similar set of problems, with some added bonus. People who speak Arabic as a first language were about 181 million in 1997 (according to this Times article), making it the Fifth language in the world after Mandarin Chinese, English, Spanish and Hindi.

Arabic is unique in that it needs the peripherals (the VT100 terminal and the printer) to support automatic contextual character shaping on the fly, and Right-to-Left orientation. It shares these qualities with other Semitic language (Hebrew, Aramaic, Syriac, Nabatean). So, a character set and a font is not enough, like the case in most western language.

Several years ago, there were lots of character sets, each in use by a different hardware vendor, and even many vendors had several character sets. A standard (called ASMO-708) emerged, and was adopted by almost all vendors using ASCII (IBM was EBCDIC, so they were different).

In the early 1990s, a company called Al Alamia developed a version of Microsoft Windows 3.x that supports many character sets, including ASMO-708. Microsoft hired (read stole!) the main developer from Al Alamia, there was a law suit.

When Windows 95 came, the battle was won (by MS!) in the Arabic arena.

When the web arrived, things got even worse (from a standard point of view) and a Netscape version (called Sindbad) was developed by Sakhr to navigate the web in Arabic, and lately released it as a plug-in to Navigator 4.x. It is terribly slow though. Microsoft won the browser wars, and virtually all the Arabic users are now using Windows 95/98/NT with MS Internet Explorer. New development of Arabic web pages is almost done entirely for MS Internet Explorer. Not good!

Dynamic fonts are great and are used by a few sites. They work great with MS IE or NS Navigator, but are not widely used.

So, where does this leave Linux? There are:

• No arabized GUI for Linux at all, which makes me still use a dual boot to get Arabic.
• No good arabized browsers under Linux either.
• Microsoft is gaining a virtual monopoly on a whole culture of 22 or so countries!

I am still using Netscape for e-mail and browsing (even on Windows, and fed up with its problems!), but have to use MS IE for browsing Arabic web pages! Sad!

I have some links on Arabic on the web (scroll to the bottom of the page on what is available for Arabic on the net.

--
Have you checked out Muslim Investor?

Port Indians to English...

[Bill-Gates]

It would be even better to port the Indians to English.

Sure, they're not very portable, but with a little re-programming, they will even be able to get onto the internet. In addition, not only will they benefit from [cough]Open Source, but then they'll be able to use my wonderful Windows Products in English.

It will save a lot of work for my programmers, to not have to port Windows to another language, and they can use the free time to implement some other features I've been wanting. Security, stability, are a couple of features I've heard good things about. Platform independance sounds kinda neat, for hardware anyway...

Regardless, it is very important that you cease your efforts to port Linux to Indian languages, but rather port Indians to English..

Sincerely,
Bill Gates
(Score 5, Monopoly)


Have you tried my newest software? Microsoft.com