Wednesday, April 22, 2009

Reaping the benefits of open source

This article written by me appeared in Network Computing's India edition. Please feel free to translate this and get it published in other geographies. There is no need to credit me as the author, but do use this to spread the message of open source and drop me a line of acknowledgment when it is sent to a publication or gets published in print/online. And if you prefer the term "free software" that's also OK with me.

Venky
=====

Reaping the Benefits of Open Source

Open Source Software is moving from the edge of the enterprise into its very heart. CIOs can benefit from this development

Recently, in a Gartner report, 'The State of Open Source 2008', it was mentioned that, "By 2012, more than 90 percent of enterprises will use Open Source in direct or embedded forms." The report goes on to add that, "Open Source is a phenomenon with a broad impact. Chances are, if you do not think you use it, then you use it; and if you think you do use it, then you use lots more of it than you know.”

Why is Open Source becoming so pervasive? The reason is that we are now entering an era of Collaborative Innovation. Open Source Software (OSS) is the leading example of this trend, but the Open Source development model based on collaboration, community and the shared ownership of knowledge is rapidly expanding to other areas like content (Wikipedia), medicine (Open Source Drug Discovery), scientific publishing (Public Library of Science) and other areas of society. With 1.2 billion people on-line, the Internet, which is the largest collaborative platform that mankind has ever seen, has enabled OSS like Linux, Apache, Mozilla Firefox, Open Office and others to flourish. In the next couple of years another 600 million people will join the Internet. Thus the trend towards increasing collaboration is only set to grow and this is reflected in the explosive growth of Open Source projects across the world.

A few years back, OSS could be found on the edge of enterprise, running workloads like mail servers and web servers. However, the growing maturity of the OSS ecosystem means that it is now moving into the very heart of the enterprise, running mission critical servers, desktop computers and even application areas like CRM, ERP, Document Management, collaborative wikis, Content Management Systems and many others.

In these recessionary times, OSS also offers a major advantage in that users can often freely download and try out the software and pay only for value-added services like support. The growing interest in OSS has thrown up a number of software startups that specialize in supporting OSS. For CIOs, this development provides an alternative because the cost of buying support for OSS is usually far less than the cost of purchasing licenses for proprietary software.

For instance, download and check out OpenOffice.org, the full-fledged, Open Source office productivity suite. Many organizations have made OpenOffice.org, the default choice on their desktops and have generated significant cost savings when compared to proprietary office suites. A leading bank, known for its technological savvyness, has almost 70 percent of its staff working on OpenOffice.org. A few years ago, they looked at the increasing cost and hardware requirements of proprietary software and decided to switch to OpenOffice.org. Initially, users had to adjust to the new software, but a four-member helpdesk enabled them quickly become comfortable with OpenOffice.org. The cost of the help desk was far lower than the licensing fees and the increased hardware costs that would have been incurred on proprietary software, which has been restricted to a small group of financial analysts within the company.

Similarly, a study done by IIM Ahmedabad found that the Government of Delhi has saved almost 80 percent by switching to OpenOffice.org. One important reason for switching to OpenOffice.org was its support for the Open Document Format, an open standard for office documents, that ensured that needless upgrades of office suites and its underlying hardware would not be forced upon them. The usage of open standards also helped the Government of Delhi avoid vendor lock-in, which invariably reduces negotiation capabilities of the customer and increases cost.

In many ways, Open Source is becoming the baseline for software development. Within the past couple of years, venture capitalists have started investing several billion dollars in Open Source startups and have reaped handsome returns. The open and inclusive development model; and the freedom to modify the source code and improve the software attracts the best minds from across the world. When compared to the closed development models of proprietary software, which depends on internal skills, the open source model proves to be superior. Eric Raymond, author of the landmark book on OSS, “The Cathedral and the Bazaar” expresses it succinctly by saying that, “Many eyes make bugs shallow.” The reliability, robustness and low number of defects per thousand lines of code in open source software like Linux and Apache bear this aphorism out.

Of course, CIOs of enterprises need enterprise-class support. The growing adoption of Open Source in enterprises like LIC, Axis Bank, Central Bank of India, Bharti (Airtel), mission critical portals like Naukri.com, Yatra.com etc. point to the growing support infrastructure for OSS. As users test and deploy OSS, the demand for support is growing immensely.

Finally, while Open Source is not a panacea for every recessionary ill, in these tough times, it makes sense to evaluate OSS instead of spending precious money on expensive proprietary software. Open Source is a long-term trend that is here to stay and those CIOs who judiciously adopt this new paradigm of software development and deployment will benefit from it.

The "computer programme per se" conundrum

In a previous blog entry, I had mentioned how the Draft Patent Manual interprets Section 3(k) of the Indian Patent Act in a manner that allows software patents a back-door entry. I had also mentioned in a previous post how the term "per se" in Section 3(k) which says, “A mathematical or business method or a computer programme per se or algorithms are not patentable” leaves a lot of room for mischief. Knowledge Commons has submitted an unambiguous definition for the term "computer programme per se" which would be in tune with the intent of the Indian Parliament, which deleted a proposed amendment to Section 3(k) that said, “a computer programme per se other than its technical application to industry or a combination with hardware; a mathematical method or a business method or algorithms.” Prabir Purkayastha, Chairman of Knowledge Commons lead the creation of this definition. Prabir is one of the few people I know who can climb the rareified heights of strategy and also roll up his sleeves the next moment and work on nuts-and-bolts issues that need to get done. This definition would just not have been possible without him. Richard Fontana of Red Hat, Mishi Chowdhury of Software Freedom Law Center, Tahir Amin and Jaijit Bhattacharya also provided valuable inputs that went into the final submission to the Indian Patent Office that reads:

Computer programme per se in the relevant clause means (a) any computer programme in the abstract, (b) any computer programme expressed in source code form, including source code recorded on an information storage medium, or (c) any computer programme that can be executed or executes on a general purpose computer, including computer programme object code designed for execution on a general purpose computer that is recorded on an information storage medium. An information storage medium means any disc, tape, perforated media or other information storage device, which, if fed into or located in a computer or computer based equipment is capable of reproducing any information, other than an information storage medium that itself represents an inventive contribution to the art. A general-purpose computer here means a device capable of running multiple unrelated programs, often simultaneously for different purposes. It will comprise at least of: (1) one or more central processing units, (2) one or more input devices that are not specific to any one program, (3) memory, (4) one or more non volatile mass storage devices, and (5) one or more output devices. However, a general-purpose computer does not include a device that itself represents an inventive contribution to the art.
Under the foregoing definitions, a claim that merely recites software elements without any reference to hardware is per se unpatentable. If a claim recites both software elements and hardware elements, but the hardware elements amount to nothing more than reference to the components of a general purpose computer on which the software is executed, or an information storage medium in which the software is stored, such that the only possibly inventive aspect of the claim resides in the software elements, then the claim is not patentable. If the software/general purpose computer is combined with other hardware, and the inventive contribution resides primarily in the software or in the software in combination either with components of the general-purpose computer or in an information storage medium, then the claim is not patentable. If the claim recites software elements and hardware elements, and the hardware elements themselves are an inventive contribution to the art, then the claim may be patentable, provided that the claim as a whole is such that the pre-requisites of novelty, non-obviousness and utility are met.