Overwriting the early internet in Africa

Anriette Esterhuysen advocates for digital rights from a Global South perspective. During the 1990s she was the executive director of SANGONeT in South Africa, where she helped pioneer one of the first internets in the region that linked up universities and NGOs with global networks. She is the former executive director of APC. 

She was interviewed by Xavier Coadic. They spoke about her early experiences with the internet in South Africa, the TCP/IP protocol, FLOSS in Africa, and open content. The full interview can be downloaded at: https://media.apc.org/w/p/hvSrvn6rTr1dYgL8875EiF   

Anriette Esterhuysen:

Linux is pretty old, and APC used to develop its own software in the 1980s and 1990s, before commercial internet, before TCP/IP. There was something called “pnotes” that APC members around the world ran. And then it also wrote various protocols to allow the APC software to interface with InterNetNews, which is what the academic internet was using at the time.

It was also involved in developing something called the Marimba, which was based on the FidoNet [FidoNet allowed for communication between bulletin board systems]. The code for all of these was shared freely. This overlapped with the emergence of Richard Stallman and the free software movement. Maybe some of the people that were involved in this might have identified as part of the free software movement and some might not have. But they did all collaborate in a very similar way, by sharing code, by sharing tools. What was really interesting was how this also coincided with the emergence of the open content movement, which Creative Commons grew out of. 

The FLOSS movement was actually part of mainstream early commercial internet in Africa, which used open source for their services – and in fact, FLOSS would have been in the mainstream in the early internet, but Microsoft developed an internet server package. It was awful. It was terrible software and very insecure, but they pushed it onto the early internet. So many of those not-for-profit early ISPs who would have used Apache or other FLOSS tools ended up using Microsoft, which was technically really bad for the development of the internet in Africa.

What changed with the arrival of the commercial internet based on TCP and IP protocols? 

It destroyed what we had. It put us back 10 years. In the case of Africa, particularly FidoNet, it was designed to provide connectivity in contexts where there were very few phones and very poor-quality phone lines. You would communicate using a modem with what we called a local “post office” using a database query. Emails would sit on a server and then that server would make a phone call to another “post office” to exchange information before disconnecting, and so on. You would get a response to your query maybe a day later or two days later or an hour later. It all depended on how often you connected. When permanent connectivity became the mainstream of the commercial internet, people without phone lines, without good quality phone lines, were unable to use the internet. So Africa was quite ahead by the end of the 1990s. We were pioneers and then we lost momentum; we lost access and we have never been able to catch up again.