NetNews Usenet Archive 1992 #26/NN_1992_26.iso/spool/comp/lang/c/16435

Text File | 1992-11-12 | 3.6 KB | 76 lines

Newsgroups: comp.lang.c
Path: sparky!uunet!ferkel.ucsb.edu!taco!rock!concert!uvaarpa!darwin.sura.net!convex!constellation!iapa!ctrbdo
From: [e-mail]%mailhost.ecn.uoknor.edu (bryan d oakley)
Subject: Re: Reasons for using C vs. Fortran or vice/versa
In-Reply-To: [e-mail]'s message of Thu, 12 Nov 1992 13:59:01 GM
Message-ID: <[e-mail]%mailhost.ecn.uoknor.edu>
Organization: FAA / Mike Monroney Aeronautical Center
References: <[e-mail]>
Distribution: usa
Date: Thu, 12 Nov 1992 22:50:15 GMT
Lines: 63

> This may or may not be the place to ask this, but here I go anyway.
>
> I was wondering what advantages Fortran has over C(if any) and
> vice-versa.
>
> I know C is much better than Fortran in the I/O dept., but is there
> anything else that makes it a better language.
>
> The reason I ask is when was in college, we used C for our primary
> language.  But the company I work for, uses a lot of Fortran(mainly
> because it was written years ago and still works).  We do a lot of
> number crunching around here and I've always been told that Fortran
> was very good at that.
>
> Is Fortran actually better than C for number crunching routines?

Let the holy wars begin!  Off the top of my head (and boy does that
hurt :-) I say that FORTRAN has the advantage that there are millions
of lines of it laying around all over the place.  It's (arguably) easy
to use and yes, it's good at number crunching.

Each language has qualities that make it better than all other
languages for certain tasks.  For FORTRAN, numerical processing seems
to be a forte, I/O a liability.  The problem with C, IMHO, is that in
the hands of the uneducated programmer (or poorly educated programmer,
for which there are legions) is a very dangerous weapon.  As for
having to maintain code, I have seen many, MANY more lines of hard to
fathom C code than FORTRAN code, but then I've seem some very
difficult algorithms coded in a most elegant way in C.  FORTRAN
requires a bit more brute force.  And true, FORTRAN GOTO's are
notorius for making code unreadable.  But then, that's another story
altogether.  Anyone who uses GOTO's with any regularity gets what
he/she has coming.

Language choice depends on the task to be solved.  Where I work, we
support 150,000+ lines of good ol' FORTRAN, but if I have to write a
little program to process text or do windows or whatever I use C.  I
actually consider myself to be a FORTRAN programmer at heart and can
write very elegant (ahem :-) stuff with it.  I recently had a
discussion with some co-workers on what language we would rewrite our
software in, given the chance.  I ended up thinking that staying with
FORTRAN isn't such a bad idea.  The suitability of any language is
dependent on the task at hand.

I know this is comp.lang.c, but posting about FORTRAN in
comp.lang.fortran is like preaching to the choir.  For what it's
worth, here are three quotes taken from the net in the past couple of
days:

"I don't know what the computer language of the year 2000 will look
 like, but I know that it will be called FORTRAN. (Tony Hoare?)"

"FORTRAN was FORTRAN when C was a pup,
 FORTRAN will be FORTRAN when C's time is up. (Seymour Cray?)"

"Fortran was a mistake of innocence and naivete.  C was created by people
 that should have known better." (anon)

--
----------------------------------------------------------------------
Instrument Approach Procedures Automation              DOT/FAA/AMI-230
----------------------------------------------------------------------
Bryan D. Oakley                    [e-mail]