On Tue, 23 Jan 2018 10:02:13 +0000 (UTC)
David Schwartz <
newsletters@thetoolwiz.com> wrote:
> The fallacy here is that a HS dropout who’s been building homes for
> 30 years could build a home as nice as any Architect just a couple of
> years out of school.
That's not the fallacy. Nobody said anything about high school
dropouts. What was being discussed were people who love to code. My
experience tells me most of these people are college grads, although
non Comp-Sci degrees. For whatever reason, a lot seem to ride bicycles
and/or play musical instruments. I've known only one HS dropout who was
a programmer. The comparison here is NOT between average HS dropouts
and people with relevant engineering degrees.
The other thing is, a software product is a very different thing than a
building.
[snip]
>
> But most run-of-the-mill construction workers have no clue when it
> comes to structural dependencies, legal codes, how to choose a
> foundation appropriate for the ground and terroir, etc.
When I was an electrician's helper, I was busy learning the codes. And
the journeymen I helped sure nuf knew the national and local codes. But
that's not even the issue. The issue is, the set of buildings resemble
each other an order of magnitude or so more than the set of computer
programs. There are no equations for a foundation: There's no
foundation in a computer program: They're not made of concrete.
IIRC there are about 13 security rules you need to follow if your code
is heavily exposed on the net. Most are fairly obvious: Sanitize all
user input, set all pointers on declaration, prevent arrays from
overrunning their bounds, don't try to access out of scope variables.
One doesn't need a college degree to know and understand these rules.
>
> You guys keep conflating human psychology and behavior with
> acquisition of some mythical sort of “knowledge” that’s somehow
> learned through osmosis over time (or even instantly).
I haven't seen anyone in this thread make such a conflation. What *has*
been said is that people who really like to code code, and that very
soon in their career they read books, read websites, read other peoples'
code, engage in discussions, to gain their knowledge. There's nothing
mythical about it, and it's done by extreme interest and putting in the
time, not by osmosis.
=-=-=
I'd like to return to the concept of "run of the mill construction
workers", specifically, the "run of the mill" part. Let's discuss "run
of the mill" Comp-Sci grads, because as a rule they're not like the
people on this list.
Once I advertised a position as my assistant programming a substantial
part of a medical management package. Entry level: I'd teach em. We got
several Comp-Sci grads from UCLA, and also several students and recent
Associate Degree folks from Santa Monica Community College. I gave them
all the exact same test: Pseudo-code a program to take an input file,
capitalize all characters, and output to an output file. You have 1/2
hour.
About 1/2 of the Santa Monica College people substantially completed
the test within a half hour. Exactly one UCLA comp-sci grad
substantially completed the test. Most of the UCLA grads bogged down
after 2 or 3 lines of pseudocode.
In the interviews, the UCLA people glibly buzzworded, enumerated the
languages they knew, and boasted of writing compilers. But when asked
to deliver, they couldn't produce a bare bones proof of concept of the
algorithm basic to most back end processing. These were run of the mill
comp-sci grads.
Let's leave run of the mill people out of the discussion. Folks who
love to code self-select themselves out of the run of the mill category.
SteveT
Steve Litt
January 2018 featured book: Troubleshooting: Why Bother?
http://www.troubleshooters.com/twb
---------------------------------------------------
PLUG-discuss mailing list -
PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
http://lists.phxlinux.org/mailman/listinfo/plug-discuss
(text/plain)