1. 45

Intro is almost 11 minutes; direct link to actual start: https://youtu.be/EY6q5dv_B-o?t=658

  1.  

  2. 14

    I was in the audience for this. It was standing room only. During the introduction, when somebody asks “does anyone here not know who Ken Thomson is”, only one person raised their hand.

    The interview was informative, even for people who are familiar with the lore, because it appears to contradict established ideas at certain points:

    • B is described at length but BCPL is never mentioned;
    • Thomson claims that pipes were not the same as the IPC mechanism in that paper (which he implemented but that he found unusable for real tasks);
    • he talks about writing grep but not naming it, which seems to contradict histories of the program that associate it with ed features.
    1. 6

      he talks about writing grep but not naming it, which seems to contradict histories of the program that associate it with ed features.

      The other thing I learned is that he did not, in fact, write grep in a single evening! Hah. I remember hearing the story about McIlroy describing a search program to Thompson, and then Thompson coming back with grep the next day. But I hadn’t heard the part where Thompson had already written grep awhile ago, but just hadn’t shared it yet. (I’m not sure if this was already known or not yet.)

      1. 1

        B is described at length but BCPL is never mentioned;

        He didn’t talk about it by name, but did mention the version of B that ran as an interpreter to fit in memory, which, based on my understanding is what BCPL essentially was?