It would have been interesting if the article mentioned the translation of OpenOffice.org into Swahili which was actually completed quite a bit before the Microsoft version. I think they suffered similar issues of adoption.

At Translate.org.za we have done similar things in Suuth Africa, translating Mozilla Firefox (a web browser), Mozilla Thunderbird (an email program) and OpenOffice.org (an office suite) into all the 11 official languages of South Africa. We've essentially demonstrated that a non-profit effort can deliver localised software.

But we've also suffered from many of the disappointments mentioned here. But when we have used the channels of existing projects we have begun to see more adoption. Thus our release of an Afrikaans version of Mozilla Firefox has lead to major changes in adoption. It has also lead to an increase the adoption of our Afrikaans spell checker.

The article does highlight one concern for me. The misalignment of goals. In the article in mentioned that the localisation of Microsoft in Swahili was targeted at 150 million people. Rubbish! That is the total number of Swahili speakers. If that is your goal you will fail. Rather set your gaols in reality which might be as simple as changing peoples attitude. In South Africa simply by translating OpenOffice.org into the 11 languages we finally got Microsoft to translate into 4 languages. That is a triumph for us as it slowly addresses the perception that our languages are in some way inferior.

I would encourage people not to give up. But set realistic goals and work with communities.