Why the third world won't save open source

The primary barrier to Linux growth is the cost of moving from a Windows ecosystem to a Linux ecosystem. Developing nations, however, have less existing IT infrastructure. Much as African nations are bypassing wired telecommunications and moving straight to wireless, why can't developing nations bypass the Windows standard and grow a Linux ecosystem?

First, don't underestimate the installed IT base present in developing nations. Nations such as Brazil aren't Zaire. Per capita GDP in Brazil is six times that of China, and over half of the Czech Republic. The IT base in most nations isn't exactly a tabula rasa, and any large organization will have already built up credible amounts of IT infrastructure.

Second, consider why English is the de facto language of business. Given the need to communicate in business situations, people naturally gravitate towards one language, and for historical reasons, that language is English. It doesn't matter whether an economy newly integrated into the global marketplace has less of a track record of using English. They will use English in business situations, because that's what the wider business world is using.

The same principle applies in computer technology. Asians outnumber Westerners by a factor of 4 to 1. Still, the West will remain a critical market for the forseeable future, and having systems that integrate seamlessly with, and can consume products created by, Western corporations will continue to be important. That means Asian IT infrastructure is likely to mirror Western IT infrastructure.

Some think that third world nations' enthusiasm for open source, driven in part by a desire to replace IT imports with domestic product (a form of IT industrial policy), will change the equation and pull the Western world into open source nirvana. Unfortunately, such industrial policies don't have a great track record of success.

Dana Blankenhorn wondered aloud in a recent blog post whether Brazil's commitment to open source might "jump-start profitable businesses." He also noted, interestingly enough, the disastrous consequences of Brazil's last industrial scheme:

In the 1980s the country instituted a market reserve policy aimed at limiting PC imports. This proved disastrous. The country missed 15 years of progress.

The goal, of course, was to replace PC imports with domestically created product. Unfortunately, hiding domestic companies from foreign competition merely makes weak domestic companies that can't survive without government preferences. I don't see why replacing the last system of preferences with a new system of preferences will fare any better

The open source community can't look to the developing world as the tugboat that drags the developed world into the Linux port. They will need to figure out a way to convince the developed world to adopt Linux. In my entirely fallible opinion, the only way to do that is to lower the cost of shifting to Linux, and that's going to depend on making it easy for Windows developers to move into open source. If the open source community doesn't want to do that, then Windows developers aren't going to move, and that means the people who use their products will stay put, too.