Linux Asia 2004 Keynote

2004-04-17

The text below is an abbreviated version of a keynote I presented at the recent Linux Asia 2004 Conference held in New Delhi.

I'm delighted for a couple of reasons. Most importantly, because when I look at Open Source today I see no more important nation than India and no more important region than Asia.

And, I'd like to add, no more important application than OpenOffice.org. OpenOffice.org is the key to getting Open Source accepted on the desktop.

It's not difficult to figure out why India and Asia are important forces shaping Open Source today; it's hard to figure out why others in the US and Europe seemingly have yet to understand that (at least, there have been scant articles published on the subject, but of those that have been, we can count Frederick Norhonha's series on OSS in Asia last year; Noronha is based in Goa, India, I believe). Perhaps it has to do with the conception of Open Source many Americans and Europeans still maintain (as does the press). That's a conception of the lone hacker or developer who has an itch to scratch, and either joins an existing Open Source project or forms his own, so that he can collaborate with others, compete with others, on building better code.

To be sure, this model still exists and even has the virtue of mapping a reality. Individual developers motivated by personal reasons do attach themselves to a project, and OpenOffice.org has it share of them; two even co-lead important core projects. But it tells only a very small part of the picture and certainly underplays the vast muscle of national interest which can mobilize people and structure funds to support or at least enable open-source projects. It also misses what has recently been seen as the local importance of Open Source. As researchers such as Riyab Ghosh, Jordi Carrasco Muñoz, and others have illuminated, Open Source is the path for national and regional independence from a IT regime that determines a dependency on remote sources of increasingly important technology, server and office software. "Dependency" because the money spent on software is sent out of the country; and dependency because one must spend on software in order to compete in the global market; actually, in order to compete at all.

Should the price of business be so high? Should it be at the sacrifice of local power, local wealth, local identity? The accepted international language for business is now English, the default templates for document creation US, testaments to economic empire. Most software is accordingly in English, though the more popular programs have been translated and customized (localized) to the major languages. But localizing any program is not a trivial task and is done by private companies only for reasons of market or profit, not out of altruism.

I read in the paper that Microsoft is planning on further localizing its Office suite. That's great, I suppose, but as I see it, it does not solve the problem at all, it rather extends it. The main problem is still that money and resources are sent out of the country and a whole ecology of software production is distorted to cater to the profit of a vastly wealthy corporation more than ten thousand miles away, on the opposite side of the planet. An economic regime is perpetuated at the expense-literally-of local wealth, intellectual as well as financial.

Open Source, in contrast, is a means not just to save local money but to create local wealth and build local identity. The money you send to Redmond, Washington, in the US, for every purchase of Microsoft Office or Windows, is better spent, we believe, building local industry, helping local talent, creating a local economy, all by using Linux and other open-source systems and applications.

But despite its appeal, Open Source won't get very far, in India, or anywhere in the world, without a bridge that connects the proprietary world with the open world. Facts are facts: Most businesses throughout the world use Microsoft Office and they exchange documents with others who use Microsoft Office. Governments use it, too, as well as continuing to use specialized (and often obsolete) proprietary products. Absent a bridge that connects the divide between world of legacy closed systems and open systems, between proprietary software and Open Source software, there will be no success for Open Source; one needs an application that allows users of whatever sophistication to exchange documents and files with their colleagues across the globe; that allows them simply to feel comfortable not just about their work but about how they do their work, and that allows all users to build upon and use an extant library of works.

OpenOffice.org is that bridge. It is easy and familiar to use and because it is Open Source, it can be made easier and better by local users. Local users can configure it to satisfy their needs not remote marketing desires. It has all the advantages of proprietary products (including support) and of Open Source ones.

OpenOffice.org was initiated late in the 20th century by Sun Microsystems, which, on 13 October 2000, gave to the Open Source community under the LGPL and SISSL licenses the source to StarOffice 5.2. The goals of the project were those touted by advocates of OSS in 1999-2000, before the collapse of the dotcom bubble: OSS would get us to market faster and better because the open-source community of developers would be attracted to the project and devote their energies and time because they wanted to, because they had an itch to scratch.

OpenOffice.org has succeeded in ways that have transcended expectations for Open Source development and brought them more in line with the new, real world of the 21st century. To be sure, a community of developers has formed; equally, the office suite OpenOffice.org has developed faster and is better because of community development. But OpenOffice.org is not like Open Source endeavors such as Linux or Apache or many others. OpenOffice.org is also sponsored by a major enterprise, which continues to lead development, and that sponsorship has affected our trajectory and successes. In summary, we have benefited by the sponsorship and have thrived by the implicit guarantee that a major corporation is backing development of the product.

Clearly, our chief success has been in the adoption of the free application, also called OpenOffice.org. Very few anticipated that OpenOffice.org 1.0, released in 2002, would be so incredibly popular. I suppose the doubters felt that because OpenOffice.org did not then have all the gleam and shine of something from Microsoft, and because it had to be downloaded and didn't come in a shrinkwrapped box that regular users would stay away from the application. But of course, all that shines is not gold and all that is does not shine; OpenOffice.org 1.0 did what most users wanted and despite the lack of corporate marketing (we do our own) word got out. For users of Windows, who remain our greatest fans, as well as for users of Linux, Solaris, and Mac OS X, it allowed them to create text, presentation, spreadsheet, and drawing documents that could be saved in the Microsoft Office format, as well as several others, all for free. Free: as in speech as in beer.

Was it any surprise then that this application has been so very popular?

Sure, we could not offer professional support; we offered, and continue to offer, free community support (in at least 20 plus languages), though we have now also included professional support, including that offered by Sun, which has extended the StarOffice support package to OpenOffice.org. (This arrangement, of course, makes a world of difference; companies will not look again if there is no professional support offered.)

