Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!cornell!uw-beaver!teknowledge-vaxc!sri-unix!quintus!ok
From: ok@quintus.uucp (Richard A. O'Keefe)
Newsgroups: comp.lang.fortran
Subject: Re: Fortran vs C for computations
Message-ID: <391@quintus.UUCP>
Date: 13 Sep 88 09:09:22 GMT
References: <1226@scolex> <3448@lanl.gov>
Sender: news@quintus.UUCP
Reply-To: ok@quintus.UUCP (Richard A. O'Keefe)
Organization: Quintus Computer Systems, Inc.
Lines: 30

In article <3448@lanl.gov> jlg@lanl.gov (Jim Giles) writes:
>People keep saying that Fortran has no preprocessor.  This is _not_
>true.  Preprocessing is a textual substitution problem.  It can be done
>on _any_ sufficiently powerful text editor.  Not only that, cpp can usually
>be applied to any text file you have - including Fortran code.  A pre-
>processor _is_ unnecessary in a language standard - in fact, standardizing
>would only complicate the compiler writer's job without actually enhancing
>the language at all.

(1) cpp need not exist on every system.
    It need not even exist on a system which has a C compiler.
    [For example, the DEC-10 C compiler did not have a separate cpp.]

(2) cpp may be applied to any text file.  True.
    cpp may be USEFULLY applied to Fortran code.  FALSE.

    For example, consider the Fortran statement
	+0000000001111111111...77777777778+
	+12345C7890123456789...01234567890+
	|      X = 'THIS IS ...ONG00123000|
	|     & FORTRAN STRI...'  00124000|
    cpp doesn't know about Fortran continuation rules or sequence numbers.

(3) If I develop a program on machine X and then later try to compile it
    on machine Y I need the *SAME* preprocessor on both machines.  Giles
    is right that it doesn't have to be in the _language_ standard, but
    there _does_ have to be a preprocessor standard.  Does anyone wonder
    why I treasure my copy of the public-domain Ratfor sources, or of the
    public-domain M4?  Does anyone wonder why I _haven't_ bothered to get
    my own copy of cpp?