Re: "Modern Fortran Explained": announcement

"Daniel Carrera" <daniel@xxxxxxxxx> wrote in message
On 03/28/2011 04:18 PM, Beliavsky wrote:
If I were teaching programming to someone interested in physics or
engineering, I might start with a Python (with the SciPy package) or

I'm not a fan of Python's syntax, or SciPy's separation of matrices vs
narrays. So I'd vote for Octave.

I wish there was an Octave-like interactive program with a slightly more
Fortran-like syntax.

One such book is "Python Scripting
for Computational Science" (2008) by Hans Petter Langtangen, but the
Fortran code in that book uses the F77 style, because that's what the
Python wrapping tools in the book handle...

I've been thinking lately about how I would design a course on scientific
computing that taught Fortran. My first goal would be to force everyone
to use strictly modern Fortran, and I think you can do that with only
three simple rules:

1. The program must compile without warnings with gfortran -std=f2008

2. *ALL* procedures must be in modules.

3. Every module must start with "implicit none".

4. The COMMON statement is forbidden.


Email: my qname at domain dot com
Domain: qomputing