But OpenOffice.org was easy enough to use then, and it is easier now, that support was not such an obstacle for individuals. Users were more excited by the fact that this free application could do more or less what Microsoft Office could do--only on Linux, Solaris, and Windows and Mac OS X. They could use it to read Microsoft Office files and, perhaps more important, save files in the Microsoft Office file format. Other users receiving files generated by OpenOffice.org would never be the wiser; they would think, if they thought about it, that these files were in fact created by Microsoft Office.

In the year and nine months since the release of OpenOffice.org 1.0, we have issued many updates and have established OpenOffice.org and its derivatives as the key to any successful open-source desktop system. We have also launched dozens of new technical and support projects. OpenOffice.org is so big and powerful that it is a veritable Open Source ecology unto itself and directly and indirectly nurtures projects, companies, and individuals. We have formed not only a community of developers but perhaps more importantly an economy of support and development that builds local wealth and local talent.

Shortly before the release of 1.0, we had formed a few Native Language projects. These projects offer support, information, resources in a user's native language. The first was in French. After 1.0, we pushed for more Native Language projects. The logic was simple. Development and education could proceed faster in a users native language than in English. To prevent chaos, we mandate that all coding discussions be done in English but discussions having to do with everything else can be done in the users native language.

There was a philosophy behind this approach, too. Guy Capra (who initiated the project) and I didn't want OpenOffice.org to be limited by the artificial constraints English imposed. It was obviously an international thing from the very start and had the promise of being a thing of the world's people. OpenOffice.org and Open Source would not just suffer but do a disservice to the people if they ignored the languages of the world. Keep in mind that an Open Source project is not like a company, where the executive allocates resources (employees) and something is done by executive will (employees work on what he or she wants). It is a bazaar, where individuals, groups, act on the basis of their own desires and abilities. Things can be stated as being important and really desirable-but who states them? The community of users and developers, though there may be spokespeople, such as myself, who articulate community desires. And who will end up doing them? The developers, working with all other contributors. In our case, Sun's own StarOffice agenda affects the product's shape but does not limit others from actually working on what they want. Our Native-Lang projects were and are efforts to coordinate interest and work in extending the product and project.

Of the Indian languages, we have now two language projects, the Hindi and Tamil Native-Language projects. These projects are responsible for supporting their respective languages, including maintaining development and QA on the localizations. Those are rather serious responsibilities, and they demand time, resources, effort. But, despite the demands of these projects, the payoff is so great that we have every reason to believe these are just a start, other Indian language projects are anticipated. Because OpenOffice.org uses primarily the LGPL, there is every reason to think that the drivers for creating these projects will include government, businesses, as well as students and independent developers. That's because the motivation is more than intellectual. The licenses allow for selling enhanced binaries of OpenOffice.org. To paint the picture plainly, a company could localize OpenOffice.org 1.1 to any regional language, add a few elements and sell it to businesses and schools. Such a localized version would be legal and further provide users with a tool by which they could create papers, articles, memos, spreadsheets, presentations, websites, and much, much more, all without having to learn yet another language.

To be sure, as I learned today, Microsoft is planning on translating its Office and operating system to 15 or so Indian languages. But it will still be a thing of the elites and pirates, still a thing over which you, the user, the developer, will have no real control. In contrast, OpenOffice.org gives users and developers full control over the source and its packaging and it is free, meaning it is not merely a tool for the elites of a nation.

Allow me to provide a key example.

Earlier this conference, I meet some remarkable people. I met Ranjan Deb, Sachin Bujone, and Nitrin Meshram, who had come to this conference from Chandrapoor. There, they run Learn IT Fast, which teaches meritorious students and others basic and advanced computer skills using often donated computer equipment. Their students go on to find jobs in the growing Indian IT economy; for them, Mr. Deb's company is invaluable.

But what have the students learned that allows them to succeed?

They have gained a certificate in using Linux and, for the advanced students, in using Oracle. (Mr Deb will be adding MySQL, too.) With this certificate, the students can find jobs. Now, Messrs Deb, Bujone and Meshram could not certify their students for Microsoft. It costs too much and for the cost does not really provide value.

Linux does.

What's more, Mr Deb can further include the free OpenOffice.org in his educational package. And because OpenOffice.org has been localized to some Indian languages, students need not have to know English to use it. They can, I hope, use their own language.

I find this story extraordinarily exciting. This, I tell myself, is what Open Source is supposed to offer: an opportunity.

But in order for Mr. Deb's work to be really effective, he and his colleagues need, of course, support. I can provide some of that, by telling you and the world of his and others' efforts. But he needs more than that.

He needs, I hazard, a government mandate. One that enables open-source development and open-source software. One that dares to provide monies to companies who have embarked on developing and deploying Open Source solutions. I am not suggesting that Open Source technology be required in all cases. I am suggesting rather that the government, federal and local, provide money to meritorious groups who are using Open Source to solve otherwise intractable problems of education, development, distribution. Why? The economic reason is by now fairly obvious. As professor Ghosh has brilliantly elucidated, Open Source development and Open Source software saves money coming and going. Mr. Deb's students will themselves further enrich the local economy by taking on jobs that would otherwise hesitate to manifest in their region; and new applications to resolve both new and old problems will be created, for with the source open, one has the power to act now, in one's own interests.

But the consequences of Open Source go beyond the merely economic. Open Source is essentially a system of production that tends to be, in some circumstances, particularly efficient. But it also promotes a collaborative environment that we need now, more than ever.

I end with a question which you can answer: How do we make these good things happen?