Fortran vs Python - Newbie Question

[Beliavsky]

On Mar 27, 6:32 am, cla... at lairds.us (Cameron Laird) wrote:
> In article <1174958090.292094.168... at e65g2000hsc.googlegroups.com>,Beliavsky <beliav... at aol.com> wrote:
>
>                         .
>                         .
>                         .
>
>
>
> >Your experience with Fortran is dated -- see below.
>
> >> I'll be more clear:  Fortran itself is a distinguished
> >> language with many meritorious implementations.  It can be
> >> costly, though, finding the implementation you want/need
> >> for any specific environment.
>
> >Gfortran, which supports Fortran 95 and a little of Fortran 2003, is
> >part of GCC and is thus widely available. Binaries for g95, also based
> >on GCC, are available for more than a dozen platforms, including
> >Windows, Mac OS X, and Linux. I use both and consider only g95 mature,
> >but gfortran does produce faster programs. Intel's Fortran compilers
> >cost about $500 on Windows and Mac OS and $700 on Linux. It's not
> >free, but I would not call it costly for professional developers.
>
> >Speaking of money, gfortran and g95 have free manuals, the latter
> >available in six languages
> >http://ftp.g95.org/. Final drafts of Fortran standards, identical to
> >the official ISO standards, are freely available. The manual for Numpy
> >costs $40 per copy.
>
> My experience with Fortran is indeed dated.  However,
> I still work with university groups that balk at $500
> for valuable software--sometimes because of admini-
> strative conflicts with licensing (example:  the group
> needs an educational license that fits its team
> perfectly, but educational license have to be approved
> by a campus-wide office that involves the group in
> expenses uncovered by its grants, and ... complications
> ensue).  Intel's compiler, for example, is a great deal,
> and recognized as a trivial expense sometimes--but
> judged utterly impossible by a research group down a
> different corridor.
>
> My summary:  practical success depends on specific
> details, and specific details in the Fortran and Python
> worlds differ.
>
> Also, Beliavsky, thanks for your report on the pertinent
> Fortran compilers.  There *are* other proprietary Fortan
> compilers extant; do you expect them to fade away,
> leaving only g* and Intel, or are you simply remarking
> on those two as the (intellectual) market leaders?

My point was that a few years ago an advantage of Python+Numeric or
Octave or R over Fortran is that the former let one work at a much
higher level, if one restricted oneself to using only free tools. The
creation of g95 and gfortran has changed that somewhat, and the
existence of commercial compilers is a plus, since they can surpass
the free compilers in performance (Intel), functionality (creating
Windows GUI programs entirely in Fortran, for example) or diagnosing
errors (NAG, Lahey/Fujitsu and Salford/Silverfrost). A research group
could purchase a single license of a commercial compiler to use in
nightly builds but use the free compilers for development.

Which commercial compilers will fade away? Decent free compilers will
hurt the market for mediocre commercial ones, which may explain the
demise of the compiler from NA Software. The Fortran 2003 standard
adds many new features to Fortran 95, thus making big demands on
vendors, and I have heard that Lahey and Salford will not be upgrading
their compilers to the new standard. The active vendors appear to be
Absoft, IBM, Intel, Pathscale, Portland, and Sun.