Adam Thornton <athornton(a)gmail.com> wrote:
I liked the Frisch and Limoncelli/Hogan books. Nemeth
less so.
Having taught a system administration grad level class
for many years now, I think there's an evolution of
the profession whereby books that were previously
practically useful are less so now.
Nemeth et al was useful in a very practical sense
(e.g., Pete's example of "fix printer spooler on
HP-UX"), but Limoncelli's books were always stronger
in principles.
Taking it further up the layer of abstractions, I then
found Mark Burgess's works more valuable, but that
very much leaves the realm of vocation (which I
maintain is hard to capture in books anyway) and
goes towards the more theoretical side.
The profession has evolved (or even splintered)
dramatically, with DevOps, SRE, and cloud computing
covering overlapping but different sets of work in
this area. All that is very different from the early
days of administering Unix systems.
I still struggle with the balance of teaching students
basic Unix skills, conveying (true) Unix multiuser
system concepts, translating administrative principles
into containerized, ephemeral environments, especially
at large scale...
-Jan