Path: utzoo!mnetor!uunet!lll-winken!lll-tis!ames!ll-xn!mit-eddie!uw-beaver!tikal!amc!pilchuck!dataio!bright
From: bright@Data-IO.COM (Walter Bright)
Newsgroups: comp.lang.c++
Subject: Re: List of differences between AT&T and GNU C++
Message-ID: <1531@dataio.Data-IO.COM>
Date: 22 Apr 88 18:02:28 GMT
References: <724@acornrc.UUCP>
Reply-To: bright@dataio.UUCP (Walter Bright)
Organization: Data I/O Corporation; Redmond, WA
Lines: 25

In article <724@acornrc.UUCP> bob@acornrc.UUCP (Bob Weissman) writes:
>(As long as I'm here, I'd like to flame AT&T for using .c as the source
>code file extension.  What a stupid idea.  Now you can't tell a C
>source file from a C++ source file just by looking at the filename.
>And you can't tell make(1) to make a program containing C and C++.  And
>each Makefile has to have a .c.o rule to override the default one.  Hey
>AT&T!  Why don't you take a tip from GNU and use .cc?)

The C++ world is developing a problem with C++ source file extensions. So
far, I have seen:

	.C	??	AT&T
	.c	.h	Guideleines
	.cc	??	Gnu
	.cxx	.hxx	Advantage
	.cpp	.hpp	Zortech

Argh! OK, here are the ground rules I propose:

	1. The extension shouldn't be the same as for C, so compilers and
	   make can tell the difference.
	2. The extension shouldn't depend on case sensitivity, so it
	   will work with case insensitive operating systems like VMS
	   and OS/2 and MS-DOS.
	3. The extension should be <= 3 chars, for operating system
	   portability.