Re: [feyerabend-project] Systems Software Research is Irrelevant
> Here is a URL to the homepage of Rob Pike, and I have found his slidesI remember reading this when he first posted it. I think it is sadly
> about "Systems Software Research is Irrelevant" very interesting. What
> do others think about it?
One thing Pike mentions is that every project feels the need to first
demonstrate compatibility with the huge weight of existing software,
or else nobody will take them seriously. I think of four OS research
projects that were all interesting in different ways -- Mach, Sprite,
Amoeba, and Plan 9 -- and yet they all established Unix compatibility
very early. Think how much more innovative they might have been if
they hadn't all had as a goal the ability to run Unix programs. (Mach
researchers made much of its ability to support multiple OS
"personalities" atop its microkernel core, but so far as I know the
only personalities that were ever produced were variants of Unix.)
I actually thought of Pike's talk last week while rereading --
strangely enough -- _A Deepness in the Sky_, a science fiction novel
by Vernor Vinge. There's a passage where he talks about the many and
ancient layers of software in the systems on their spaceships, and it
culminates with this:
Take the Traders' method of timekeeping. [...] down at the very
bottom of it was a little program that ran a counter. Second by
second, the Qeng Ho counted from the instant that a human had
first set foot on Old Earth's moon. But if you looked at it still
more closely ... the starting instant was actually about fifteen
million seconds later, the 0-second of one of Humankind's first
computer operating systems.
The novel takes place "thousands of years from now," and the Unix
epoch is still in place as the basis for a computer's notion of time.
It's just a novel, but I have a hard time convincing myself that
Vinge's wrong about that.