Summary

  • Czechoslovakia’s EARN connection and its Internet connection were related but not interchangeable. Early in 1992, one international route was divided into independent channels so the established EARN service and the newer IP service could coexist.
  • Jan Gruntorád helped turn political opening after 1989 into a sequence of operational commitments: an EARN organisation, an experimental Linz connection, the official launch on 13 February 1992, ministry funding and the nationwide CESNET academic backbone in June 1993.
  • The episode shows that connectivity arrives in layers. Permission, protocol, circuit capacity, operating responsibility and domestic distribution each had a different owner and date.

A circuit with two answers

An international line can look unambiguous on a map. Prague is at one end, Linz at the other, and a stroke between them appears to answer the question: connected or not? The equipment in Prague in early 1992 gave a different answer. CESNET’s institutional history says the route was doubled and split into two independent channels. One remained attached to EARN, the European branch of BITNET. The other was set aside for the Internet.

The distinction was not semantic housekeeping. EARN had already brought valuable international academic services to a Czechoslovak mainframe. Users could exchange electronic mail and retrieve resources through a store-and-forward environment. Internet Protocol connectivity offered another operating model: end-to-end communication among networks, with different routing, addressing and application possibilities. Keeping both channels alive acknowledged that a new architecture rarely replaces an established service at the instant it becomes available.

That coexistence is the centre of Gruntorád’s story. He is often called the father of the Czech Internet, a nickname that compresses years of collective work into a family metaphor. The line itself resists that compression. It records migration as it was: political constraints changed, an institution accepted responsibility, a foreign university provided a path, an older network kept serving users, and a newer network had to earn its own capacity.

Permission came before bandwidth

Gruntorád’s later account begins in Denmark. During a research stay in the late 1980s, he saw conference participants using terminals to send email. When he asked for an address, he was told that it would be of little use because there would be nobody in Czechoslovakia to correspond with through the system. He returned to Prague and persuaded his university to request a connection to EARN. According to his recollection, the reply was polite but negative: political regulations made a connection impossible.

That episode should be read with the correct evidentiary weight. The surviving public source is a retrospective biography carrying Gruntorád’s account, not the original refusal letter. Yet it captures an important control point. The obstacle was not simply a missing modem or a weak line. Under the communist state, international data communication was also a question of permission.

The Velvet Revolution of November 1989 altered that condition. It did not create a network by decree. It made the next acts possible. In 1990, Gruntorád became the first national director of EARN and led the establishment of its Czechoslovak presence. CESNET’s timeline places the country’s connection to the American university network BITNET, through EARN, in that year. Political change opened the gate; engineers still had to build and operate what passed through it.

EARN was a beginning, not a synonym

In 1991, an IBM mainframe at the Czech Technical University in Prague was connected to EARN at 9.6 kb/s. CESNET describes the service as locally available but as the country’s first opportunity to reach worldwide network infrastructure. That sentence holds two truths at once. The link was transformative, and it was bounded.

It is tempting, looking backward from a world dominated by IP, to rename every earlier international network connection “the Internet.” Doing so erases the very decision that mattered. EARN and BITNET had their own architecture and operational culture. A CESNET anniversary recollection describes the later CESNET network as bringing full end-to-end access that replaced earlier mediation through email and the EARN Trickle server. The service before IP was real; it was simply not the same service.

Late in 1991, the Prague team established an experimental Internet connection through the EARN node at the University of Linz. Official accounts place the experiment in November or describe a successful link in December. The variation is a warning against forcing an experimental process into a ceremonial timestamp. What is clear is the order: the EARN relationship and circuit provided an institutional and physical bridge; Internet service was then tested across it.

Nineteen point two, divided in half

The official connection is commonly described as 19.2 kb/s. CESNET also records a preceding 9.6 kb/s experimental connection. These figures are not rival claims when the configuration is reconstructed. At the beginning of 1992, the transmission route was strengthened to twice its earlier capacity and split. Half served EARN–BITNET, and half served the Internet.

This was an architecture of transition. The team did not strand existing users in order to present a cleaner origin story. Nor did it allow the older service to consume the whole route indefinitely. Capacity was allocated to two networks with different assumptions, giving the new IP service an operating space of its own.

Scarcity made that allocation consequential. Gruntorád later recalled that the Czech Technical University paid 174,000 Czechoslovak koruna a month for its connection in 1992, when an average monthly salary was about 4,000 koruna. The comparison is a recollection rather than an audited tariff schedule, but its scale explains why a channel was governance, not merely configuration. Every fraction of the line embodied a budget decision about which community and which architecture could grow.

What 13 February actually changed

On 13 February 1992, the Czech Technical University hosted the official opening of Internet services in Czechoslovakia. The date deserves its place in public memory. An experimental path became an acknowledged national capability, presented in front of technical and institutional witnesses. Gruntorád’s team had made the demonstration possible; he organised the opening.

But “official” should not be confused with “first.” Czechoslovakia had already joined EARN. The Linz experiment preceded the ceremony. The split route was an operating fact. The event therefore marked a change in legitimacy and commitment: the Internet connection could now be treated as infrastructure worth expanding, not only as a technical trial conducted by a small group.

The consequences appeared in a funding decision. Plans for a countrywide academic backbone had circulated in 1991, but CESNET’s account says the February demonstration helped make the proposal intelligible to a government occupied with a far broader political and economic transformation. In June 1992, the education ministry released 20 million koruna. The team had asked for 26 million; the smaller sum was still enough to begin.

The ceremony mattered because it moved another gate. It translated a functioning connection into an institutionally fundable project.

From one border crossing to a national network

The next milestone came after Czechoslovakia itself had divided. On 15 June 1993, the Czech Technical University began operating the nationwide CESNET academic network. It connected ten university cities in Bohemia and Moravia. The main Prague–Brno route ran at 64 kb/s, while the early network’s other connections were measured at 19.2 kb/s.

Internet Hall of Fame credits Gruntorád with leading the choice to base CESNET on IP. That decision mattered more than the initial speeds. A national backbone turned one international connection into a shared domestic distribution system. Universities no longer needed to solve the foreign-border problem independently; they could connect to a common network built for research and education.

The institution and the network must also remain separate in the chronology. The network called CESNET began in 1993 under the Czech Technical University. The association of the same name was founded in 1996 and then took over its operation. Gruntorád’s work crossed both forms, which is precisely why reducing it to the launch ceremony understates his contribution.

CESNET was initially the only practical path to the Internet within the country. Its spare capacity was later offered for commercial purposes, allowing private-sector users to connect. Consumer access had not been the original mission; a research network’s unused capacity became a bridge into a wider market. Infrastructure built for one constituency acquired public significance because its operators found a controlled way to admit another.

The contribution was sequencing

Gruntorád did not invent EARN, IP or the international circuit, and he did not act alone. The Czech Technical University, the University of Linz, IBM, EARN colleagues, ministries, funders and numerous engineers each controlled part of the path. His distinctive role was to keep the sequence moving across their boundaries.

He converted a glimpse of email into an institutional request. After political change, he converted permission into an EARN organisation. His team used an established academic relationship to test Internet connectivity. They preserved the old service while allocating capacity to the new one, made an official launch credible, and used that legitimacy to support a national build. Later work around NIX.CZ and CZ.NIC continued the pattern of giving shared technical functions durable institutional homes.

The title “father” suggests origination. The record points instead to orchestration: knowing which gate was closed, which actor could open it, and which operating arrangement had to exist before the next request would be credible. Czechoslovakia did not cross from isolation to the Internet in one leap. It moved through a chain of partially overlapping networks—and someone had to prevent the chain from breaking between stages.

Sources