Very kind words from Brantley and Clem. It's an
interesting notion to regard Unix as gestational until it came
out in public talks (1973) and was exported to universities.
Maybe I could claim to have laid the groundwork for Unix
by causing Multics to be written in PL/I, a language big and
sprawling, like the project itself. That unintentionally
provided plenty of stimulus for thinking small. Ken was
absolutely on his own when he began to fiddle with building
a tiny operating system on the GE 645. I heard about it only
after the fact.
After Multics, I ran interference to keep our
once-burned higher management from frowning too much on
further operating-system research. I was aware that Ken,
Dennis and Rudd were discussing the subject down the hall
from my office, but I did not participate in the
discussions. At the same time, I was noodling over what
would later be called shell pipelines; but I did not come
up with the vivid term "pipe" or a halfway workable
syntax for another three years. While these actions may
have contributed to a welcoming environment for Unix, they
in no way "started" it.
Doug