Multics Timeline

By michelele

  • John McCarthy proposes time-sharing on MIT's IBM 709

    John McCarthy proposes time-sharing on MIT's IBM 709
    John McCarthy writes his memo to Philip Morse proposing time-sharing as a way to make computing more efficient. https://multicians.org/project-mac.html

  • Fernando Corbato (Corby) begins development on CTSS at MIT

    Fernando Corbato (Corby) begins development on CTSS at MIT
    Fernando Corbato and his team begin work on CTSS and complete the first test of the system in November of 1961. They were able to support four users on the same length of tape. https://multicians.org/thvv/7094.html

  • J.C.R. Licklider and Professor Robert Fano's meeting

    J.C.R. Licklider and Professor Robert Fano's meeting
    Licklider and Professor Fano meet to discuss IPTO funding a collective of computer researchers at MIT. Licklider proposes $5 million over two years and forms Project MAC. The Computer Systems Research Group begins to plan a second-generation time-sharing system. https://multicians.org/project-mac.html

  • Project MAC Summer Study

    Project MAC Summer Study
    The official start of Project MAC was the 6-week summer study in 1963. This exposed 57 visiting researchers to the MIT's CTSS and SDC's Q-32 time-sharing systems. https://multicians.org/project-mac.html

  • Multics takes shape

    Multics takes shape
    Planning for Multics had begun before Project MAC had started. Visits to vendors were made in 1963 to select hardware, and GE was chosen in August of 1964. Bell Labs became associated two months later. https://multicians.org/project-mac.html

  • GE-635 Delivered to MIT

    GE-635 Delivered to MIT
    GE was chosen for their 635 computer, which impressed Project MAC for its use of multiple CPUs. This would allow symmetric multiprocessing, which was a proposed feature of Multics.

  • 1965 Fall Joint Computer Conference

    1965 Fall Joint Computer Conference
    The first papers drafted describing and naming Multics are presented. CTSS will be used as a development tool for the new operating system. The goals of Multics are also stated: continuous operation, a wide range of configurations, an internal file system, and tools for systems administration. https://multicians.org/project-mac.html
    https://multicians.org/history.html

  • MIT Receives the GE-645

    MIT Receives the GE-645
    The GE-645 was finally delivered to MIT in January of 1967. This confused Multics development due to the length of time spent using the much slower GE-635. For a length of time, the 645 had to emulate the 635 in order to run the 645 emulator used to debug Multics. https://multicians.org/ge635.html

  • Multics' first boot

    Multics' first boot
    Multics planning notebook milestone 1 was achieved in December 1967. This milestone required the system boot the operating system in a single process on the bare hardware of the GE-645

  • Multics development becomes self-hosting

    Multics development becomes self-hosting
    By December of 1968, Multics was being entirely developed on the GE-645. Developers were using the MIT 645 as a time-sharing service, phasing out the CTSS. https://multicians.org/ge635.html

  • Bell Labs withdraws and Multics releases

    Bell Labs withdraws and Multics releases
    Bell Labs chose to withdraw from the Multics project in April of 1969. The time of their contract with MIT had passed, but Multics was not ready to use. Multics released to MIT customers in October of that year, with numerous problems related to development. https://multicians.org/history.html

  • GE Sells its computer division to Honeywell

    GE Sells its computer division to Honeywell
    General Electric sells their computer division to Honeywell and finalizes the merger in 1970. Honeywell gains access to the former GE properties, including the integrated circuit based GE-655 and Multics. https://multicians.org/history.html
  • Period: to

    Multics connects to the ARPANet

    Project MAC programmers worked to develop software which connected Multics to the ARPANet. The GE-645 was connected to the ARPANet in 1971, and the MIT 6180 was connected in 1974. https://multicians.org/history.html

  • Multics' commercial announcement

    Multics' commercial announcement
    Several commercial announcements for Multics were made in 1973 after securing major customers like Ford and the US Air Force. The Honeywell Model 6180 was announced in January 1973 running Multics. https://multicians.org/history.html

  • Honeywell makes decision to cancel Multics

    Honeywell makes decision to cancel Multics
    Due to aging hardware architecture, Honeywell was not able to keep up with other systems in the 1980's. In November of 1984, Gene Manno decides to cancel Multics in July of the next year. https://multicians.org/history.html

  • Multics achieves a B2 rating

    Multics achieves a B2 rating
    In 1985, Multics acquires a B2 rating, allowing top secret information to be kept on the operating system. Growing out of support from ARPA and contracts with the US Air Force, Multics slowly became hardened enough to receive a B2 rating from the Department of Defense's "Orange Book". Ironically, experience hardening Multics in the 1970's informed the writing of the very same book. https://multicians.org/history.html

  • Bull releases source code

    Bull releases source code
    In 2007, rightsholder Bull HN released the source code for Multics at MIT.

  • Multics CPU Simulator is released

    Multics CPU Simulator is released
    The Multics CPU Simulator is released, allowing the boot of Multics on non-Multics CPUs. The most recent Multics CPU Simulator update was released in August of 2023.