Microsoft recently announced plans to offer a Kiswahili version of their Office product in the next six months, perhaps followed by Windows XP. Microsoft is evidently looking for other ways to make their products more attractive to a Great Lakes audience, changing the Windows startup sound to a Taraab melody…
Microsoft’s manager for East Africa explains that the company is looking at the 100m Kiswahili speakers as a large possible market, and that the company is looking at other large language markets, like Hausa, Yoruba and Amharic. While there are certainly a large number of possible Kenyan, Tanzanian and Ugandan users, the high level of piracy in the region suggests that the motivation may not be pure profit. Instead, this looks like an excellent way for Microsoft to maintain mindshare in the region while waiting for the market to mature to the point where they can later demand license fees from government offices and schools.
I had dinner a couple nights ago with Duane Bailey from the excellent translate.org.za project, who suggested that the project may be a direct reaction to some of the activities his non-profit has recently undertaken. Translate is working on localizing OpenOffice into all the eleven official languages of South Africa. They’re focusing on OpenOffice, instead of Linux, for instance, because they’re interested in tools that enable the creation of content in local languages. OpenOffice is also a great “stealth” way to introduce people to the power of open source – folks can try it out without changing platforms from Windows, making it lots easier to experiment with than setting up a Linux box.
Duane wonders whether Microsoft’s announcement is a response to the Kiswahili spellchecker he and Jason Gisethko recently developed at Tactical Tech’s Africa Source gathering in Namibia. In under a day, Duane and Jason were able to combine a list of Kiswahili terms Jason had been collecting for years with some existing spellchecking code and produce a functional and popular kiswahili spellchecker.
While that’s a great example of one of the advantages of Open Source – the ability to quickly localize existing code – the community as a whole doesn’t appear to have embraced Kiswahili as a high priority. It doesn’t show up as one of the 66 languages fully or partially supported by Mandrake, and the translation effort at Debian is one of the least popular, with less than 2% of essential strings translated.
David Gyewu, Ghana’s very sharp Deputy Minister of Communications, mentioned yesterday that he thought Open Source folks might be overfocusing on the importance of localization. He argued that, while most Ghanaians spoke at least some Twi (the language of the dominant Ashanti people), most didn’t read it, and that he didn’t therefore see a ton of need for a Twi Open Office. He also felt that IT was a great incentive for people to improve their English, which in turn improves their chances to succeed in global business. I’m not entirely sure I agree – I think there’s pretty good evidence that people learn to read best in their native languages, and that there’s an importance to protecting and preserving existing languages – but I need to chew on his comments for a while and see if I think localization is still the best argument in favor of Open Source.