shell architecture (to glob or not to glob)

[Brandon S. Allbery KB8JRR]

As quoted from <4f5c65f2.20b6d at apollo.HP.COM> by vinoski at apollo.HP.COM (Stephen Vinoski):
+---------------
| In article <4584 at lib.tmc.edu> jmaynard at thesis1.hsch.utexas.edu (Jay Maynard) writes:
| >for a user. It took me two years of running a Unix system at home before I
| >got comfortable with it, and I'm a systems programmer by trade. How long
| >does it take a user?
| 
| Thank you for reinforcing my opinion of those who call themselves "systems
| programmers."  Two years?  Sheesh!
+---------------

At the other end of the scale, I picked up Unix --- not just using the shell,
but also the basic "tools" philosophy which makes it what it is (or at least
*was*) in three weeks.

But don't be *too* hard on people.  My mother was a comfirmed computer
illiterate and was getting absolutely nowhere with a computer at work until I
taught her my "secret":  that you can't hurt the computer by trying things.
(Within reason, of course.  Don't try "rm -rf /" at home, kids!  ;-)  But
then, if what you're trying is OS/shell features, up-to-date backups are a
very good idea anyway.)

I can learn to use a program that's completely new to me in less than a week
by setting up a "play area" --- then "doing my worst" in that area.  For
example, fire up a word processor or spreadsheet on a new file or on a copy of
an existing one, then start trying out commands.  I usually get a new OS to
play with on a new computer, so I load the system and play with it --- then
reload it once I know what's what, usually tuning the reload for the system's
intended purpose.

In learning computers or computer programs, caution is a good idea --- but TOO
MUCH caution is almost as bad, because (like so many other things in the
world) manuals are no substitute for experience.  But with computers, you can
get that experience if you're willing to sit down and just try things.  The
problem is that, to a novice, computers are awfully expensive devices whose
internals are completely incompresensible, and there is a quite natural fear
that doing even the slightest thing "wrong" will wreck it beyond repair.
Similarly for programs, operating systems, etc.:  the fear is always there
that doing something "wrong" will ruin it.  But all it takes is a little
caution to make sure nothing important will be touched --- then you can
discover that computers and computer programs are far more forgiving than most
other things.  After all, a mistake while learning how to drive a car can kill
you --- but a mistake while learning how to use a computer can be easily
corrected if a little common sense is used beforehand.

BTW, I don't know if *any* of this applies to Jay; the real world contains an
awful lot of complications.  As an example:  I regularly crashed ncoast when I
first discovered it (1981), because the OS it was running was extremely buggy.
This would have made me extremely hesitant to "play" with it, except that I
caught one of the OS bugs in action.  [The system panicked whenever the
process IDs rolled around.]  Nor are computer/OS/program bugs the only source
of complications --- Murphy is always looking for a chance to step in and
wreak havoc in otherwise simple situations.

++Brandon
-- 
Me: Brandon S. Allbery			    VHF/UHF: KB8JRR on 220, 2m, 440
Internet: allbery at NCoast.ORG		    Packet: KB8JRR @ WA8BXN
America OnLine: KB8JRR			    AMPR: KB8JRR.AmPR.ORG [44.70.4.88]
uunet!usenet.ins.cwru.edu!ncoast!allbery    Delphi: ALLBERY