[Haifux] My "Stop Using (and Teaching) C-Shell and Tcsh" Page

Eyal Rozenberg eyalroz at technion.ac.il
Fri Oct 19 20:06:36 MSD 2007

I'll comment on 2 points:

- bash installation
- many tools vs. few limited tools

AFAICT, the bash installation issue is rather moot. First of all, for 
several (many?) years, bash (and certainly sh, probably for even longer) 
has been installed on t2 and tx. Secondly, if Kimchi or any other 
faculty member in the department, or even a large group of students, 
would have asked for it, I don't think the computer center would have 
objected to installing sh or bash even earlier. It's just a shell after 
all, pretty harmless.


What Orr is arguing is valid, IMO, when teaching compiled programming 
languages. For interpreted languages, I'm not sure I agree, and for 
shell scripts I definitely disagree. The idea of shell scripting is 
'sewing together' the numerous tools available on a system. When 
teaching people to use shell scripts, it is not your intention to make 
them learn to squeeze code/mem usage/etc., but rather to get things done 
in a simple manner on typical (UNIXish) machines without resorting to 
writing complex C or C++ programs (which is in theory the only other 
kind of programming the students are capable of when taking MATAM). Now, 
I agree that telling them to employ perl or awk is overkill, since 
they're not supposed to be learning yet another scripting language. But 
standard binaries with reasonably rudimentary syntax should certainly be 
in their toolbox. In this respect, teaching shell scripting is also the 
teaching of use of nice UNIX tools.

I don't accept the "find isn't available on Windows", since MATAM is 
UNIX-focused even regardless of shell scripting, plus I'm guessing 
there's probably some find-like binary in Windows, just like 
C:\Windows\System32\find.exe is a sort-of-a-grep.

More information about the Haifux mailing list