In 1995, Brazil was introduced to a technology with which it would later fall in love. For the first time, the common citizen could have access to the internet through the dial-up IP modality. Due to the Internet Management Committee (CGI.Br) creation, in May 1995, the 20 years of the commercial internet in Brazil are celebrated. Back then, the period was instable for the telecommunication sector. However, what is rarely discussed is that the internet existed in the country long before this, in the academic environment, where it was generated and found means to develop, in the molds we know today.

Seven years prior, in 1988, the São Paulo Research Foundation (in Portuguese, FAPESP) and the National Laboratory for Scientific Computing (LNCC) were two pioneering institutions in connecting to Bitnet, a technology that preceded the internet and the World Wide Web, which allowed the transference of text files through electronic mail. A year later, the National Research Network arose, financed by the National Council for Scientific and Technological Development (CNPq), with the goal to spread the use of networks in the country, under the general coordination of Tadao Takahashi.

In the following months, Tadao appointed Demi Getshko, a FAPESP IT Manager at the time and currently NIC.Br President, to coordinate the Operations. In addition to FAPESP in São Paulo, led by Getshko, and the LNCC in Rio, where RNP Former Engineering and Operations Director Alexandre Grojsgold stood out, other institutions in Rio quickly connected to Bitnet: the Federal University of Rio de Janeiro (in Portuguese, UFRJ), led by professor Edmundo Souza e Silva, and the Pontifical Catholic University (in Portuguese, PUC-Rio), where the current R&D Director Michael Stanton was a professor.

“There was an interesting environment there, because the university had local networks and we started to interconnect them. So, it was possible to communicate between departments, which was something very new”, Michael recalls. “Starting from 1990, we had an external connection for this set, which allowed accessing the LNCC”, he stresses. Quickly, the novelty spread to other universities and, in 1991, RNP already connected 40 institutions to Fapesp and the LNCC, which had external connections for Bitnet.

While Bitnet came to connect thousands of universities throughout the world, but had limitations to transfer data, in the United States another technology had been developed for three decades, able to support various applications, TCP/IP. This technology picked up steam in 1985, when the National Science Foundation (NSF) financed the construction of a backbone interconnecting the North American universities.

The financing was maintained for ten years, until 1995, when the pressure from the telecommunications to make the internet a profitable business grew, and there was no need to finance an academic network with research grants anymore. Facing this, the reaction from universities was to create Internet2, currently in operation. The main justification was that the service quality provided by commercial operators did not meet the researchers’ needs. In Brazil, the same movement occurred, and RNP felt the pressure from the telecommunications sector for the commercialization of the internet. In 1995, the TCP/IP academic backbone already existed, since 1992, with two international outputs, one in Brasília and another in São Paulo, and a capacity that reached 2 MB/s, considered high at the time in comparison to the data networks of the period.

“The transition to TCP/IP was not something predetermined. Each country had its own solution for the data networks. The American, TCP/IP, was the most widespread. They had developed a system that worked, and everybody wanted to copy it, because like this you could speak to everyone”, Stanton comments.

Under the same argument that more band capacity is necessary for the progress of Science, RNP sustained itself amid the turmoil. Holder of the monopoly in the country at the time, the Brazilian Telecommunications Company (Embratel) was slow to understand a key feature that differentiates the internet from telephony: decentralization.

“They wanted to apply telephony how it used to be, and wanted to be the holders of such control. While telephony was vulnerable and dependent from the physical infrastructure, with the internet the connections were supposedly not immutable. If a connection fell, technology found another way to reach the destination without the need of human intervention”, the R&D director explains.

RNP operated commercial traffic for three years, between 1995 and 1998, as an alternative to Embratel, which had finally adopted the TCP/IP technology and was prevented by the government to exercise the monopoly of data network services in Brazil. Its actuation was limited to the provision of backbone services for commercial internet, a role RNP also played during this period. “In many other countries, the academic networks ended up becoming commercial providers, and there was a possibility this would occur in Brazil, once the then Ministry of Science and Technology (MCT), through CNPq, was losing steam to maintain RNP in operation”, Michael adds.

However, with the transformation of RNP into a social organization and the interest from the Ministry of Education to maintain it to cater to universities and other institutions, the Interministerial Program came to life, in 1998. To RNP General Director Nelson Simões “the decision on the assembly of PIRNP is due to the strategic vision of MCTI, which saw the internet as an important element for the future of education and research, and, in large part, to the need of the Ministry of Education (MEC) to enable access to electronic scientific journals”. Thereafter, RNP could return to its essence: to keep contributing for the innovative use of advanced networks in the country and for the constant evolution of the internet.