2020-09-07 12:01:18 UTC
in an undergrad internship) was on a naked PDP11 with no OS.
That starting position is where I'm at, and have been now
for nearly 50 years, in that I feel comfortable with full
and in-depth knowledge of all software running in any computer.
That state of play makes one very efficient in resolving bugs
because one can drill down to the deepest depths to
understand what has gone wrong.
This was even true with the RSX11-M OS, having been on the
System Programmer's Course (The Butts Centre, Reading, UK
in 1978) where the full internals of the OS were revealed.
That state of play lasted until 1996 when working on an
embedded 80186 in a PABX.
I suspect that it is no longer feasible to have that
background in a computer system of today, resulting in an
unease about the efficacy of any software (especially
with W10, QV :-) )
In fact, 13 years ago, in a contract in the automotive
world, I encountered resistance to my wish to ensure
the safety of the changes that I proposed because when
I sought to trace the effect of the changes into other
part of the system, I was denied access to the source
code of object models that had already been signed off.
(ISTR that the Object systen in use was one by Rational)