Summary

  • Bulgaria’s research network grew alongside, and sometimes separately from, commercial EUnet access; email, a private university circuit and Internet reachability were different operating states.
  • Later records describe wider academic links by 1995, but contemporary reports disagree about whether those links reached the public Internet. A retrospective network history says state funding stopped in 1997 and external connectivity was lost.

“Connected” is too blunt a word for Bulgaria’s early research networks. A computer could exchange email through EARN, an academic body could hold a private line to a foreign university, and a commercial provider could offer TCP/IP access to customers, all without those paths forming one Internet-reachable academic network. The records tell a more useful story when each path is named and dated.

A 2003 European research-network report traces a Bulgarian research-network prototype to 1988. A historical study by Juan Rogers and Ivan Tchalakov says a Bulgarian Academy of Sciences laboratory established an email connection with an Austrian institute in Linz that year, through the European Academic and Research Network. That was an important international research connection, but the study describes a DNA-based network, not a national TCP/IP Internet gateway. The distinction matters: an early cross-border message route did not by itself make the later Internet available to every institution.

In April 1992, RIPE’s account listed two parallel routes. The Bulgarian EARN node in Sofia used dial-up X.25 to Linz; EUnet’s Bulgarian network reached the Internet through its Varna backbone and an EUnet node in Greece. The report said the projects were considering coordination around a shared fixed IP connection. Its 1993 update described EARN at Sofia University and about ten academic institutions, but said there was no gateway between EARN and EUnet. Network communities could occupy the same country and still operate as separate systems.

Digital Systems’ own chronology dates its first full TCP/IP Internet service to 1992; that commercial milestone does not establish that EARN and UNICOM-B were then reachable from the Internet.

A November 1994 update to the NSF-supported connectivity database makes the split more concrete. EUnet Bulgaria manager Daniel Kalchev called EUnet the only ISP then providing Internet access in the country. He described UNICOM-B’s academic connection as operational for email only, through the BGEARN BITNET node. In February 1995 he added that a 9.2 kb/s Sofia–Vienna line, if in service, was private between UNICOM-B and the University of Vienna, leaving UNICOM-B unreachable from the Internet. In the same message he described EUnet’s 19.2 kb/s Amsterdam–Varna line and domestic backbone.

This is a provider’s contemporary account, not a neutral audit, but it distinguishes commercial Internet service from academic connectivity with unusual precision.

A later inventory complicates the picture. SERENATE’s 2003 profile says the academic network had two 64 kb/s Amsterdam circuits and a 64 kb/s Sofia–Vienna satellite channel at the end of 1995, serving twelve cities. Yet the DANTE-managed PHARE programme still listed two Sofia–Amsterdam 64 kb/s terrestrial circuits as planned in December 1996. The surviving records do not establish whether these were the same circuits, different project phases, or different definitions of the academic network.

Capacity listed in a retrospective profile cannot resolve the 1995 contemporary report that the academic route was private, or the later project table that still called Amsterdam links planned.

The clearest break comes in SERENATE’s account: early in 1997 the Ministry of Education and Science stopped funding the network, and its external connectivity ended. A July 1997 local description still reported an academic network: three large Sofia universities shared a 100 Mb/s FDDI ring, while many other participants reached Sofia over analog links. The evidence therefore supports a loss of external academic reach, not a nationwide Internet outage or the disappearance of every campus network. Commercial access continued on its own track.

The organizational timeline is also unsettled. SERENATE says the ministry established a legal entity called UNICOM-B at the end of 1998, which lasted until December 2001. Rogers and Tchalakov describe UNICOM-B’s academic-provider role as ending in June 1998. These accounts may refer to different forms of activity, but they do not prove an uninterrupted service or show that a new legal status restored the old international route. A later name, a local ring and a reachable external network are separate facts.

The useful unit of history here is not the country’s first connection date. It is the path that a particular institution could actually use: its protocol, gateway, upstream, funding and local reach. Bulgaria’s records show why a circuit count or organizational title cannot stand in for that evidence.

Sources