Path: utzoo!mnetor!uunet!husc6!cmcl2!rutgers!mcnc!thorin!polk!leech
From: leech@polk.cs.unc.edu (Jonathan Leech)
Newsgroups: comp.lang.c++
Subject: Re: C++ source file names
Message-ID: <471@thorin.cs.unc.edu>
Date: 19 Dec 87 23:40:07 GMT
References: <593@inria.UUCP>
Sender: news@thorin.cs.unc.edu
Reply-To: leech@polk.UUCP (Jonathan Leech)
Organization: University Of North Carolina, Chapel Hill
Lines: 28
Keywords: .C versus .c
Summary:

Expires:

Sender:

Followup-To:

Distribution:


In article <593@inria.UUCP> shapiro@inria.UUCP (Marc Shapiro) writes:
>
>There is a proposal (by the FSF people) to use the .C (capital C) extension
>to name C++ source files, instead of .c.
>...
>The distributed
>CC could be prepared to accept either .c or .C in order not to break any
>exisiting programs, but future programs should only use the .C extension.

    The 1.20 version of CC already does this. Admittedly the
`basename $A .[cC]` syntax does not work correctly on many systems,
but this is easy to correct.

>What do you think?

    I think it's fine as long as C++ compilers always accept both
extensions. Otherwise, someone trying to move code from a
case-insensitive system like VMS or MS-DOS will lose.

    On a tangent, I was surprised that Michael Tiemann posted the GNU
C++ announcement to comp.emacs but not here (look it up yourself,
don't ask me for a copy). Has anyone tried using it for something
substantial such as compiling cfront?
--
    Jon Leech (leech@cs.unc.edu)    __@/
    ``Thus Mathematics helps / our brains and hands and feet
      and can make / a race of supermen out of us.''
	- The Education of T. C. Mits