Summary

  • In 1996, the University of Gezira had a limited GreenNet email node polled from London; in May 1997, an external ping reached sudanet.net through NSN.
  • A government agricultural-information report said public Internet access began in October 1997 and counted 250 subscribers, while later accounts placed Sudanet’s establishment and network introduction on different dates.

The first packet to answer from a country can be a crisp historical event. It is also easy to ask it to prove too much. Sudan’s surviving record sets a university email relay, an externally observed Internet route and a reported public service beside one another. They are connected stages, but the sources do not give them a single shared start date.

The earliest trace here was not interactive Internet access. A University of Gezira node in Wad Medani joined GreenNet’s store-and-forward network in spring 1996. Messages were exchanged with a node in London, which polled Gezira twice daily. A May record describes the service as experimental and restricted to academic use. It names poor telephone lines as a concern and says the node was intended to support communication in health and education. The June update says the installation was in place and that GreenNet’s London node passed messages to and from the Internet.

This was a scheduled email path: useful, international and materially different from a dial-up customer browsing the live network.

That path was fragile at the local end. In November, the project’s coordinator reported that the server was being moved to Khartoum, where telephone lines were better, and that the email system was then out of commission. The note records a transition and a service interruption; it does not say when the node returned or whether the planned expansion happened. A connection chronology that counts only successful milestones would miss this documented interval when the route was unavailable.

The next evidence is sharper but narrower. On 22 May 1997, Mike Jensen told network colleagues that Sudan was “finally on the net” and named www.sudanet.net, reached via NSN. Three days later, Network Startup Resource Center director Steven Huter recorded ping and traceroute output to that host. Six of seven ping packets returned, an observed 14 percent loss; the replies averaged roughly 716 milliseconds. The route left the US network through Seattle, crossed Verio and MAE-West, entered NSN, and reached the Sudanet address.

That is good evidence of a working end-to-end IP path from the measurement point to one host on that date. It is not a survey of users in Sudan, a test of the local telephone network, or proof that any household or university could buy a stable session. The ping answers the question it was designed to answer: did packets reach this address and return from this vantage point? It leaves access, price, service quality and geographic reach open.

A report for Sudan’s Ministry of Agriculture, Natural Resources and Animal Wealth supplies a different operational view. The consultancy mission ran from 11 September to 18 October 1997; the report was issued by the Food and Agriculture Organization in February 1998. It says Sudanet had been established in March 1997, that NSN supplied the Internet connection, and that preparations included a satellite ground station, servers, routers and a dial-up network. Its stated public-service start is October 1997, with 250 subscribers using browsing, email, file transfer, NetMeeting and chat.

The same report makes the subscriber figure less sweeping than it first appears. It was examining agricultural documentation and research institutions. At the time of the mission, it said email and Internet connections were not installed at the surveyed information units; online Internet access remained a goal for the ministry and the Agricultural Research Corporation. So a report could describe a public service with 250 subscribers while the agricultural institutions it visited still lacked the connection needed for their own work. “Public service” did not mean uniform institutional availability.

Other accounts complicate the chronology. A 2003 Sudan Internet Society presentation by a Sudatel networking manager retrospectively says Sudanet introduced Internet in 1996 with 128 kbps of capacity. It dates Sudatel’s own service to early 1998 at 265 kbps, describes a June 1998 agreement under which Sudanet became Sudan’s only ISP, and says that by November 1999 Sudatel was the country’s only point of presence with a 2 Mbps link from EMIX, opening service provision to other providers, enterprises and universities.

A Central Bank of Sudan report later says the company was established in December 1996 and that the international network was introduced in March 1997.

These statements may describe incorporation, an international circuit, technical introduction, a public retail launch and a later wholesale arrangement. The records do not explain how those milestones fit together, and the 2003 presentation does not supply the agreement or a contemporary service log. The differences should remain visible rather than be repaired by choosing the cleanest date.

A final contemporary trace warns against treating a network diagram as a service guarantee. In November 1999, an NSN representative told the NSRC that the Sudanet link appeared to have been down for at least a month. He wrote that Sudanet and NSN were working through challenges and that the date for stable Internet service was unknown. That is evidence about the particular link and the provider’s own view at that moment—not proof that every Internet service in Sudan had stopped. It also sits uneasily beside the later presentation’s account of a November 1999 point of presence. The sources do not reconcile those accounts.

The sequence matters because “connected” names several different conditions: email can cross borders on a schedule; a router can answer an external probe; an ISP can advertise access; an institution can obtain a line; and the line can remain usable. A packet trace establishes one of these conditions unusually well. It cannot substitute for the others. Sudan’s early Internet history is therefore best read as a set of partially overlapping paths, services and operating claims—with the boundaries, interruptions and unresolved dates left in the record.

Sources