Choosing a programming language as a competitive tool

[Andrew Dalke]

Douglas Alan wrote:
>I just can't see this comparison of Lisp to Perl.  Lisp was the
>easiest language I ever learned, except for Python, which was about
>equivalently easy.
 ...
>Perl, on the other hand, I find to be the most atrocious thing on the
>planet.  It's completely baroque and convoluted.  Programs I wrote
>yesterday seem like I wrote them 20 years ago.

As a data point for comparison, I started learning Perl to do
CGI scripts.  I kinda figured out cgi-lib.pl and made it do the
basics of what I wanted it to do.  I then bought the Perl book
and after reading a few chapters understood it well enough to
finish what I needed to do.

I tried learning Lisp from an essay in Douglas Hofstadter's
"Metamagical Themas."  I learned to car, cdr and setq, and
that was about it.  I couldn't see how to apply that to problems
I wanted to solve (probably because the essays didn't talk
about I/O) so I put it aside for a while.

That while turned out to be a long time.  The only time I've
needed it since was to modify my .emacs file (which contains
a few magical phrases I've picked up over time) and once to
try out a bio related package (which apparently was for a
Lisp implementation different from what I had, and I couldn't
figure out how to modify it to get it to work.)

So while you say Lisp is easier to learn, I found it easier to
use Perl to get things done.  Since I learn best by doing
things and working on projects, that means for me Perl was
easier to learn - because there was more incentive for me to
learn it despite it's complexity and cruftyiness.

It's the same reason people use MS Windows over Unix, (or
Unix over VMS for you DEC lovers :)

                    Andrew
                    dalke at acm.org