011.30 BRAZIL by Eduardo Tadao Takahashi Brazil was incredibly late to get plugged into the academic networking world. The first networking efforts in the country started in the early 80's, led by LARC, a consortium formed by major universities and research centers for the exclusive purpose of advancing such activities. However, these efforts were severely undermined by many problems, not the least of which was the restriction that the telecommunications law in Brazil imposed until recently upon data communication services: in a strict interpretation, they were a monopoly of the union. It was only in 1988 that two BITNET connections to the US were established from Brazil, one by FAPESP (Sao Paulo State Research Funding Agency) and another by LNCC (National Laboratory for Scientific Computation), while a favorable interpretation of the law allowed interested academic institutions to get connected to either of the two pioneering gateways. Within two years, most universities and research institutions managed to get connected to what became a national network. A third connection to the US was set up at about the same time by UFRJ (Federal University of Rio de Janeiro), but it would be used mostly for UFRJ alone, offering BITNET and HEPNET services to Brazilian researchers at negligible cost to end user institutions. While the academic community gathered around FAPESP and LNCC, a non- governamental organization in Rio, IBASE, obtained support from UNDP to set up a node of services at its site and through an UUCP connection to IGC in San Francisco. In 1989, the Brazilian Research Network (RNP) was formally launched by the Ministry of Science and Technology, having as its main purpose: . in the short-term, to integrate the then scattered networking efforts of the academic community in Brazil; . in the medium-term, to plan and direct the consolidation of an academic network in the country, interconnecting regional networks and individual institutions, and providing direct connections to major networks in the US, Europe, and neighboring countries in Latin America. . in the long-term, to help devise and implement a national strategy for the deployment of informatics resources in education and research, ranging from high-performance computing to K12 education. The RNP initiative was organized as a project under the direct coordination of the National Research Council (CNPq), which has financed most activities since its inception. As of 1991, RNP approaches the conclusion of its first main objective, that is, the installation of a nationwide 9.6 - 64 Kbps backbone with points- of-presence in twelve major cities in the country, supporting multiple protocols and running TCP/IP as its lingua franca. As the backbone starts to operate, most institutions are expected to deactivate their current connections (which are too low speed and badly engineered as a rule) and seek a link to the nearest point-of-presence. In February, a National Operations Center (NOC) will start working at FAPESP, taking responsibility for the management of the national backbone and its connection to other networks abroad. At the same time, a National Informations Center (NIC) will start to operate at LNCC, dedicated in the beginning to the preparation and dissemination of training materials. The NIC and the NOC are being organized with the partial support of IBM, ORACLE and INTERLEAF, which are donating hardware, software, and services. Complementing the efforts of RNP at the national level, several states in the country (most notably Rio de Janeiro, Rio Grande do Sul, and Sao Paulo) have been pursuing their statewide networks, which are integrated through the national backbone. Thus, by the end of 1992, one expects to have in Brazil a two-level network composed by some 100 nodes, mostly supporting INTERNET services, while some remaining as BITNET/BITNET-II installations. In 1992, RNP will start to deal with a number of issues, including: . faster connections, as the local public infra-structure gets improved; . integration with existing commercial services (eg, X.400 messaging) and with other national projects (eg, library access; copying service; automation of the federal administration along GOSIP lines, etc.) ; . experimentation with low-cost technologies such as packet satellite communications; . systematic experimentation with packages such as ISODE, PP, QUIPU, etc., as a preliminary step to the launching of a concerted effort in topics such as experimental X.500 services. In future notes, we shall discuss these issues as seen from the Brazilian perspective, which in a way is unique: not many national networking initiatives have to simultaneously cope with first world requirements and wishes and "fourth world" needs and priorities. Eduardo Tadao Takahashi, Rede Nacional de Pesquisa - CNPq, Cidade Universitaria, Campinas