Re: one-liner for characater replacement

On May 26, 5:00 pm, "James Giles" <jamesgi...@xxxxxxxxxxxxxxxx> wrote:
Gary Scott wrote:
[...]    I see it attracting more users not fewer as it finally
acquires essential features that pulled users away in the first
place.  [...]

I actually don't see that happening.  Switching languages is
a big dislocation (that's the proper economic term for a big
change that costs a lot to accommodate).  People that I know
who abandoned Fortran have no interest in ever coming back.
The only thing that would lure them would be some market
leading, not playing "me-too" with features they already
have in whatever language they switched to.

To be sure, in large institutions with lots of "linguistic inertia",
they haven't abandoned Fortran (not wholly anyway) like their
more nimble colleagues outside.

Would it be accurate to say the the large institutions you have in
mind are non-profit ones (universities, government funded research
orgs etc.) ?


 That's where interest in new
Fortran features seems the highest.  That, I suspect, is where
vendors see  a future market for Fortran compilers.

rather like the Readers' Digest - the only reason its still around is
that its subscribers are too senile to remember to send in their
cancellation.

I suspect that the vendors are humoring the remaining fortran diehards
as loss -leaders - they'll give you the newer fortrans for free if you
buy something else (the box, or middleware or something).



[...] Given the need to remain backward compatible, it will never achieve
true beauty, but at least you can now look it in the face without
laughing.<snip>

Some of the features are either comedy or tragedy.  Between

That something is seriously wrong with where Fortran is headed is like
Peak Oil - some early pioneers saw it, the established kicked and
screamed in denial - but now the signs are too clearly visible to keep
up the party line.

My top candidate for Fortran tragi-comedy is the "and then" operator
- i had huge fun at its expense when it was proposed - but I guess
the laugh is on me because it looks like it has been adopted.

But at least thats a small relatively harmless thing - some of the big
ticket items such as the stilted "me too" imitations of fads like OOP
seem directly intended to destroy what remains of the language.

15 and 20 years ago I predicted that by the time Fortran had
standardized OOP features in it, and implementations were
finally becoming available, the rest of the programming
community would be moving on to other things.  In those days
I regarded OOP and inheritance as interchangeable.  Well,
implementations of Fortran's inheritance features are now
becoming available.  And OOP (they've change the definition)
is moving away from inheritance as fast as they can.

I also predicted that the next paradigm after inheritance would
be generic programming.  Guess what the OOP people are now
embracing as the big new thing?  If we had gone directly for
generic programming 15 years ago, F2003 would now be
leading the pack.  At least some might find that worth a laugh.

Well, Fortran can't now be expected to have standardized generic
programming features - with implementations widely available - for
perhaps 10 years.  And I've lost the gift of prophecy.  I have no
idea what the market leadng paradigm will be 10 years from now.
I'd bet it isn't generic programming anymore.  So, Fortran will
be trailing the pack again.


Its actually not too late. A group can secede from the committee and
go "back to the future" and create an alternative evolutiuon path from
f77. I only wish I had the time to do it.

--
J. Giles

"I conclude that there are two ways of constructing a software
design: One way is to make it so simple that there are obviously
no deficiencies and the other way is to make it so complicated
that there are no obvious deficiencies."   --  C. A. R. Hoare

.

Relevant Pages

  • Re: Object oriented programming in Fortran 95
    ... Better support for OOP is one of the major features of Fortran 2003, ... and some of the new features have already been implemented in the NAG ... I downloaded the papers and they're really good. ...
    (comp.lang.fortran)
  • Re: OT: FORTRAN
    ... needed to support a more standard version of Fortran -- and at that, ... CDC always supported all ANSI Fortran features. ... They frequently added additional features and a lot of them became standard features in the next version of the ANSI standard; ... We always ran the ANSI standard Fortran tests everytime we made an update, even a bug fix, to the Fortran compiler. ...
    (comp.lang.pl1)
  • Re: one-liner for characater replacement
    ... acquires essential features that pulled users away in the first ... who abandoned Fortran have no interest in ever coming back. ... I regarded OOP and inheritance as interchangeable. ... be generic programming. ...
    (comp.lang.fortran)
  • Re: Object oriented programming in Fortran 95
    ... Better support for OOP is one of the major features of Fortran 2003, ... and some of the new features have already been implemented in the NAG ... Fortran compiler, as described at ...
    (comp.lang.fortran)
  • Re: A petition to J3 apropos FORTRANs future
    ... > Those features that individually were not decisive before ... > As long as non-programmer managers can mandate language ... FORTRAN has always been the ugly duckling once the ... Commercial users needed "dirty" things like database APIs, ...
    (comp.lang.fortran)