This paper takes a frank look at what has and has not been achieved in African language technology during the past two decades. Several questions are addressed: What was the status of technology for African languages 20 years ago? What were the major initiatives during that time? What were their successes and failures? What can we learn from these experiences? How does this inform the work that we are planning going forward? Examining in particular the history of Swahili, it is argued that technology projects have often achieved their expressed aims, but have collectively not significantly advanced the normalization of African languages as operable within the technical sphere, even while Africa has become blanketed with mobile technology. It is argued that future projects will succeed only by asserting the goal that technology of 2035 must be fully operational in users' primary languages, and gearing policy, funding, and individual project efforts toward gathering and deploying linguistic data for a large number of African languages to meet cutting-edge technologies as they emerge