(j3.2006) article in _Nature_

[Tobias Burnus]

Bill Long wrote:
> The comment about Fortran seemed odd.  What did they mean "would not 
> work on their machines"?  gfortran is free and works just about 
> everywhere. Did they mean that it could not be compiled with gfortran? 
> In my experience, Python is less available on systems than Fortran. 
> And why would someone trust the translation? 

Regarding the "would not work": My impression is that most scientific 
Fortran programs use that little system-dependent functions and 
typically also only common vendor extensions; thus, a Fortran program - 
however old - usually just works. By contrast, Python does not keep 
backward compatibility: There are some incompatibilities between Python 
2 and 3.


I also do not like the: "code in a newer, more transparent programming 
language - Python".

The "newer" is a fact - though it conveys the message "better". And I 
doubt that Python is really more transparent; a newly (re)written 
program usually looks better than one which has been developed for years.

Tobias,
who started using Fortran, when Python was 12 and  Fortran 46 years old.

PS: Is it on purpose that at http://j3-fortran.org/doc/meeting/193/ the 
updated documents are hidden in a tar ball ("meeting193.tar.gz")?