Summary

  • Datacom’s Internet gateway depended on a loan for the satellite ground equipment and a separate, conditional offer to help pay the recurring satellite-channel bill.
  • A January 5 network message and the January 17 public ceremony mark different steps: the gateway was reachable in one test before Datacom’s public opening, while local access still had to be built and financed.

On January 5, 1996, Steve Goldstein at the National Science Foundation forwarded a terse engineering update: “The Mongolian Link is up and running!” The same message carried a qualification. Sprint could reach Datacom’s router in Ulaanbaatar, but Datacom still needed to add its default route back to Sprint. The reported test sent 50 packets and received all 50, with an average round trip of 719 milliseconds. It was evidence of one working path to one gateway, not a measure of service for Mongolian users. (NSF message archived by the Network Startup Resource Center)

Twelve days later, Datacom held an opening ceremony attended by President Punsalmaagiin Ochirbat and officials from government, science and education. The company’s director, Dangaasuren Enkhbat, reported more than 100 guests from 60 organizations. The embassy’s account says the equipment had been installed in December and tested by Sprint engineers before the event. The technical connection and the public launch were related, but they were not the same milestone. (Enkhbat’s 18 January message; U.S. Ambassador Donald C. Johnson’s account)

The path to that ceremony began before the dish arrived. In late 1994, Datacom connected its domestic message system to the United States over a dial-up UUCP gateway. Email and Usenet newsgroups moved in batches, initially twice a week; the schedule became more frequent as users increased. Datacom wanted a permanent connection, and the later IDRC report describes a 128-kbit/s leased satellite link through PanAmSat 2, arranged with Sprint. That was a change in operating conditions: instead of waiting for the next mail exchange, a connected host could use full Internet services, including the Web. (IDRC’s 1997 account of the Pan Mongolia project)

The financing was also a two-part design. One part paid for equipment. The U.S. Embassy and Mongolia’s Ministry of Trade and Industry administered a revolving loan fund built from proceeds when the Mongolian government sold surplus butter and butter oil donated by the U.S. Department of Agriculture. Ambassador Johnson said Datacom first borrowed to obtain a Sun computer, then sought a loan for the satellite dish and ground station. He estimated those items, transport and installation at about US$100,000.

Approval came in August or September 1995; the equipment was ordered from Comstream and reached Mongolia in late November, with shipping delays before the December installation.

The second part addressed the bill that would return each month. Johnson said Steve Goldstein expected the NSF might cover a satellite-channel charge of US$5,000–US$7,000 a month if Datacom offered very low-cost service to a defined scientific community: the Academy of Sciences, the Ministry of Nature and Environment, the National University and the National Technical University. The later IDRC account describes the condition as free access for educational institutions during 1996/97. The wording and duration differ, and neither account is an audited record of payments. A 2022 Mongolian-language industry retrospective reports a US$50,000 butter-fund loan at 4.8%; that figure should not be added to the Embassy’s broader equipment-cost estimate as if they measured the same thing. (Gemnet retrospective)

That distinction matters because a dish and a circuit solve different problems. A one-time loan could buy an asset that remained in Ulaanbaatar. It could not make the monthly satellite service free to operate. A sponsor could help keep the international route available, but the conditional access offer applied to a named institutional group, not every school, company or household.

The launch also did not remove the local network bottleneck. The IDRC account describes unreliable telephone lines, erratic power and a small number of computers. It says the satellite node created a base for later technical and content projects, while local infrastructure still had to spread access. The Washington Post’s contemporary report presented the connection as a way to reach libraries and communicate with researchers, but those ambitions should not be mistaken for a completed national network. (The Washington Post, 29 February 1996)

Mongolia’s 1996 connection was therefore more than a ceremony and less than universal access. A recoverable local loan assembled the gateway; conditional outside support helped address its recurring international cost; and a local operator had to turn that route into services people could actually use. The link was a beginning whose reliability depended on who paid the next bill and who could reach Datacom’s equipment.