9.1 Time-sharing and resource-sharing demand
Interactive computing created pressure to connect researchers with remote machines and specialised resources rather than duplicate every expensive computer at every site.
9.2 Network planning
ARPA-sponsored researchers developed functional requirements, host interfaces and a packet-switching subnet architecture during the late 1960s [1][2].
9.3 First RFCs
RFC 1 in April 1969 described host software before the full network existed, making open technical discussion part of the infrastructure itself [1][4].
9.4 IMP deployment
BBN built IMPs that connected hosts to leased lines and performed packet switching. The first installation occurred at UCLA in 1969 [4][5].
9.5 Four-topic network
UCLA, SRI, UC Santa Barbara and the University of Utah formed the first ARPANET cluster by the end of 1969 [4].
9.6 NCP operation
Host-to-host communication stabilised around NCP, supporting Telnet, file transfer and other applications inside ARPANET [3][6].
9.7 Email and collaboration
Network mail quickly became a major use, revealing that shared communication could matter more than remote machine access alone.
9.8 Protocol strain
Growth exposed retransmission, sequencing, duplicate and multipacket-message problems, motivating protocol reassessment [6].
9.9 Transition pressure
Packet radio, satellite and other networks could not simply become ARPANET segments. Internetworking required a new architecture.
9.10 NCP-to-TCP transition
RFC 801 documented a coordinated transition from NCP to TCP/IP, demonstrating that protocol migration is institutional choreography as much as software replacement [7].