OPEN SOURCE SOFTWARE

This is another in a series of entries occasioned by Jeffrey James’ book, “Bridging the Global Digital Divide”. James returns again and again to open source software, providing more of a sales pitch for developing countries to adopt open source, rather than a balanced view of the issues (in my opinion).

He focuses on Linux and office suites. These are probably the “poster children” for the open source movement. It may well not be the case that the lessons from such software generalize to other kinds of software. For example, a large number of programmers have cooperated in the development of Linux, while the vast majority of open source efforts languish for lack of programmer interest.

There are some definitional issues. “Open Source” means that the source code for a piece of software is available. For most open source software, this is accomplished by publishing the code. However, I have read that Microsoft is to make the source code for Windows available to the Government of India; thus for the GoI, Windows becomes “open source”.

Then there is the issue of intellectual property rights. Proprietary software is protected. But most open source software is also protected, in part automatically, and in part by the team developing the source code, in order that they can control the development and define the most appropriate versions. SCO has sued IBM for infringement on SCO’s intellectual property rights to parts of Linux. So Open Source code is not necessarily freely available to use in any way one wants. Some software has license conditions that do make it freely available to build upon and to further disseminate the products incorporating the open source software.

Finally there is the issue of cost. Of course we think of open source software as being available without charge by downloading from a website. On the other hand there are a number of firms that charge for Linux, and provide support as well as the software in return for the charge. On the other hand, a lot of companies give their software to educational institutions and other civil society organizations as donations. The cost of Windows can be quite low: thus Microsoft is reportedly selling windows and its office suite in Thailand for US$40.

There are real economic costs to developing, maintaining, supporting, and disseminating software. People have to do the work, they have to have computers to do it with, the process has to be managed, etc. There are several ways this can be done:
1. Commercial firms can do this as part of their business;
2. Organizations can develop software for their internal purposes, and give it away to others without commercial gain (indeed, in the old days, users groups used to share software in this way, and it was the primary way to obtain software);
3. Governments may develop and disseminate software as a public good;
4. Software can be developed by civil society, and if you think about it, most of the open source efforts can be so classified.

I think the reason that most open source projects don’t work is that civil society is not well organized to mobilize resources for software development, nor is it well organized to manage the development process.

I would note that corporate support resulted in the development of Open Office software, and has been important in the dissemination of Linux, Apache server, and router software. Indeed, companies have many reasons for supporting efforts to make software available to large number of users.

James notes that Microsoft has been more effective in disseminating its commercial software products than have open source networks in disseminating competitive products. I suggest this is exactly because Microsoft charges for its products, and is able to use the resources it so obtains to market (i.e. disseminate) it software still more widely. While it may be the case that “if you build a better mousetrap, people will beat a path to your door,” it is clearly not the case that all the people who need mousetraps will use that path!

Governments, with few exceptions, have shown very little interest in supporting the development and dissemination of software for developing nations as a public good! Nor do they nor the educational institutions they occasionally fund for this purpose seem very good at the job. The exceptions to the rule might warrant some study to figure out what best practices are in this field.

I am consulting with the Development Gateway Foundation, which is an experiment in such financing, and it will be interesting to see how successful we are in the process.

I will also point out that Linux is built upon UNIX, and thus on software approaches that are comparatively old. Microsoft bit the bullet long ago, and changed to Windows, building upon developments made by others, to improve the user friendliness of the PC operating system. While antiquated technology is quite suitable for many purposes (navel officers are still trained on sailing ships), one would not want to condemn developing countries to use of outmoded technology because they depended on open source systems.

James suggests that use of unlicensed software, while cheap, is ultimately unproductive. I will point out that the United States is supposed to have avoided protecting European books via copyright protection through much of the 19th century, because there were more benefits to Americans from pirating than from publishing copyrighted materials. Of course I would not suggest that developing countries follow the American lead.

So lets see lots of Software Police in Africa! What else have African nations to do with their resources other than protect the international software industry?

There are lots of people who know a lot more about Linux and open source office suites than I do, but I would not be surprised if these eventually do become the standards for government and education.

I just saw results of a survey of demand for training in ICT from developing nations. The top ranking field was “security”. Will developing countries find that massive use of software, for which the source code is freely available to everyone, will contribute to computer and network security problems? Perhaps not.

In any case, I think that banks and other financial institutions, manufacturing firms, and many others in developing countries will continue to find most of their software needs to be best met commercially. I would like to have seen James spend more time on the need for developing countries to create the domestic software industries to develop and adapt software for such needs.