On Tue, 14 Jun 2016 15:50:37 +0200
Irrwahn <irrwahn@???> wrote:

> We could play this game day in, day out for months or even
> years, without having you grasp the basics necessary to
> fully understand the implications of even moderately
> sophisticated C programs.

This discussion is getting theoretical, so allow me to discuss my
theories...

I think Edward's being sucked into a vortex of basics at lightning
speed, and within a couple weeks will know as much about the basics
necessary for Felker PID1 as you do (which would be twice what I know,
I found your explanation really hit the spot). Different people learn
differently: Perhaps Edward learns much faster from doing than from
reading or receiving instruction. If he ends up in the same place, we
have one more knowledgeable person in the group.

I had been programming office automation packages in C for 20 years in
2014, and when I saw Felker PID1, it looked like Greek to me. I had to
have a guy named Joel Rees explain what the hell it was doing, pretty
much like you just explained to Edward. From then on, I knew what
fork() and wait() and waitpid() did, without ever studying them
beforehand.

> It'd be like teaching integral
> calculus to someone not having mastered multiplication yet:
> frustrating for both student and teacher.

I'd use a different analogy, but sometimes, for some people, it's better
to learn the calculus first.

>
> There is no substitute for learning things from the bottom
> up. There is no silver bullet! You *have* to go through the
> basics to even have a chance of understanding more advanced
> concepts.

Yes, but in what order? It's very possible and practical to learn C in
a need-to-know order, rather than studying all the system calls just to
program a PDF watermarker.

A real beauty of this thread started by Edward is it motivated you to
give a very succinct explanation of the Felker PID1. Don't think for a
minute that Edward was the only one who learned from your explanation:
It reinforced my knowledge too, and I doubt I'm the only one.

And your question about "why a shellscript?" got me to think about and
get clear about the entire motivation for my writing the "Manjaro
Experiments."

Everywhere I go, with the exception of the supervision@???
mailing list, I'm one of the most init-knowledgeable people in the
group. I got that way by doing exactly what Edward's doing, and when I
started in September 2014 I knew a heck of a lot less than Edward.

It's called Curiosity Driven Learning, and even though it can seem
crazy to people having all the foundational knowledge, it's probably
the quickest way to acquire knowledge, and soon enough the foundational
knowledge is acquired.

Disclaimer: I have a dog in this fight because my book "Rapid Learning
For the 21st Century" depends in large part on Curiosity Driven
Learning.

This is a really fun discussion. You don't hear anything like it over
on Debian-User.

SteveT

Steve Litt
June 2016 featured book: Troubleshooting: Why Bother?
http://www.troubleshooters.com/twb