The UNIX Time-sharing system

From: Greg Green (ggreen_at_cs.washington.edu)
Date: Tue Jan 06 2004 - 19:43:31 PST

  • Next message: James Welle: "The UNIX Time-Sharing System"

    This paper describes the design and implementation of UNIX in 1974. At
    the time it was operating in about 40 installations of PDP-11
    systems. The file system was described in some length. The main
    differences between the system as described and modern systems was the
    small file size and only having user and other permissions. The main
    elements were essentially the same as they are now. Process management
    and pipes for IPC were described. The main concepts of a shell were
    described, running other processes, redirecting output, filters, and
    backgrounding processes. The beginnings of the signal subsystem are
    described, but obviously still in a nascent state.

    I have read a couple of books on UNIX design and implementation, so I
    found nothing particular interesting in this paper, other than being a
    concise explanation of some of the design decisions. It was
    interesting that there was some data on reliability of the system, ie
    longest up time of 2 weeks and a 98% average up-time. About on par
    with modern Windows NT systems, but much worse than what we see with
    Linux systems. I suppose that is due to the fact that the system was
    so small (42K bytes) and the implementers were competent people.

    A point was made that the system was successful because it just grew,
    instead of being designed for a particular purpose. He also states
    that since they are programmers, the system was designed to support
    programmer users, as opposed to the general population. That hasn't
    changed much in the intervening years.


  • Next message: James Welle: "The UNIX Time-Sharing System"

    This archive was generated by hypermail 2.1.6 : Tue Jan 06 2004 - 19:43:33 PST