Acceptance of OpenOffice.org

[Paul Lussier]

Jerry Feldman <gaf at blu.org> writes:

> IMHO, the schools should teach the important aspects 
> of computer programming, including data structures early, then once a 
> student learns the first language, then transition that student to a 
> real-world language. 

My understand is that this is exactly how MIT has structured their CS
curriculum.  They don't teach "languages", they teach methodology,
concepts, and how to think about and solve problems.  From this point
of view, languages are more or less all the same, it's just the syntax
which differs.  Any given problem can be solved in any language,
what's important is not the syntax you use, but the data structures
and engineering approach.

Once you've figured out how to solve the problem, then worry about
syntax and language.  We've all seen Fortran written C and shell
scripts written in Perl (heck, I've recently seen Lisp written in
Perl!)  which seems to be proof, or at least anecdotal evidence, that
any language *can* be used to solve any problem.

This isn't to say that some languages aren't better fits for certain
types of problems.  I certainly wouldn't want to write a Web interface
in Fortran (or PHP for that matter ;) and I can't imagine trying to do
AI stuff in C.  Writing a Java app to munge /etc/passwd or convert
between /etc/hosts and DNS zone files seems like a big lose too.

IMO, I think MIT has the right approach.  Teach the concepts and ideas
and teach how to *think* about problems.  If you can master that, you
don't need to worry about language and syntactic sugar.
-- 

Seeya,
Paul